Citation
Safety assessment of freeway merging and diverging influence areas based on conflict analysis of simulated traffic

Material Information

Title:
Safety assessment of freeway merging and diverging influence areas based on conflict analysis of simulated traffic
Creator:
Atamo, Markos Alito
Publication Date:
Language:
English
Physical Description:
1 electronic file : illustrations ;

Subjects

Subjects / Keywords:
Traffic safety -- Evaluation ( lcsh )
Roads -- Interchanges and intersections ( lcsh )
Roads -- Interchanges and intersections ( fast )
Traffic safety -- Evaluation ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references.
General Note:
ABSTRACT The safety of merging and diverging influence areas, intersections, interchanges, and other traffic facilities is assessed by tracking and analyzing police-reported motor vehicle crash records. Since the nature of road crashes is random and infrequent, this process is slow to reveal the need for remediation of the roadway design and traffic control strategy. Moreover, this process is not applicable to assess new designs that have not yet been built, or deployed in the real world. This study summarizes a technique combining micro-simulation and automated conflict analysis to assess the safety of traffic facilities without waiting for a statistically above-normal number of crashes to occur. The technique is also valuable in assessing the relative performance of one design versus another. Traffic Conflict Technique (TCT) is among the most common surrogate measures to study the safety of roadway facilities. The software referred to as Surrogate Safety Assessment Model (SSAM) was developed by FHWA and was used in this study. The SSAM software application was designed to perform statistical analysis of vehicle trajectory data. Trajectory data is the output from microscopic traffic simulation models. Among the various traffic simulation modeling tools, VISSIM was chosen because of its compatibility with SSAM, its versatility for analyzing networks of big sizes and its ability to provide users with the capability to model any type of geometric configurations. Forty-two merging and forty-two diverging influence areas in Colorado, USA, were modeled in VISSM and conflict analysis was performed using SSAM under AM-peak traffic conditions. Five field validation tests were conducted. Conflicts predicted by SSAM approach were compared with actual crash records at merging and diverging influence areas in all statistical validation tests. The results of the validation effort of safety assessment based on conflict analysis of simulated traffic suggested that this technique is recommended for safety assessment at merging and diverging locations.
Statement of Responsibility:
Markos Alito Atamo.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
856144334 ( OCLC )
ocn856144334

Auraria Membership

Aggregations:
Auraria Library
University of Colorado Denver

Downloads

This item has the following downloads:


Full Text
SAFETY ASSESSMENT OF FREEWAY MERGING AND DIVERGING
INFLUENCE AREAS BASED ON CONFLICT ANALYSIS
OF SIMULATED TRAFFIC
By
Markos Alito Atamo
B.Sc., Addis Ababa University, Ethiopia, 2002
M.Sc., Addis Ababa University, Ethiopia, 2005
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Civil Engineering
2012


2012 by Markos Alito Atamo
All rights reserved


This thesis for the Doctor of Philosophy degree by
Markos Alito Atamo
has been approved for the
College of Engineering and Applied Science
by
UP, 2 o l Z-
Date


Atamo, Markos Alito (Ph.D., Civil Engineering)
Safety Assessment of Freeway Merging and Diverging Influence Areas Based
on Conflict Analysis of Simulated Traffic
Thesis directed by Professor Bruce N. Janson.
ABSTRACT
The safety of merging and diverging influence areas, intersections,
interchanges, and other traffic facilities is assessed by tracking and analyzing
police-reported motor vehicle crash records. Since the nature of road crashes
is random and infrequent, this process is slow to reveal the need for
remediation of the roadway design and traffic control strategy. Moreover, this
process is not applicable to assess new designs that have not yet been built,
or deployed in the real world.
This study summarizes a technique combining micro-simulation and
automated conflict analysis to assess the safety of traffic facilities without
waiting for a statistically above-normal number of crashes to occur. The
technique is also valuable in assessing the relative performance of one
design versus another. Traffic Conflict Technique (TCT) is among the most
common surrogate measures to study the safety of roadway facilities. The
software referred to as Surrogate Safety Assessment Model (SSAM) was
developed by FHWA and was used in this study. The SSAM software


application was designed to perform statistical analysis of vehicle trajectory
data. Trajectory data is the output from microscopic traffic simulation models.
Among the various traffic simulation modeling tools, VISSIM was chosen
because of its compatibility with SSAM, its versatility for analyzing networks of
big sizes and its ability to provide users with the capability to model any type
of geometric configurations.
Forty-two merging and forty-two diverging influence areas in Colorado, USA,
were modeled in VISSM and conflict analysis was performed using SSAM
under AM-peak traffic conditions. Five field validation tests were conducted.
Conflicts predicted by SSAM approach were compared with actual crash
records at merging and diverging influence areas in all statistical validation
tests. The results of the validation effort of safety assessment based on
conflict analysis of simulated traffic suggested that this technique is
recommended for safety assessment at merging and diverging locations.
This abstract accurately represents the content of the candidates
dissertation. I recommend its publication.
Signed
Bruce N. Janson


DEDICATION
I dedicate this thesis to my loving and devoted wife, Azeb, for her unfaltering
support and understanding while I was completing it. I also dedicate this to
my daughters Yohana, Heldana, and Yosabet and to my parents who gave
me an appreciation of learning and taught me the value of perseverance and
resolve.


ACKNOWLEDGEMENT
My thanks to each of the members of my doctoral committee for their support
during the completion of this thesis.
Very special thanks to my advisor, Professor Janson, for his wise guidance,
instruction and unlimited support. I am very grateful for having the opportunity
to learn from him.


TABLE OF CONTENTS
Figures xiii
Tables xv
Chapter
1. Introduction................................................... 1
1.1 Background..................................................... 1
1.2 Traffic Safety Overview........................................ 2
1.3 Statement of the Problem....................................... 4
1.4 Research Objectives and Contributions to the
Transportation Industry........................................ 5
1.5 Glossary of Terms- Quick Reference Guide....................... 7
1.6 Study Flow Structure...................................... 11
2. Review of the Literature...................................... 13
2.1 Traffic Conflict Technique.................................... 13
2.2 Merging and Diverging Movements on Freeways................... 15
2.3 Survey of Modeling Data....................................... 16
2.4 Interchanges and Ramps........................................ 17
viii


22
23
25
27
29
32
35
36
38
42
42
43
44
49
51
51
51
51
55
Background of Safety on Freeway Ramps.............
Log-linear Regression Models......................
Poisson Regression Model..........................
Negative Binomial Regression Model................
Goodness of Fit Measure...........................
Traffic Crash Prediction Model....................
Empirical Bayes Technique.........................
Empirical Bayes Safety Estimate...................
Bayesian Identification of Accident-Prone Locations
Simulation Modeling...............................
Choice of a Simulation Tool.......................
VISSIM Micro-Simulation Tool......................
SSAM Conflict Analysis Tool.......................
SAS/STAT Software and the GENMOD Procedure..
Research Approach and Methodology.................
Field Validation..................................
Purpose...........................................
Data Assembly and Geometric Definition............
Simulation Modeling of Interchanges and
Modeling Assumptions..............................
IX


3.5 Identification and Removal of Outliers.................... 59
3.6 Methodology............................................... 60
3.6.1 Field Validation Test 1: Merging and Diverging
Location Ranking by Total Incidents....................... 60
3.6.1.1 Conflict Ranking.............................................. 61
3.6.1.2 Crash Ranking................................................. 61
3.6.1.3 Ranking Comparison............................................ 62
3.6.2 Field Validation Test 2: Merging and Diverging
Location Ranking by Specific Incident Types............... 63
3.6.3 Field Validation Test 3: Crash and Conflict
Prediction Regression Model Paired Comparison............. 64
3.6.4 Field Validation Test 4: Crash and Conflict
Prediction Regression Model Comparative
Analysis For Total Incidents.................................. 65
3.6.5 Field Validation Test 4: Crash and Conflict
Prediction Regression Model Comparative
Analysis For Specific Incident Types...................... 69
3.7 Goodness of Fit Measure....................................... 69
4. Test Results and Discussion................................... 72
X


4.1 Field Validation Test 1: Merging and Diverging
Location Ranking by Total Incidents............................ 75
4.2 Field Validation Test 2: Merging and Diverging Location
Ranking by Specific Incident Types............................. 76
4.3 Field Validation Test 3: Crash and Conflict
Prediction Regression Model Paired Comparison.................. 80
4.3.1 Illustration................................................... 91
4.4 Field Validation Test 4: Crash and Conflict Prediction
Regression Model Comparative Analysis
for Total Incidents........................................... 93
4.4.1 Development of Conflict Prediction Regression
Models for Total Incidents.................................... 93
4.4.2 Identification of Accident Prone Locations (APL).............. 95
4.4.3 Illustration.................................................. 96
4.4.4 Ranking Locations............................................. 97
4.5 Field Validation Test 5: Crash and Conflict
Prediction Regression Model Comparative
Analysis For Specific Incident Types....................... 101
4.5.1 Development of Conflict Prediction Regression
Models for Specific Incident Types......................... 101
XI


4.5.2 Identification of Accident Prone Locations (APL) to
Rear End Incidents.......................................... 104
4.5.3 Ranking Locations........................................... 104
5. Summary, Conclusions, and Recommendations................... 108
5.1 Summary and Conclusions..................................... 102
5.2 Recommendations............................................. 113
Appendix
A Interchange traffic flow information............................. 116
B Average 3 year crashes recorded at
merge and diverge locations......................... 117
C Average yearly crashes recorded at
merge and diverge locations......................... 118
D Conflicts of five replications at merge and diverge
locations recorded by SSAM software....................... 119
E Average hourly conflicts at merge and diverge
locations recorded by SSAM software....................... 120
Bibliography......................................................... 121
xii


LIST OF FIGURES
Figure
1.1 STUDY WORKFLOW STRUCTURE........................ 12
2.1 TYPICAL INTERCHANGE CONFIGURATIONS.............. 19
2.2 TYPICAL RAMP CONFIGURATIONS..................... 21
2.3 EVENT-FILE BASED INFORMATION
FLOW DIAGRAM................................... 45
2.4 ILLUSTRATION OF CONFLICT ANGLE DIAGRAM.......... 49
3.1 ILLUSTRATION OF MERGE INFLUENCE AREA............ 54
3.2 ILLUSTRATION OF DIVERGE INFLUENCE AREA.......... 54
3.3 SCREEN CAPTURE. VISSIM GEOMETRIC
MODEL OF I-25&E HAMPDEN AVE.................. 57
3.4 SCREEN CAPTURE. VISSIM TRAFFIC
MODEL OF I-25&E HAMPDEN AVE.................. 58
4.1 SCREEN CAPTURE. SSAM SCREEN-
xiii
CONFIGURATION TAB
73


4.2 SCREEN CAPTURE. SSAM SCREEN MAP TAB
VIEW AT 1-25 & E HAMPDEN AVE................... 74
4.3 SCREEN CAPTURE. VISSIM CONFLICT SCENARIO
RESULTING FROM LANE CHANGE MANEUVER............ 77
4.4 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT
AND MERGING ADT................................ 83
4.5 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT
AND DIVERGING ADT.............................. 84
4.6 TOTAL CRASH AS A FUNCTION OF TOTAL
CONFLICT AT MERGE.............................. 88
4.7 TOTAL CRASH AS A FUNCTION OF TOTAL
CONFLICT AT DIVERGE............................ 89
XIV


LIST OF TABLES
Table
4.1 Distribution of crashes and conflicts by incident type.......... 79
4.2 Correlation coefficients by incident type....................... 80
4.3 Prediction model of total crashes as a function of
mainline ADT and merging ADT.................................... 82
4.4 Prediction model of total crashes as a function of
mainline ADT and diverging ADT.................................. 82
4.5 Prediction model of total crashes as a function of
conflicts at merge................................................... 86
4.6 Prediction model of total crashes as a function of
conflicts at diverge................................................. 86
4.7 Prediction model of conflicts as a function of
mainline PHV and merging PHV......................................... g4
4.8 Prediction model of conflicts as a function of
mainline PHV and diverging PHV................................... 94
XV


4.9 Comparison between total crash and total conflict
prone locations at merge..................................... 98
4.10 Comparison between total crash and total conflict
prone locations at diverge................................... 99
4.11 Summary of APLs of total incidents........................... 100
4.12 Average total crashes and total conflicts
based on incident proneness of locations..................... 100
4.13 Prediction model of rear-end crashes as a
function of mainline ADT and merging ADT..................... 102
4.14 Prediction model of rear-end crashes as a
function of mainline PHV and merging PHV..................... 102
4.15 Prediction model of rear-end crashes as a
function of mainline ADT and diverging ADT................... 103
4.16 Prediction model of rear-end crashes as a
function of mainline PHV and diverging PHV................... 103
4.17 Comparison between rear-end crash and rear-end
conflict prone locations at merge............................ 105
4.18 Comparison between rear-end crash and rear-end
conflict prone locations at diverge.......................... 106
XVI


4.19 Summary of APLs of rear-end incidents........................ 107
4.20 Average rear-end crashes and rear-end conflicts
based on incident proneness of locations..................... 107
xvii


1. Introduction
1.1 Background
According to the National Highway Traffic Safety Administration (NHTSA
2006) report, motor vehicle travel is the primary means of transportation in the
United States, providing an unprecedented degree of mobility. Despite all its
advantages, deaths and injuries resulting from motor vehicle crashes are the
leading cause of death for people of every age from 2 through 34. Traffic
fatalities accounted for more than 90 percent of transportation-related
fatalities.
National Highway Traffic Safety Administration (NHTSA 2009) stated that in
the United States more than 5.5million police-reported motor vehicle crashes
occurred. Of these 1.5million (28%) crashes resulted in an injury, 30,797
(fewer than 1%) resulted in a death and 3.4million (73%) involved in property
damage only. The fatality rate per 100 million VMT was 1.03.
There were 8.6million vehicles involved in single and two-vehicle crashes.
The number of vehicles involved in a merging/lane changing maneuver were
312,000 (3.6%). Resulting from these maneuvers were 769 fatalities, 48,000
injury, and 263,000 property damage only crashes. Similarly, the number of
vehicles involved in a going straight maneuver was 4.3million (50.4%).
Resulting from these were 27, 124 fatalities 1.25million injury, and 3.0million
property damage only.
l


The World Health Organization (2004) projected that between 2004 and 2030
global deaths increase by 28% which is predominantly due to the increasing
number of road traffic accident deaths. Road traffic accident deaths are
projected to increase from 1.3 million in 2004 to 2.4 million in 2030, primarily
due to the increased motor vehicle ownership and use associated with
economic growth in low- and middle-income countries. Road traffic accidents
are projected to rise from the ninth leading cause of death globally in 2004 to
the fifth in 2030. According to the United Nations General Assembly (2008),
by 2020 road traffic deaths and injuries will exceed HIV/AIDS as a burden of
death and disability.
1.2 Traffic Safety Overview
A traffic accident occurs when a road vehicle collides with any physical object
and results in injury, property damage, and death. Factors contribute to the
risk of collision include; vehicle design, speed of operation, road design, and
driver impairment. Worldwide motor vehicle collisions lead to significant death
and disability as well as significant financial costs to both society and the
individual.
For the last two decades, several legislations have been making safety a
central, explicit, comprehensive, and integrated part of transportation
planning. As a result of this, safety management systems have advanced to a
great extent. Traffic safety data and analytical tools have been improved and
refined, and the effects of countermeasures have become better understood.
2


Since 1991 there have been three major Acts announced in the Unites States
of America: (1) the 1991 Intermodal Surface Transportation Efficiency Act
(ISTEA); (2) the 1998 Transportation Equity Act for the 21st Century (TEA-21),
and (3) the 2005 Safe, Accountable, Flexible, Efficient Transportation Equity
Act (SAFETEA-LU.)
The Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991 moved
the historical focus of highway and transit programs away from construction,
capacity, and congestion. The Act changed the emphasis towards mobility
and access, system performance, and consideration for the environment and
quality of life.
The Act did not specifically mention safety as part of the planning process but
mandated six comprehensive management systems including a Safety
Management System (SMS) as a prerequisite for funding. The SMS was part
of the strategy to improve the management, operations, and safety of the
highway system through improved data analysis and collection, through
improved coordination, cooperation, and communication among agencies,
and through the development of collaborative strategic plans (Depue 2003).
In 1998, the Transportation Equity Act for the 21st Century (TEA-21) called
for comprehensive safety consciousness. The Act required state DOTs (and
MPOs) to increase the safety and security of the transportation system for
motorized and non-motorized users. This was the first time that safety
became an explicit part of transportation plans. Prior to TEA-21, safety was
sometimes a prominent factor in project development and design, but this
legislation calls for safety consciousness in a more comprehensive, system-
wide, multimodal context, (FHWA 2001a).
3


The Safe, Accountable, Flexible, Efficient Transportation Equity Act
(SAFETEA-LU) became law in 2005. SAFETEA-LU has a strong focus on
integrated and comprehensive safety planning. It was built on previous
legislation in giving specific and increasing recognition to safety issues.
The Act establishes the Flighway Safety Improvement Program (HSIP) as a
core program and nearly doubles the funds available for infrastructure safety
and comprehensive strategic highway safety planning. The purpose of the
HSIP is to reduce fatal and serious/life changing crashes. The program
includes planning, implementation, and evaluation of safety programs and
projects. (Bahar, G., and Morris, N. (2007)).
Generally, these legislations have encouraged or insisted on giving safety
priority through better data, better analysis, better reporting systems, and the
adoption of a structured comprehensive approach. A detailed study of the
safety of highways would help to understand safety and help to make safety
legislation a success.
1.3 Statement of the Problem
Until recently the safety of merging and diverging influence areas,
intersections, interchanges, and other traffic facilities is most often
assessed by tracking and analyzing police-reported motor vehicle crashes
over time. Since the nature of crashes is random and infrequent, the
process of analysis of safety based on the police report is slow to reveal
the need for remediation of either the roadway design or the flow-control
strategy. This process is also not applicable to assess new designs that
4


have yet to be built, or to assess new flow control strategies before they
are employed on-site.
This study summarizes a technique combining micro-simulation and
automated conflict analysis to assess the safety of traffic facilities without
waiting for a statistically above-normal number of crashes to occur. The
technique is also valuable in assessing the relative performance of one
design versus another in comparing alternative designs.
1.4 Research Objectives and Contributions to the Transportation
Industry
The objective of this research was to assess the safety of freeway merging
and diverging influence areas using conflict analysis technique and compare
the results with the actual crash experience. SSAM conflict analysis software
was used for predicting safety performances at the predefined locations. The
Surrogate Safety Assessment Model (SSAM) is a software application
designed to perform statistical analysis of vehicle trajectory data output from
microscopic traffic simulation model, in this study VISSIM.
VISSIM is a traffic simulation modeling tool which was chosen because of its
compatibility with SSAM, its versatility for analyzing networks of big sizes and
its ability to provide users with the capability to model any type of geometric
configurations or unique operation/driver behavior encountered within the
transportation system. It is a time-based microscopic simulation tool that uses
various driver behavior and vehicle performance models to accurately
5


represent urban traffic and public transport operations. This study identified
areas in need of further research.
The study will make the following contributions to the industry:
Helps in assessing the relative safety performance of one design versus
another.
Provide transportation agencies, planners, engineers, researchers with
information about the type, frequency and relative location of conflicts.
Provide solution that improves safety, access, and mobility of the
merge/diverge influence area.
Provide the safety analyst with a fast report to reveal the need for
remediation of the roadway design.
6


1.5 Glossary of Terms Quick Reference Guide
Accident (traffic): Event between road-users that results in injury, fatality
or property damage.
Accident causation: Underlying reasons that pre-empt a traffic accident,
most usually involving an unforeseen chain-of-events. Accident causation is
often attributed to one of the three main components of the traffic system:
road-user, vehicle or roadway, or a combination of thereof.
Accident outcome: Result of an accident in terms of injury severity, fatality
and in some cases also property damage.
Accident rate: Number of accidents in accordance with a measure of
exposure.
Accident severity: Level of injury sustained in a traffic accident: usually
categorized as minor, serious or fatal.
Calibration: Process used in Traffic Simulation to (statistically) ensure that
the functioning and behavior of a particular model and/or sub-model
corresponds with observed empirical measurements or predetermined values.
Car-following: Term used to describe the status of a vehicle that has a
time/distance gap or headway less than a predetermined maximum value.
Collision: Impact event between two or more road-users/vehicles, or a road-
user (vehicle) and stationary object.
7


Collision course: Existence of a common projected conflict point in time and
space for two (or more) road-users/vehicles, usually based on momentary
measures of trajectory, speed and distance.
Conflict: An observable situation in which two or more road users approach
each other in time and space to such an extent that there is risk of collision if
their movements remain unchanged.
Conflict distance: A momentary measurement of (spatial) distance to a
common conflict point for a road-user/vehicle in a conflict situation.
Conflict point: Common spatial location of projected trajectories given
momentary measures of speed and distance for two or more road-
users/vehicles.
Conflict severity: Seriousness of a potential collision or near-accident
measured by temporal or spatial proximity.
Conflict speed: Momentary measurement of velocity for a road-user (vehicle)
in a conflict situation.
Conflict zone: Common area used by road-users/vehicles approaching from
different trajectories.
Crash: Term that is sometimes preferred to (traffic) accident due to the fact
that it implies an element of causality rather than an unforeseen random
occurrence.
8


Driver behavior: Largely misused and over-simplified term used in traffic
engineering that is used to describe the actions and/or variability of drivers in
different driving situations. It should relate to the study of individual behavioral
processes that underlie driver actions (performance).
Exposure: Measure of spatial or temporal duration in the traffic system in
relation to the number of dynamic system objects road-users, vehicles
(axles), etc.
Fatality: Death resulting from a traffic accident (usually within a 30 day period
after the accident occurrence).
Gap-acceptance: Process that describes and measures interaction between
prioritized and non-prioritized road-users. Generally involves spatial or
temporal measurement of gaps or lags in prioritized streams that are
accepted or rejected in relation to a particular yielding maneuver.
Incident: the term used in a more abstract sense to refer to either crashes or
conflicts.
Injury accidents: Traffic accidents that result in minor or serious injury to one
or more parties.
Macroscopic (macro-) simulation: Simulation at a less detailed
(aggregated, macroscopic) level.
Microscopic (micro-) simulation: Simulation at a very detailed (microscopic)
level.
9


PET: It is the minimum post-encroachment time observed during the conflict.
Post encroachment time is the time between when the first vehicle last
occupied a position and the time when the second vehicle subsequently
arrived to the same position. A value of zero indicates a collision. A post-
encroachment time is associated with each time step during a conflict. A
conflict event is concluded when the final PET value is recorded at the last
location where a time-to-collision value was still below the critical threshold
value. This value is recorded in seconds.
Police reported accidents: Accidents that are reported by the police and are
recorded in the accident database of accident.
Safety: Freedom from accident or loss.
Simulation (traffic): Abstract imitation of the operation of a real-world
process or system over time and event occurrence. Traffic simulation is
concerned with the modeling of processes in the traffic system and can be
conducted at different levels of abstraction depending on the purpose of the
study.
TTC: It is the minimum time-to-collision value observed during the conflict.
This estimate is based on the current location, speed, and future trajectory of
two vehicles at a given instant. A TTC value is defined for each time step
during the conflict event. A conflict event is concluded after the TTC value
rises back above the critical threshold value. This value is recorded in
seconds.
10


1.6 Study Workflow Structure
Figure 1.1 illustrates Study Workflow Structure. Accordingly, this research is
divided into three Parts. Part 1 included introduction, information and data
search. It consists of Chapter 1 and Chapter 2. Chapter 1: Introduction, it
provided an introduction to the study by presenting the background
information, traffic safety overview, statement of the problem, study objectives
and contributions to the transportation industry; and glossary of terms.
Chapter 2: Literature Review: a comprehensive literature review on the safety
performance of merging and diverging influence areas from published
information was provided. It presented the overview of the merging and
diverging movements on freeways, the survey and collection of available
modeling data, background of safety on freeway speed change locations,
regression models, Bayesian accident-prone locations, and simulation and
conflict analysis tools. Also preliminary qualitative analysis of the data was
carried out to identify the general validity of the selected variables. Part 2
included data analysis. It consists of Chapter 3: Research approach and
methodology, this section described the study methodology used in the field
validation test and the steps taken to complete the study. Part 3 consists of
Chapter 4 and Chapter 5. Chapter 4: Test results and discussion, this section
validated the actual crash data and discussed the results of the analysis.
Chapter 5: Presented the summary, conclusions and recommendations
drawn from the results of the study.
li


FIGURE 1.1 STUDY WORKFLOW STRUCTURE
12


2. Review of the Literature
2.1 Traffic Conflict Technique
According to Amundsen and Hyden (1977), a conflict is defined as an
observable situation in which two or more road users approach each other in
time and space to such an extent that there is risk of collision if their
movements remain unchanged.
The traffic conflict technique method initially developed at the Detroit General
Motors laboratory in the late 1960s for identifying safety problems (Perkins
and Harris, 1968). It was based on observer judgments using time-lapse
filming, and proves a costly and time consuming technique. They correlated
conflict patterns to accident types. Spicer (1973) studied about the safety of
six intersections and concluded that there was good correlation between
serious conflicts and reported injury accidents. Sayed (1997) conducted a
study to estimate the safety of un- signalized intersections using traffic safety
technique. A computer simulation model was used to study traffic conflicts
with time-to-collision as the critical traffic event in simulating driver behavior.
Using the data collected from 30 conflict surveys, he established traffic
conflict frequency and severity standards for un-signalized intersections.
These standards allow the relative comparison of the conflict risk at various
intersections. Miglez et al. (1985) developed a methodology to predict traffic
accidents from observed conflicts. Also, statistical procedures were
developed to determine conflict rate values that could be considered
"abnormally" high. The study concluded that traffic conflicts are good
13


surrogates of accidents in that they produce estimates of average accident
rates nearly as accurate as those produced from historical accident data.
Gettman, Pu, Sayed, and Shelby (2008), studied a method of safety
assessment utilizing a traffic conflicts analysis technique applied to simulation
models of intersections, interchanges, and roundabouts. They developed a
software application to automate the task of traffic conflicts analysis and
conducted validation testing to gauge the efficacy of the assessment method.
Several surrogate traffic conflict measures have been developed to assess
the safety of highway facility. The common surrogate conflict measures are
gap time, encroachment time, deceleration rate, proportion of stopping
distance, post-encroachment time, initially attempted post-encroachment
time, and time-to-collision.
There is general consensus that higher rates of traffic conflicts can indicate
lower levels of safety for a particular facility, given the fact that conflicts
generally result from a lack or misunderstanding of communication between
the different road users (Risser 1985 and Archer 2000)
In this study, traffic conflicts occur with adequate frequency to overcome the
statistical challenges posed by infrequent crashes. Also, since adequate
conflict data can be collected in a relatively short time, conflict analysis is not
subject to the problem of changing of the underlying conditions (e.g., traffic
volumes, geometry) that affect long-term crash records.
14


2.2 Merging and Diverging Movements on Freeways
According to Roess, R., Prassas, E., and McShane W. (2004), merging
occurs when two separate traffic streams join to form a single stream.
Merging can occur at on-ramps to freeways or multi-lane highways, or when
two significant facilities join to form a single traffic stream. Merging vehicles
often make lane changes to align themselves in lanes appropriate to their
desired movement. Diverging occurs when one traffic stream separates to
form two separate traffic streams. This occurs at off-ramps from freeways and
multilane highways, but can also occur when a major facility splits to form two
separate facilities. Again, diverging vehicles must properly align themselves in
appropriate lanes, thus indicating lane-changing; non-diverging vehicles also
make lane changes to avoid the turbulence created by diverge maneuvers.
Therefore, a merge area would always follow an on-ramp whereas a diverge
area would always precede an off-ramp. Road users in such areas are
required to select and adjust speeds without causing undue hazards to
themselves or to other road users. Such hazards would normally increase in
terms of frequency and intensity with increasing volumes of vehicles entering
and exiting a specific segment on the freeway. This increase may result from
the combination of increasing number of required lane changes and
decreasing probability of a suitable gap between moving vehicles to perform a
merge or diverge operation (Sarhan, M.; Hassan, Y.; and Abd El Halim, A.O.
2008).
15


2.3 Survey of Modeling Data
Since the nature of road crashes is random, the choice of the analysis period
has a significant impact on the accuracy and reliability of the safety
assessment. Accident counts that are recorded for overly extended time
period may introduce biases in the analysis when current traffic and
geometric conditions differ from those prevailing when the crashes occurred.
Similarly, an overly short period reduces the number of crashes considered
and hence the statistical accuracy of the model would be affected. According
to the Permanent International Association of Road Congresses PIARC
(2003) Road Safety Manual, the minimum accepted analysis period is 3
years. For this study, a three year data from year 2008 through 2010 was
used which was obtained from the Colorado Department of Transportation
(CDOT) accident database.
The greatest statistical challenge of the safety assessment of a facility in
accurately pinpointing an underlying crash rate is the nature of motor vehicle
crashes. Crashes are so infrequent and exhibiting variable hourly and yearly
crash counts. The problem is increasingly difficult for areas with lower (more
infrequent) crash rates. There is possible bias in comparing peak-hour
conflicts to annual crashes. The statistical models that correlate the peak-
hour crash with the peak hour conflict on the merging and diverging locations
may not converge as there are too few peak-hour crashes (which are less
than 25% of the total crashes). Field validation through estimation with a
larger data set would result in a more significant parameters estimates and
more accurate crash prediction.
16


Twenty-one interchanges, forty-two merging and forty-two diverging influence
areas were selected. A preliminary assessment was made of the types of
interchange elements that were in sufficient numbers and had sufficient data
available for statistical modeling of accidents to be conducted. Accordingly
diamond interchange configurations were selected.
Using each interchanges milepost as a common identifier, modeling data was
collected for each interchange taking the following into consideration:
1. Interchange layout: number of lanes on mainline and ramps, ramp
length, and ramp configuration.
2. Traffic volumes: average daily traffic (ADT), design hourly volume
(DHV), percent trucks, and on-ramp hourly traffic volume recorded by
ramp meters.
3. Data on crash frequency on mainline and ramps.
4. Location of crash occurred on the highway. Only crashes at the merge
and diverge influence areas of the interchange were considered.
5. Accident Type: accidents were regrouped into three: rear-end, crossing
and lane-changing. Only vehicle-to-vehicle accidents were considered.
6. Direction: the direction of traffic flow at the time of crash: North,
Northeast, Northwest, East, West, South, Southeast, Southwest.
2.4 Interchanges and Ramps
Grade separation of intersecting roadways reduces crashes caused by
crossing and turning movements. An interchange is a system of
interconnecting roadways in conjunction with one or more grade separations
that provides for the movement of traffic between two or more roadways or
highways on different levels. Access control is a highly desirable feature
17


along the crossroad at an interchange to provide efficient traffic operations
and safety. The most common roadways connections at an interchange
consist of freeway-to-freeway (access-controlled freeways), freeway-to-
arterial and arterial-to-arterial. Freeway-to-arterial and arterial-to-arterial
connections are non-access-controlled highways. This research focused
exclusively on a freeway-to-arterial interchanges.
There are several basic interchange configurations to accommodate turning
movements at a grade separation and it is determined by number of
intersection legs, expected volumes of through and turning movements,
composition and type of traffic, topography, design control, proper signing,
and physical constraints such as existing rivers, railroads, and roadways.
All freeway interchanges with non-access-controlled highways should provide
ramps to serve all basic directions to prevent wrong-way movements.
According to AASHTO 2004, the term ramp includes all types,
arrangements, and sizes of turning roadways that connect two or more legs at
an interchange. The different ramp patterns of an interchange (i.e. the
different types of interchange configurations) are made up of various
combinations of ramps. Figure 2.1 illustrates typical interchange
configurations: from the simplest full-diamond interchange to complex, multi-
level directional interchanges. Many variations of each of these basic
interchange configurations are possible.
18


(4-Quad)
i / \ t
\
Full Diamond
with Slip Ramps
ZL
Partial Cloverleaf (Parclo)
h\
'
Full Cloverleaf with
Collector/Distributor
Roads
A U
\ r
Directional
FIGURE 2.1 TYPICAL INTERCHANGE CONFIGURATIONS (BAUER, M.
AND HARWOOD, W 1998)
19


A one way road that leaves a mainline highway facility is called an off-ramp or
exit ramp. A one way road that joins a mainline highway facility is called as an
on-ramp or entrance ramp. This distinction is important because according to
Cirillo (1967) & Lundy (1966), the accident rate of on-ramps was consistently
lower than off-ramp accident rates. Accordingly an on-ramp accident of 0.59
accident per MVMT and off-ramp accident of 0.95 accident per MVTM have
been confirmed. Vehicles typically travel along off-ramps at higher speeds
than along on-ramps, so that accidents are more likely to occur on off-ramps.
Figure 2.2, illustrates a number of typical ramp configurations. Each of the
ramps of the ramp configurations illustrated traffic exiting from a mainline
freeway, but an analogous ramp configuration for traffic entering the mainline
freeway also exists.
20


Direct Connection*
Semi-Direct
Connection*
(to two-way
frontage road)
(to two-way
frontage road)
(to one-way
frontage road)
* When used in directional Interchanges
b Scissors connection
FIGURE 2.2 TYPICAL RAMP CONFIGURATIONS (BAUER, M. AND
HARWOOD, W 1998)
21


2.5 Background of Safety on Freeway Ramps
Numerous studies have studied the safety performance of freeway ramps
during the past several decades. Bared, J., Giering G., Warren, D., (1999)
developed the statistical model of accidents to estimate accident frequencies
for entire ramps as a function of speed change lane length among other
variables. According to the accident model developed in that study, the longer
speed change lane showed the less accident frequency.
The result of the study by Lord, D., Bonneson, J., (2005), showed that the exit
ramps are more dangerous and the non-free-flow type ramp experience twice
as many accidents as other types of ramps. Previous researchers have
developed several crash prediction models to relate crash frequency at ramp
sections to different explanatory variables such as traffic volumes and ramp
design elements.
Bauer, M, Harwood, W. (1998), studied the relationship between traffic
crashes and highway geometric design elements and traffic volumes for
interchange ramps and speed change lanes. The statistical modeling
approaches used in that research included Poisson and negative binomial
regression. Several models were developed to predict crashes on ramp
sections and speed change lanes. The variables that were included in crash
models included mainline freeway AADT, ramp AADT, area type
(rural/urban), and ramp type (on/off), ramp configuration, right shoulder width,
and lengths of ramp and speed-change lane. The regression model is
presented in Section 2.6 of this study.
Accident rates decrease as length of weaving area, or acceleration and
deceleration lane increases (Cirillo 1967, 1968, 1970). Cirillo developed
statistical model and provided a relationship between the length of weaving
22


and acceleration and deceleration lanes and traffic volume levels and
percentage of merging and diverging traffic. Increase in traffic volume was
associated with an increase in crash. The effect of acceleration lane length on
accident rate was significant when merging traffic percentage exceeded 6%
of the main line traffic volume. As the percentage of merging traffic increase
beyond this volume, the additional length of acceleration lane provides a
significant reduction in accidents. The effect was not as great for deceleration
lanes compared to acceleration lanes.
Studies suggested that a good portion of the accidents associated with ramps
occur at the entrance and exit of ramps. Mullins and Keese (1961) found that
23% of all through-lane accidents occur near entrance terminals. 52% of the
on-ramp accidents according to Lundy (1966) occur at entrance terminal.
McCartt, A., Northrup, V., Retting, R. (2004) studied about the types and
characteristics of ramp-related motor vehicle crashes on urban interstate
roadways in Northern Virginia. The study found that 48% accidents occurred
in the process of exiting freeway, 36% occurred when the traffic entering and
16% occurred in the midpoint.
2.6 Log-linear Regression Models
Bauer, M. and Harwood, W. (1998) developed statistical model for
interchange. These models are Log-linear regression models. They included
Poisson and negative binomial regression models. Statistical background on
both the Poisson and negative binomial models is provided next.
23


The following notations are used:
n = interchange elements (e.g., ramp proper segments, entire ramps),
i = is a set of q parameters (xn ,Xi2,...,Xiq ) (e.g. the geometric design,
traffic volume, and other related characteristics of that element.
Yt= the number of accidents occurring at the ith element during a specific
period (say 3-year): where i = 1,2, ...,n.
yj= the actual observation of y{ during the 3-year period, where
Yi = 0,1,2,... and i = 1,2,..., n.
E(Yi) = |ii = the expected number of accidents at the ith element and the
q parameters,
Note: In section 2.5 of this study all logarithms are natural logarithms and
are denoted by log.
The main objective of a statistical model is to develop a relationship E(Yi) =
|ii and q parameters, X{1 ,Xi2, ....,Xiq.
This relationship can be formulated through a general linear model (GLM) of
the form:
Function (^) = (B0 + Mil + + (3qXiq
(2.1)
24


where:
(B0, Pi, p2 -Pq, are the regression coefficients and estimated from the
data.
The estimation procedure used to obtain the regression coefficients is
dependent on the assumption made about the distribution of the Yj,
2.6.1 Poisson Regression Model
The assumption of a log-normal distribution [i.e., the assumption that log(Yi)
follows a normal distribution] is not valid when the average number of
accidents at a ramp is small. The Poisson model then becomes a natural
choice as it models the occurrence of rare discrete events well. The Poisson
distribution has a mean and variance, a2 both equal to |.
According to the model formulated by Bauer and Harwood (1998):
Assuming the number of accidents,^, follows a Poisson distribution with
mean |, the probability that a ramp with y} accidents can be expressed as
q
(2.2)
25


P(Y; = y;; w) =
(2.3)
where: yp denotes the factorial of yi.
Maximum likelihood estimation method is used to estimate the Poisson
regression coefficients (B0 P!,...,pq. Maximization of the likelihood function is
performed by taking the derivative of the likelihood function with respect to the
parameter value. However, it is convenient to take the natural log of the
likelihood function before differentiating. Since the natural log is a monotonic
function, maximizing the log-likelihood is equivalent to maximizing the
likelihood.
So, the likelihood is:
The log-likelihood over all n interchange elements in the category of elements
is therefore:
n
(2.4)
log(L) = -n\i + (y1 + + yn) log(g) log^! x ... x xn!)
(2.5)
26


Which is equivalent to:
log(L) = ^[y; logOO |ii log(yi!)]
i=l
(2.6)
2.6.2 Negative Binomial Regression Model
As it has been discussed in Section 2.5.1, in Poisson distribution the mean
and the variance of the distribution are equal. However, previous traffic
accident research has shown that this is not always the case.
Paternoster and Brame, (1997); Osgood, (2000) suggested the negative
binomial distribution as an alternative to the Poisson when there is evidence
of overdisperson.
Overdispersion is a phenomenon that implies there is more variability around
the model's fitted values than is consistent with a Poisson formulation.
Suppose a Poisson model is used for modeling accidents and if the variance
(or dispersion) of the data exceeds the estimated mean of the accident data
distribution, then the data are said to be overdispersed, and the underlying
assumption of the variance being equal to the mean for the Poisson
distribution is violated (Bauer, M. and Harwood, W. (1998)). The negative
binomial is proposed as a means to correct for this problem (Osgood, 2000).
Negative binomial distribution has two parameters, unlike the Poisson
distribution.
27


For negative binomial distribution the relationship between the expected
number of accidents occurring at the ith element and the q parameters is
same as Equation (2.1). The probability that a ramp Y; = y} accidents can be
expressed as:
PM = yd =
r(y+E> ftuF
rry + ijr(i)n + iqly*i
Vi = 0,1,2,...
(2.7)
The mean and variance of the negative binomial distribution of accident
frequency are given by:
Mean = E(Y) = (2.8)
Variance = Var(Y) = \i[ + k[i2 (2.9)
where:
k = the negative binomial dispersion parameter
The model regression coefficients (B0 are estimated by the method
of maximum likelihood minimizing the negative of the log likelihood.
Bauer, M. and Harwood, W (1998) developed the log likelihood for the
negative binomial distribution and is given by:
28


n
log(L)
^ Yi log
i=l
a
(a + 1).
nk log(a + a) + function of yj, k
(2.10)
Modifying Equation (2.10):
log(L) = ^ yi log
i=l
gk
.(gk + 1)J k
n
log(a + gk) + function of yj, k
(2.11)
Which is equivalent to:
log
- Po + PlXil + P2Xi2 1------h PqXiq
(2.12)
2.6.3 Goodness of Fit Measure
The scaled deviance and Pearsons chi-square statistic are helpful in
assessing the goodness of fit of a given generalized linear model. According
to Nelder and Wedderburn (1972), for a fixed value of the dispersion
parameter , scaled deviance is defined to be twice the difference between
the maximum achievable log likelihood and the log likelihood at the maximum
likelihood estimates of the regression parameters.
29


Let Z(x,ju) the log-likelihood function,
where:
/1 = the mean of the distribution
y = the response values,
Then the scaled deviance is defined by:
SD(y,ii) = 2l{l(y,y) - (2.13)
SD(y,ii) =
D(y. M)
0
where:
(2.14)
D(y,n) =model (residual) deviance
0= the generalized linear model dispersion parameter
Accordingly, scaled deviance for the Poisson distribution is given by Equation
(2.15):
D = 2
Zy'ln-Z(y'-,)
i=l i=l
(2.15)
30


Also, scaled deviance for the negative binomial distribution is given as:
D = 2I
i i
where:
ylog(f)-(y+v)hB
k = the negative binomial dispersion parameter
(2.16)
Since

distributions the scaled deviance is equal to the deviance.
Pearsons chi-square statistics is given by:
2 v 2
*
i=i
and the scaled Pearsons chi-square is defined as:
where:
fi = the mean of y
V(yd = the variance function of the distribution.
(2.17)
(2.18)
31


In both Equation (2.16) and (2.17) the weight to each interchange is to be
one.
Since 0 is one for both Poisson distribution and negative binomial
distributions the scaled Pearsons chi-square is equal to the Pearsons chi-
square statistics.
2.7 Traffic Crash Prediction Model
Bauer, M. and Harwood, W. (1998) developed models to predict accidents.
Initially accidents on ramps and speed-change lanes were modeled
separately with the thought that accident predictions from the separate
models could be added together to determine the combined safety
performance of a ramp and its adjacent speed-change lane. The study proved
that the separate models, by themselves, didn't provide an adequate fit to the
data. Therefore, they developed the best models of the safety performance of
interchange ramps and speed-change lanes by combining the accident
experience of ramps and their adjacent speed-change lanes into a single
model.
Independent variables included in the model include:
Ramp AADT
Mainline freeway AADT
Area type (rural/urban)
Ramp type (off/on)
Ramp configuration
Length of speed-change lane
Ramp length
32


It was determined that these seven independent variables are all statistically
significant at the 20-percent significance level in the model for total accidents,
and all of these independent variables, except ramp configuration, are
significant in the model for fatal and injury accidents.
To predict the average accident frequency for a ramp including the adjacent
speed-change lane, the regression coefficients .Pq are replaced in
equation (2.19) by the estimated values of the coefficients and the variables
X1X2, ->Xq are replaced by their appropriate values or levels.
function(p*) = exp{fQ){AADTramp)Plexp{f2Xi2) ... exp(fqXiq) (2.19)
For example, the expected 3-year total accident frequency can be estimated
as
Y = e-7-27(X1)-78(X2)013exp(0.45X3)exp(0.78X4)exp(-0.02X5)
exp(0.69X6)exp(-0.37X7)exp(0.37X8)
exp(-2.59X9)exp(1.62X10) (2.20)
where:
Y = expected number of total accidents in a 3-year period on entire ramp
plus adjacent speed-change lane
X = ramp AADT (veh/day)
X2 =mainline freeway AADT for the direction of travel in which the ramp is
located (veh/day)
33


X3 = 1 if the ramp is a diamond ramp; 0 otherwise
X4 = 1 if the ramp is a parclo loop ramp; 0 otherwise
X5 = 1 if the ramp is a tree-flow loop ramp; 0 otherwise
X6 = 1 if the ramp is an outer connection ramp; 0 otherwise
X7 = 1 if the area type is rural; 0 otherwise
X8 = 1 if the ramp is an off-ramp; 0 otherwise
*9 = speed-change lane length (mi)
X10 = ramp length (mi)
The estimate of dispersion after fitting, as measured by the deviance and
Pearsons chi-square divided by the degrees of freedom, are 1.0 and 0.95
respectively for total accidents which are well within the acceptable range.
Also the Ft2 goodness of fit is approximately 0.38.
Among these variables, it was found that mainline and ramp AADT to be the
most important predictor of accident on ramps. The increase in crash
frequency was associated with higher accidents. The effect of acceleration
length was more complicated since on one hand longer lengths provide safer
traffic maneuvering and on the other hand corresponding increased length
may provide longer exposure and more accidents.
It was noted that the 3-year accident prediction models were divided by 3 to
obtain the corresponding accidents per ramp per year.
34


2.8 Empirical Bayes Technique
Bayes theorem is a probability theory that shows how new information can be
used to update or revise an existing set of probability. The prior probability
may have to be estimated based on very little information. These probabilities
would be improved as more information becomes available.
When it comes to traffic safety, Empirical Bayes (EB) is a theoretical
framework that combines the actual crash count of a location with an
estimated crash frequency determined from the safety performance function
of the same location. Sayed (1995) stated that the expected number of
accidents at a location is a random variable that fluctuates around some
unknown mean. This randomness is reason that historical accident data at a
location does not always accurately reflect its long-term accident
characteristics. Higle and Witkowski, (1988) stated that a location that has
low accident frequency during long periods of time may have had high
accident rates during portions of this period and vice versa and confirmed that
Empirical Bayes approach accounts for random variations of accident rate.
The technique is based on Bayes theorem which can be mathematically
described as:
P(0|x) =
P(x|0) X P(0)
ZP(x|0)xP(0)
(2.21)
35


Where:
0 = a parameter such as the number of accidents at a location,
P(0) = the prior distribution of 0,
P(x|0) = the probability of making x observation for a specific value of
0 (observation distribution), and
P(0|x) = the posterior distribution of 0 which represents the resolution of the
prior distribution given the observations.
According to Calvin (1990) the distribution of the observation is assumed to
be a Poisson or Binomial distribution and the prior distribution will be a
gamma or beta distribution. In the Empirical Bayes approach, the parameters
are estimated using a sample of observations from population of similar
locations.
2.8.1 Empirical Bayes Safety Estimate
Hauer (1992), Brude and Larsson (1988) stated that the safety of a location
provide clues about its traffic and road characteristics, and its historical
accident data. The Empirical Bayes (EB) approach makes use of both kinds
of clues and is used to refine the estimate of the expected number of
accidents at a location by combining the observed number of accidents at the
location with the predicted number of accidents obtained from the GLM model
to yield a more accurate, location-specific safety estimated. Equation 2.31
confirmed this. Brude and Larsson, (1988) showed that EB method
significantly reduces the regression to the mean effects that are inherent in
observed accidents count. Regression to the mean is a statistical
phenomenon that refers to the tendency of extreme events (high number of
36


accidents) to be followed by less extreme values (a lower number of
accidents) even if no treatment is applied to the field configuration that
generates the accidents at that location.
According to Hauer (1992), the EB safety estimate can be calculates by.
EBSafety estimate OC.pved + (1 Cc). COUTlt, (2.22)
where,
1
11 ~ i+var(pd>
pred
where,
count = observed number of accidents at the location
pred= predicted number of accidents as estimated from the GLIM model
var(pred) = the variance of the GLM estimates
According to (Sayed et.al., 1998),
var(pred) =
(pred)2
K
37


Where:
k = shape parameter
Then, Equation (2.22) can be rearranged as:
EBsafety estimate
k \ / pred
count
(2.23)
In addition, the variance of the EB estimate can be calculated using (Kulmala,
1995):
2.8.2 Bayesian Identification of Accident-Prone Locations
Higle and Witkowski (1988) described that Empirical Bayes procedure is a
useful technique to identify accident-prone locations.
The following notations are used:
1; = accident rate at a location i (treated as a random variable),
Nt = number of accidents at location i during the period of time in question,
Vt = number of vehicles passing through location i during the period of time in
question,
f t (A|Ni.Vj) = probability density function associated with the accident rate at
(2.24)
location i given the observations Nt and Vt (Posterior
distribution) ,
38


fR (A) = probability density function associated with the accident rate across
the region (Posterior distribution) ,and
a,p = parameters of gamma distribution.
Two assumptions were made:
1. At any given location, the actual number of accidents (JV£) follows a
Poisson distribution such that at any given location, where the accident rate
is known (A£ = A) and the expected value is given by XVt then the
observation probability distribution is given by the following:
The parameters a and p of the prior distribution are estimated using the
method of moments estimates (MME), and are chosen so that the mean and
variance associated with the gamma distribution are equal to the mean (x)
and variance (s2) of the sample.
(2.25)
2. The probability distribution of the regional accident rate (the prior
distribution), fR(X), follows a gamma distribution and is given by:
(2.26)
39


If (x) and (s2) are the sample mean and variance of the observed accident
rates respectively, and m is the number of locations under observations, then:
m
X m 2., Vt
i=1
T(N ^
m
i=1

(2.27)
(2.28)
Morris (1988) and Berger (1985) modified the parameters of gamma
distribution to avoid the bias and improve the estimation. Accordingly, the
probability density function associated with the accident rate at a location i is
given by:
fMNuVd
J-l__wn-lp-PiX
r fo)
(2.29)
And thus, location i will be identified as accident-prone if there is a significant
probability that the locations accident rate,If, exceeds the observed regional
accident rate, XR.
40


Thus, location i is identified as accident prone if:
P{Ai>XR\Ni,Vi}>8
(2.30)
or equivalently if:
at
1- I ^-Aa_1e_^dA
-I
o
rfe)
> s
(2.31)
where:
S = the confidence level desired, such as 0.95 or 0.99, and
The value of the regional accident rate, XR, is calculated using:
JT-iNt
(2.32)
Sayed (1995) modified the Higle and Witkowskis method by substituting
a + Ac.Vi for at in Equation (2.31) and is given by:
xR
-I
p{a+)ic.Vj)
r {a + Ac.Vi)
1(+Ac.Vi-l)e(.-ptX)
dA
(2.33)
41


2.9 Simulation Modeling
2.9.1 Choice of a Simulation Tool
A need for simulation modeling has been increasing to understand the
behavior of traffic as transportation systems have become more complex.
Many researches in the past have studied different similarities and differences
of simulation tools and their capacity to model different combinations of traffic,
highway type, geometric configuration, etc. of the system.
Gettman D., and Head L., (2003) conducted a study to evaluate the various
simulation models capabilities for producing measures of intersection safety
and specify algorithms for calculating the measures. The nine multipurpose
micro simulation software includes CORSIM, SIMTRAFFIC, VISSIM,
HUTSIM, PARAMICS, TEXAS, AIMSUN, WATSIM, and INTEGRATION. It
was noted that VISSIM, AIMSUN, PARAMICS, and TEXAS outputs the
proper TRJ files required by SSAM. It was concluded that there are a number
of advantages that VISSIM possesses over other micro-simulation tools.
VISSIM appears to be a full-featured microscopic simulation model with the
ability to obtain detailed state variable information on each vehicle on time
scales with better than second-by-second accuracy. Moreover it has been
interfaced to other external codes such as hardware signal controllers, thus
the developers have experience in development collaboration. The priority
rules feature of VISSIM appears to allow complex modeling of junction
behavior, including friendly merging (situations where following vehicles will
slow for merging vehicles to create a gap), as it occurs in the real world. It is
not apparent that other simulation models are able to represent such
behavior. VISSIM allows multilane merging behavior; it is typical for vehicles
entering the mainline flow to cross the path of an oncoming vehicle traveling
42


in the same direction as the intended direction of travel of the entering vehicle
and start accelerating in the adjacent lane. In this way, the oncoming vehicle
can continue at its current speed without having to break for the turning
vehicle. VISSIM appears to support most of the modeling features required
for obtaining surrogate measures at a reasonable level of fidelity.
Because of its competitive advantages, combined with its capability to
analyze networks of all sizes, VISSIM was chosen to perform the simulation
modeling of traffic.
2.9.2 VISSIM Micro-Simulation Tool
According to VISSM User Manual (2009), VISSIM is a microscopic, time step
and behavior-based simulation model developed to model urban traffic and
public transport operations. The program can analyze private and public
transport operations under constraints such as lane configuration, traffic
composition, traffic signals, etc., thus making it a useful tool for the evaluation
of various alternatives based on transportation engineering and planning
measures of effectiveness.
The model was developed at the University of Karlsruhe, Germany during the
early 1979. PTV America, Inc., a subsidiary of PTV AG located in Karlsruhe,
Germany, is the North American distributor and developer of PTV Vision
software products. VISSIM version 5.20 was used in this study.
Unlike many other simulation tools, it is not strictly node-link based, it is based
on link-connector geometry. In VISSIM the user builds a network based on
43


aerial photographs at the exact point where they are needed to have the
desired effect on road-users.
Traffic flow in VISSIM is simulated by moving driver-vehicle-units through a
network. Every driver is stochastically assigned to a specific vehicle.
Consequently, the drivers behavior will be adapted to the vehicle
performance characteristics. Attributes characterizing each driver-vehicle-unit
can be classified in to three components: (1) technical specification of the
vehicle (e.g. length, engine power, weight, maximum speed, potential
acceleration actual position in the network, and actual speed and
acceleration), (2) behavior of driver-vehide units (e.g acceleration based on
current speed and drivers desired speed, (3) interdependence of driver-
vehicle units (e.g. reference to leading and following vehicles on own and
adjacent travel lanes)
2.10 SSAM Conflict Analysis Tool
The Surrogate Safety Assessment Model (SSAM) is owned by FHWA.
According to Lili Pu and Rahul Joshi (2008), SSAM operates by processing
data describing the trajectories of vehicles driving through a traffic facility and
identifying conflicts. The vehicle trajectory input data for SSAM are generated
by traffic simulation software in a trajectory file format (where files are labeled
with a .trj file extension), specially designed for SSAM. SSAM calculates
surrogate measures of safety corresponding to each vehicle-to-vehicle
interaction and determines whether or not each interaction satisfies the
criteria to be deemed an official conflict. A table of all identified conflicts and
their corresponding surrogate safety measures is then presented to the user.
44


Figure 2.3 illustrates the workflow for using SSAM. It was reported that the
users of the SSAM software would include traffic safety engineers, traffic
simulation analysts, and traffic researchers.
FIGURE 2.3 EVENT-FILE BASED INFORMATION FLOW DIAGRAM
(Gettman D., and Head L., (2003)).
Surrogate Safety Assessment (SSAM) software uses two threshold values for
surrogate measures of safety to delineate which vehicle-to-vehicle
interactions are identified as conflicts. These two thresholds are Time-to-
Collision (TTC) and Post-Encroachment Time (PET). SSAM utilizes a default
maximum TTC value of 1.5 seconds as suggested by (Sayed, Brown, Navis,
1994) and (Hayward, 1972) and the maximum PET value of 5 seconds as
suggested by Hyden (1987).SSAM filters out conflicts that do not fall in the
specified ranges. These default values were used in this study.
45


Gettman, Pu, Sayed, and Shelby (2008) developed conflict identification
algorithm of SSAM and described as follows:
First, SSAM determines the dimensions of the analysis area based on the
TRJ file and constructs a zone grid to cover the entire rectangular analysis
area. By dividing the region into zones, the number of vehicle-to-vehicle
comparisons necessary to identify potential conflicts is reduced considerably.
Second, SSAM analyzes a single time step of a trajectory file. For each
vehicle in the analysis region, it projects that vehicles expected location as a
function of its current speed, if it were to continue traveling along its (future)
path for up to the duration of the configured time-to-collision (TTC) value.
Third, for each vehicle, SSAM calculates the rectangular perimeter
delineating the location and orientation of that vehicle at its projected future
position. Any time a vehicle is added into a zone that currently contains one
or more other vehicles, SSAM checks for overlap of the new vehicle
(rectangle) with each of the other vehicles (rectangles) in that zone. It is
possible that two vehicles may partially occupy the same zone without
overlapping. However, two overlapping rectangles indicate that a future
collision is projected for this pair of vehicles, and therefore, a potential conflict
has been identified. SSAM maintains a list of all conflicting vehicle pairs (all
conflict events) for the current time-step. Each time-step, the list is
prepopulated with all conflicting vehicle-pairs from the prior time-step. If the
current vehicle being added to the zone grid overlaps with any other vehicle,
that vehicle-pair is added to the conflict list for the current time-step (if not
already in the list).
46


Four, SSAM continues to perform a more detailed processing of each
conflicting vehicle-pair in the list for the current time-step as follows:
(a) The TTC of the vehicle-pair is updated by iteratively shortening the future
projection timeline by a tenth of second and re-projecting both vehicles as
before over successively short distances until the pair of vehicles no longer
overlaps in their projected locations. In this way, a more accurate TTC value
is established for this time-step. Instead of the large overlap, the vehicles can
have just barely come into contact. Note that if the projection timeline reduces
to zero seconds and the vehicles still overlap, then this is a crash.
(b) The minimum TTC (taking the minimum of the current TTC value and that
of the prior time-step, if applicable) are calculated and updated. Also, the
current (actual) positions of both vehicles are recorded for post-encroachment
analysis.
(c) If it was found that the vehicle-pair does not overlap over any projection
time between zero and maximum TTC, then this vehicle-pair has made its
way into the conflict event list by virtue of being in the list during the prior
time-step. In this case, the event remains in the list, watching for the one
vehicle (the trailing vehicle) to eventually occupy (or encroach on) on a
position formerly held by the other vehicle (the leading vehicle). The time
differential between when the leading vehicle occupied this location and the
trailing vehicle arrived is the post-encroachment time (PET). If a post-
encroachment was observed, then the minimum PET is updated, and this
conflict event remains in the list, as the post-encroachment could potentially
reduce as the vehicle trajectories progress over time.
47


(d) If a vehicle-pair in the conflict event list is no longer on an imminent
collision course, and it is clear that PET to any prior positions could not further
reduce the minimum PET, or the maximum PET has elapsed, then this
vehicle-pair is identified for removal from the conflict event list. Prior to
removal, all final surrogate measures are computed, including conflict starting
and end points, and conflict angles.
After SSAM identifies the conflicts, it classifies them in to four types: rear end,
lane change, crossing and unclassified conflicts. The link and lane information
is utilized for classification in the case that the vehicles both occupy the same
link at either the start or end of the conflict event. The simulation model
(VISSIM) that produced the vehicle trajectory data can generally provide link
and lane information for both vehicles. If both vehicles occupy the same lane
at the start and end of the event, then it is classified as a rear-end event. If
either vehicle ends the conflict event in a different lane than it started (while
having not changed links), then the event is classified as a lane change. If
either of the vehicles changes links over the course of the event, then the
conflict angle determines the classification using the threshold values
mentioned next.
The event type is classified based on the absolute value of the Conflict Angle.
A conflict is classified as unclassified if the conflict angle is unknown, rear-end
if ||Conflict Angle|| < 30, crossing if ||Conflict Angle|| > 85, lane change if
30 < || conflict angle || < 85. Figure 2.4 illustrates the conflict angle. The
threshold angles (30 and 85) were determined by experimentation.
48


180

o
9 = conflict angle
9, = rear end threshold angle
9j = crossing threshold angle
Refer to User Manual for
more detai
FIGURE 2.4 : ILLUSTRATION OF CONFLICT ANGLE DIAGRAM (SSAM
SOFTWARE)
2.11 SAS/STAT Software and the GENMOD Procedure
According to the SAS/STAT software user manual (2011), the software
provides comprehensive statistical tools for a wide range of statistical
analyses, including analysis of variance, categorical data analysis, cluster
analysis, multiple imputation, multivariate analysis, nonparametric analysis,
power and sample size computations, psychometric analysis, regression,
49


survey data analysis, and survival analysis. A few examples include nonlinear
mixed models, generalized linear models, correspondence analysis, and
robust regression. The current release of SAS software, SAS 9.3 version, was
used for this study.
A component of SAS software called GENMOD procedure was used. The
GENMOD procedure fits generalized linear models, as defined by Nelder and
Wedderburn (1972). The class of generalized linear models is an extension of
traditional linear models that allows the mean of a population to depend on a
linear predictor through a nonlinear link function and allows the response
probability distribution to be any member of an exponential family of
distributions. Many widely used statistical models are generalized linear
models. These include classical linear models with normal errors, logistic and
probit models for binary data, and log-linear models for multinomial data.
Many other useful statistical models can be formulated as generalized linear
models by the selection of an appropriate link function and response
probability distribution.
The GENMOD procedure fits a generalized linear model to the data by
maximum likelihood estimation of the parameter vector. There is, in general,
no closed form solution for the maximum likelihood estimates of the
parameters. The GENMOD procedure estimates the parameters of the model
numerically through an iterative fitting process. The dispersion parameter is
also estimated by maximum likelihood or, optionally, by the residual deviance
or by Pearsons chi-square divided by the degrees of freedom. Covariances,
standard errors, and p-values are computed for the estimated parameters
based on the maximum likelihood estimators.
50


3. Research Approach and Methodology
3.1 Field Validation
The field validation effort for the study sites were assessed with SSAM, and
the results were compared to actual crash histories. Also, surrogate safety
estimates were compared with traditional volume-based models for crash
prediction. The field validation testing was based solely on modeling with the
VISSIM simulation.
3.2 Purpose
The main purpose of the field validation effort was to compare the predictive
safety performance capabilities of the SSAM approach with actual crash
experience of merging and diverging freeway sections. This effort consisted
of a series of statistical tests to assess the correlation between actual crash
frequencies at a series of merge/diverge sections and the corresponding
frequency of conflicts observed in simulation models of these sections.
Traditional volume-based crash prediction models were used as a basis for
comparison.
3.3 Data Assembly and Geometric Definition
Merging and diverging influence areas along freeways at interchanges were
selected based on the following guidelines: (1) a sufficient number of
51


interchanges were needed to obtain the range of parameters necessary to
test the methodology, and to establish statistical measure of significance. (2)
interchanges were not selected based on their crash performance (e.g.,
most dangerous interchanges in an area) as this will lead to the
regression-to-the-mean bias (Hauer, E. 1980). The regression-to-the-mean
bias is the tendency of a randomly high accident frequency occurring at a
location during a specific time period to be followed by a smaller accident
frequency during a consecutive period of equal duration, even if no treatment
is applied to the field configuration that generates the accidents at that
location.
Appendix A provided a summary of the twenty-one interchanges used in this
study. It shows, for each interchange: the highway names, cross street
names, truck percentage, average daily traffic (ADT) and peak hour volume
(PHV) for the mainline, on-ramp and off-ramp movements.
Prior to fitting the model all merge/diverge related accidents were isolated
from the CDOT accident database. This eliminated the influence of non-
merge/diverge accidents from the analysis. Appendix C presents the crash
data in terms of average yearly crash counts for each merge/diverge
locations. The crash counts were derived by filtering through all interchange
crash records to include only two (or more) vehicle crashes. Thus, single-
vehicle crashes, such as run-off-road crashes, fixed-object crashes, and
animal, pedestrian, or bicycle related crashes, were excluded.
In addition to the crash count, properly defining the interchange geometry was
an important part of the modeling process. Traffic exiting a highway is
required to reduce the speed. The speed-change lane that enables the driver
52


to reduce the speed between the highway and the turning roadway in a safe
and comfortable manner is called deceleration lane. On the other hand traffic
entering a highway is required to increase the speed. The speed-change lane
that enables the driver to increase the speed between the highway and the
turning roadway in a safe and comfortable manner is called acceleration lane.
Roess and Ulerio ( 1993) studied the on-ramp and off-ramp junctions and
showed that lane changing and turbulence extend to 1500 feet downstream
from the tip of the merge gore and 1500 feet upstream from the tip of the
diverge gore. Moreover, the mainline traffic starts to change lane upstream of
the gore at merge. Similarly at diverging location it is common to see some
traffic to change lanes downstream of the gore for adjustment. As a result of
these, at both merge/diverge locations crashes at a distance of 1500ft on
either sides of the gore (total 3000ft) were compared with the corresponding
conflicts recorded for the same distance (3000ft). Figure 3.1 and Figure 3.2
illustrates the model geometric configuration of merge and diverge influence
areas respectively. Collection of traffic counts from CDOT crash database
and conflict analysis by SSAM were based on the definition of these Figures.
The model considered only right-hand traffic flow conditions for both
merge/diverge influence areas.
53


traffic movement
gore
=> number of lanes vary

FIGURE 3.1 ILLUSTRATION OF MERGE INFLUENCE AREA
FIGURE 3.2 ILLUSTRATION OF DIVERGE INFLUENCE AREA
54


3.4 Simulation Modeling of Interchanges and Modeling Assumptions
Forty two merging and diverging locations were coded in VISSIM simulation
modeling tool. The process of modeling of interchanges in VISSIM was
started by tracing an aerial photo of the proper geometric configuration by
defining the number, width, and length of lanes. After the geometric
configuration of the interchange was defined, other modeling parameters
such as traffic flow (peak-hour volume), traffic composition, percent truck,
routing decision, etc., were allocated.
Several modeling parameters were defined during the process including:
speed profiles, vehicle type, traffic composition, routing decisions and other
aspects. The traffic composition was assumed to be composed of passenger
cars and trucks. The percentage of trucks was coded at each interchange.
The desired speed profiles were assumed to range from 50mph to 75mph for
mainline traffic and 30mph to 45mph for on-ramp and off-ramp traffic. Car-
following and lane-changing behavior were set to model freeway (free lane
selection) traffic flow using the Wiedemann 99 model with all default
parameters used (VISSIM 5.2 User Manual, 2009). A simulation resolution of
5 time steps per simulation second was used. Among the different types of
routing decisions a static routing option was used which defines the vehicles
routes from a start point to any of the defined destinations using a static
percentage for each destination.
The lane-changing algorithm in VISSIM allows for two types of lane changes:
necessary lane changing and free lane changing. A free lane changing option
was chosen. In a free lane change, VISSIM checks for the desired speed
55


safety distance of the trailing vehicle on the new lane. This safety distance
depends on the vehicle speeds.
Following nominal verification, each interchange was simulated five times for
a period of 3600 simulation seconds (1 hour). Instead of running multiple
simulation runs with different random seeds, a multirun option was used. This
option lets VISSIM automatically run the requested five successive runs at
each location. When run this way, VISSIM automatically uses a new random
seed number at the starts of each run.
The modeled interchanges were tested for realistic and reasonable vehicle
behaviors. The TRJ output files from VISSIM run model were imported into
the SSAM application to identify traffic conflicts and calculate the
corresponding surrogate safety measures. The output from SSAM was
formatted and analyzed by SAS GENMOD statistical software to produce
safety performance functions.
The traffic flow inputs to VISSIM were based on peak hour volumes, as
recorded by the CDOT traffic database. The volumes used for simulation are
included in Appendix A.1. Figure 3.3 and Figure 3.4 are screen captures of
geometric layout and flow of traffic modeled in VISSIM respectively.
56


S?


FIGURE 3.4 SCREEN CAPTURE. VISSIM TRAFFIC MODEL OF 1-25 &
E HAMPDEN AVE
58


3.5 Identification and Removal of Outliers
Before any statistical inferences were applied, the outliers were identified and
removed from the dataset. Outliers are observations or subsets of
observations which appear to be inconsistent with the remainder of the data
(Barnett & Lewis, 1978). There are three ways for outliers to arise in a sample
(Barnett and Lewis (1994, pp. 33-34)): (1) data due to a reading error, a
recording error, or a calculation error in the data; (2) a data value that
incorrectly included in the data set; (3) a correctly recorded data value that
belongs in the data set. Since these causes of outliers are possible to occur in
the data, due attention were given to identify and remove them.
The Standardized Residual Criterion (z-score method) was used to identify
outliers. The standardized residual, the STUDENT statistic, is the residual
expressed in units of standard deviation from the mean value. Thus, a
standardized residual with an absolute value of 3 is 3 standard deviations
above the mean value. Observations with standardized residual with an
absolute value of 3 or greater are potential outliers and were removed from
the analysis.
X; X
z = --------
er
(3.1)
where: Z= It denotes the number of standard deviations a data value x£ is
from the mean, x.
59


3.6 Methodology
The methodology used for the field validation task was based on relating
surrogate safety measures that SSAM derived from simulation models with
actual crash occurrence observed on the field at merging and diverging
locations. The field validation effort involve the analysis of 21 interchanges
(42 merging and 42 diverging) locations, modeled with the VISSIM simulation.
Gettman, Pu, Sayed, and Shelby (2008) provided five statistical field
validation tests for the analysis of intersections. These tests were modified
and used in this study for merging and diverging locations:
1. Field Validation Test 1: Merging and Diverging Section Ranking by Total
Incidents
2. Field Validation Test 2: Merging and Diverging Section Ranking by Specific
Incident Types
3. Field Validation Test 3: Crash and Conflict Prediction Regression Model
Paired Comparison
4. Field Validation Test 4: Crash and Conflict Prediction Regression Model
Comparative Analysis of Total Incidents
5. Field Validation Test 5: Crash and Conflict Prediction Regression Model
Comparative Analysis of Specific Incident Types
3.6.1 Field Validation Test 1: Merging and Diverging Location Ranking
by Total Incidents
In this field validation test the ranking of merging and diverging locations from
SSAM according to predicted total conflicts was compared to the same
60


locations using actual crash frequency. Three steps were performed in this
test:
3.6.1.1 Conflict Ranking
In this step, SSAM was used to predict the expected total number of conflicts
at each merging and diverging locations. Each location was simulated for five
replications, each lasting 3600 simulation seconds (1 hour). Thus, the sum of
all conflicts recorded over all five replications was divided by 5 to determine
the average hourly conflict frequency in terms of conflicts per hour. The
merging and diverging sections were then ranked based on their average
hourly conflict frequency in descending order.
3.6.1.2 Crash Ranking
Crash ranking in this step was performed in such a way that the average
yearly crash frequency for each merging and diverging locations was first
determined by dividing the total number of crashes over the observation
period by the number of years in the observation period. The 3 years
observation period were used for all study locations. Merging and diverging
locations were then ranked based on their average yearly crash frequency in
descending order.
61


3.6.1.3 Ranking Comparison
The ranking derived from SSAMs conflict prediction (section 3.6.1.1) was
compared to the rankings based on average yearly crash frequency (section
3.6.1.2). The Spearman rank correlation coefficient was used to determine
the level of agreement between the two rankings. The Spearman rank
correlation coefficient is often used as a nonparametric alternative to a
traditional coefficient of correlation and can be applied under general
conditions. The Spearman rank correlation coefficient (ps) is calculated as
shown in Equation 3.2. A score of 1.0 represents perfect correlation and a
score of 0 indicates no correlation. An advantage of using (ps) is that when
testing for correlation between two sets of data, it is not necessary to make
assumptions about the nature of the populations sampled.
n(n2 1)
(3.2)
Where:
di =the difference between two rankings for pair i.
n = the number of paired ranked.
For hypothesis test, the Spearman rank correlation coefficient,ps, was
compared with the critical value ,ps (critical)-
For n < 30, ps (critical) can be read from the critical values table.
62


For n > 30, use:
Z
Ps (critical) ~ i /,
VoT^T)
(3-3)
Where:
Z corresponds to the level of significance.
For example, if a = 0.05, then Z = 1.96.If the absolute value of the
test statistics ps exceeds the positive critical value, then reject H0: ps = 0 and
conclude that there is a correlation.
3.6.2 Field Validation Test 2: Merging and Diverging Location Ranking
by Specific Incident Types
Comparative ranking procedures for Rear-end, crossing and lane-changing
crash/conflict types were repeated in Field Validation Test 2 the same as
Field Validation Test 1. These three conflict types were classified and counted
by SSAM for each type. The average hourly conflict frequency for each
conflict type was computed as in Field Validation Test 1, and these results
were used to rank the merging and diverging sections for each conflict type.
To provide type-specific crash frequencies, all crash reports from CDOT
crash database were reviewed to determine whether it was a rear-end
incident, crossing incident, or lane-changing incident. If the classification
based on the crash type was not obvious, engineering judgment was applied
to determine the most representative type. Average yearly crash frequencies
was tabulated for each incident type, rank ordered, and compared using the
63


Spearmans rank correlation test, as in Field Validation Test 1. The results of
this step demonstrated the capability of SSAM to accurately identify and rank
merging and diverging sections .
3.6.3 Field Validation Test 3: Crash and Conflict Prediction
Regression Model Paired Comparison.
In this field validation test, the correlation between conflicts obtained from
SSAM and crashes were developed. Regression equations were established
to estimate the average yearly crash frequencies at merging and diverging
locations as a function of the average hourly conflict frequencies. The R2 ,
Pearson chi-square and scaled deviance goodness-of-fit testing were used to
determine the strength of the relationship between conflicts and crashes.
Regression equations to predict crashes based on volume were developed
and compared with the conflict-based crash prediction model to study the
relative capabilities of surrogate safety assessment. Generalized linear
modeling GLM technique (Hauer, 1997; Sayed and Rodriguez, 1999) were
used to calculate the expected crash frequency at each merging and
diverging location, using the GENMOD procedure in the SAS 9.3 statistical
software package.
64


3.6.4 Field Validation Test 4: Crash and Conflict Prediction Regression
Model Comparative Analysis of Total Incidents
This Field Validation Test consists of four steps and the details of the test are
as follows:
Step A: Conflict prediction models for merging and diverging locations were
developed using standard GLM procedures relating conflicts calculated by
SSAM as a function of hourly traffic volume.
Step B: Crash prediction models for merging and diverging locations were
developed using standard GLM procedures relating actual crash data
obtained from CDOT accident database as a function of daily traffic volume.
Step C: The two prediction models were compared to determine whether or
not the conflict prediction models can predict risk in a manner similar to the
crash prediction model for merging and diverging locations with the same
characteristics. The comparison consists of identification and ranking of
crash/conflict prone locations. A crash/conflict prone location is defined as
any location that exhibits a significantly higher number of crashes/conflicts as
compared to a specific normal value. The normal value was provided by the
volume-based crash/conflict prediction models. As stated previously, the
Empirical Bayes (EB) technique improves the location-specific prediction and
thus was used to identify hazardous locations. The EB refinement method
identifies problem sites according to the following four-step process:
1. Estimate the predicted number of crashes/conflicts and its variance
using the crash/conflict prediction model. This prediction can be
65


assumed to follow a gamma distribution (the prior distribution) with
parameters a and /?, where:
^ Var{ A) E{ A)
E{ A) k
and a = (3.E(A) = zc
(3-4)
Where:
Kar(A) = the variance of the predicted crashes/conflicts.
£(A) = predicted accident frequency
zc = shape parameter
2. Determine the appropriate point of comparison based on the mean and
variance values obtained in step 1. Usually the 50th percentile (P50) or the
mean is used as a point of comparison P50 is calculated such that Equation
3.5 is satisfied.
(3.5)
0
3. Calculate the EB safety estimate and the variance by rearranging
Equations 2.23 and 2.24 respectively as shown in Equation 3.6
and Equation 3.7.
66


EBsafety estimate ( K+E(A)) E (A) + (K+E(A)) (.COUTlt)
(3-6)
where:
count = observed accident frequency.
V or (EBsafe fy
estimate
g(A) \2 ( E{A)
k + E(A)/ W U + E(A)
(count)
(3.7)
This is also a gamma distribution (the posterior distribution) with parameters
ax and & defined as follows:
EB
/?i = -------
Var(EB)
K
~e{A)
+ 1 and at = ^.EB = /c + count
(3.8)
Then, the probability density function of the posterior distribution is given by
the following:
/eb W
G(A)
(K+count)
^jK+count-lg
E(k + count)
K
.£( A)
+11
(3.9)
4. Identify the location as crash/conflict-prone if there is significant
probability that the locations safety estimate exceeds the ^50 value (or
the mean).
67


The location is prone if:
p (w^+1)(K+C0Unt)^K+count~le (£(a)+1)a
rPsoUALJ------------------------------ dX
J0 r(jc+count)
> S
(3.10)
Where:
S is the desired confidence level (usually selected at 0.95).
The merging and diverging locations were then ranked in terms of priority for
treatment once crash/conflict-prone sites were identified. Ranking problem
sites enables the road authority to establish an effective road safety program,
ensuring the efficient use of the limited funding available for road safety. Two
techniques were suggested (Sayed and Rodriguez, 1999) that reflect different
priority objectives for a road authority. The first ranking criterion is to calculate
the ratio between the EB estimate and the predicted frequency as obtained
from the GLM model (a risk-minimization objective). The ratio represents the
level of deviation that the merge/diverge is away from a normal safety
performance value, with the higher ratio representing a more hazardous
location. The second criterion, the cost effectiveness objective, is a potential
for improvement (PFI) criterion, which is calculated as the difference
between the observed crash/conflict frequency and the volume-based
estimate of normal crash/conflict frequency.
Step D: Two comparisons were undertaken. The first compared the locations
identified as crash-prone to merging and diverging locations identified as
conflict-prone. The second compared both the risk ratio and PFI merging
68


and diverging location rankings obtained using crash data to the
corresponding rankings obtained using conflict data.
3.6.5 Field Validation Test 5: Crash and Conflict Prediction
Regression Model Comparative Analysis of Specific Type
Incidents
Field Validation Test 5 repeated the same comparative analysis as Field
Validation Test 4 for the rear end, crossing and lane change crash/conflict
types.
3.7 Goodness of Fit Measure
Three statistical measures were used to assess the goodness-of-fit of the
regression models to the actual crash data to the merging and diverging
locations.
The first measure was the Pearson chi-squared and computed by the
following model:
Where:
E(Ai) = the predicted crash frequency of merging or diverging location i
(3.11)
69


yt = the actual crash frequency of merging or diverging location i
n = the number of merging or diverging location
The second statistical measure used was the scaled deviance. This is the
likelihood ratio test statistic measuring twice the difference between the log-
likelihood of the data under the developed model and its log-likelihood under
the full or saturated model. The scaled deviance is computed by the following
model:
There are several approaches to estimate the shape parameter k of the
negative binomial distribution with the method of maximum likelihood being
the most widely used. The method of maximum likelihood will be used in this
analysis.
The third goodness-of-fit measure is the R-squared which was defined by
Miaou, 1996. It estimates the percentage of variation explained by a
regression model. The model is based explicitly on the overdispersion
parameter
The R-squared goodness-of-fit test was computed as:
n
(3.12)
a
(3.13)
70


Where:
a = the overdispersion parameter estimated in the model
amax = the overdispersion parameter estimated in the negative binomial
model, the model with only a constant term and an
overdispersion parameter.
71


4. Test Results and Discussion
This section presents the results of the field validation testing effort. The
experimental procedure, prior to statistical testing, is summarized as follows.
A set of twenty-one interchanges, selected from field sites in Colorado, North
America, and were modeled in VISSIM. Each merge/diverge model was
simulated for five replications for 1 hour of simulated time. The corresponding
five output files (i.e., TRJ files) from VISSIM were then imported into SSAM
for identification of conflicts and computation of the surrogate measures of
safety for each conflict event. SSAM was configured to use conflict
identification threshold values. The maximum time-to-collision (TTC) and
maximum post-encroachment time (PET) values used were 1.5 seconds and
5.0 seconds, respectively. The results of the SSAM analysis consists of the
number of total conflicts and the number of conflicts of each type of vehicle-
vehicle interaction: crossing, rear end, and lane changing. Before any
statistical inferences were applied, the outliers were identified and removed
from the dataset.
Figure 4.1 shows five files added to the case files section. These files were
obtained from a model simulated for five replications. Note that the files
selected have a .trj extension. The default maximum TTC and maximum PET
values were not changed. Also the threshold angles (30 and 85 ) which
were determined by experimentation were not changed.
72


El SSAM C:\Maricos\CU\RESEARCH COMPONENTS\raSSERTA'nON-INTERCHANGES\E5 Hampden Ave (routing test)\I25 & Hampden Ave--NB(RS-MultiiTO)l
File Window Help
[iaaosnifflSa
0] E Yale Ave 1-2 _*
0} 1225 & 6th Ave-i
QJI22S&E Alameda;
0] 1225&E Alameda:
0] 1225 &E Colfax A
6}I22S&E Colfax I,
Q]l22S8>EMESBSt
[|]I22S&EtAss>S9f
0]I2S&E Arapahoe
0] I2S & E Arapahoe!
0]I25&EBeleview!
0] 125 AEBefeviev/
0]I2S&E County Li,
0]I2S&E County U.
0JI2S&E Dry Creel
0] 125 &E Dry Creel:
0] 125 &E Evans (R.L
0]I25&EttffAve-:"
0]I2S&EHffAve-J
0]I2S&EUncoHA/
0] 125&E Lincoln Av
0] 125&E Lincoln A/
0] 125 &E Orchard /
0] 125 &E Orchard/
0] 125&E Yale Ave-I
01125 & Hampden A1
OHHMHIgHt
0] 125 & Hampden A:
0] 125 AS Colorado I
0] 125&S Colorado I
0] 125&S Parker Re;
0] 125&S Parker Rcj
0]I25&SYosemtel
0]I25&SYosemite
0] 125 & University I
0] 125 & University I
0]I25&ridgegateF
0]I25&rdgegateF

0] 125 & Hampden Ave*-N8(RS-Multirun) | o
Case Information
Case Pile: C:^arkosV^J^ESEARCH COMPONB^TS'piSSERTATION-INTCHANGESy25 Hampden Ave (routing test)\I25 &Hampden Ave-NB^S-Muftrun)
| Comments: [Yotr comments here.
Analyst: |Your name here.
Analysis Date: Fri Feb 0316:56:43 MST 2012
[ Edt Info ~j Apply Change
Case Hies
Trajectory Res: fcWarkosVOJVESEARCH COMPON&ITSÂ¥)ISSStTATION-INTERCHANGES\I2S Hampden Ave (routing test)y25 & Hampden Ave-NB (RS-Multirun).l.trj
____________ C;\Marko$VCU1iRESEARCH COMPONS^TStiISS0tTATION-INTSlCH^IGES\I25 Hampden Ave (routing test)\I25 & Hampden Ave-TC (RS-MJtinjn)_2.trj
| Add | CiV^arkosVCUVtESEARCH COMPONB I-----------1 ClWarkosVCUVtESEARCH COMPON0n3PISSBLTATlON-INTERCHANGE5\I25 Hampden Ave (routing test)\!25 8. Hampden Ave-NB (RS+UlirunJjUrj
| Delete | C;ysarkosCUVtESEARCH COMPONENT5IDISSERTATION-INTERCHANGES\I25 Hampden Ave (routing test)\I25 & Hampden Ave-N8 (RS-Multinjn)_5.trj
Conflict Thresholds
Maximum time-to-colsion (TTC):
Maxrxrn post-encroachment time (PET):
Conflict angle Thresholds:
Rear end angle:
Crossing angle:
1.50-:-
5.00 -fcj
| Edit | | Apply | Analyze Q Check This to ouput frj files in Text Format
FIGURE 4.1. SCREEN CAPTURE. SSAM SCREENCONFIGURATION
TAB
SSAM contains a built-in feature to visualize conflicts. The conflicts can be
visualized within a Map tab or imported to SSAM as VISSIM .inp file.
Typically, the Map file is the graphic file of the simulated network for the case
document. The VISSIM .inp file is the VISSIM network file that has been
simulated and has generated the vehicle trajectory file that was used for
73


SSAM analysis. Figure 4.2 shows the display of conflicts in the Map view
section. The average hourly conflict counts for each merge/diverge locations
is provided in Appendix E.
f 0 SSAM C:\PemACIARESEARCH COMPONENTSDfS5ERTAT10N-tNTERCHANGES'\I25 Hampden Ave (routing test)\E5 & Hampden Ave~NB routing test
£4e fflindow £jetp
6] 122546th/
1225 46th/
122S 4 E 41;
1225 4 E At;
|]I22S4ECo
SI 1225 IE Co
1225 4 E Me
1225 4 E Mb
125 4 E A;*
0125 4EM
SI I254EBefc
1254EBe*
125 4 E Cry
I2S & E Evai
S]I254EEvai
I2S&E0fF
I25 4EIHf
125 4ELTO
I25 4ELro
3] 125 4 E ad
Q] I254EOrd
I2S & E rate
125 4 Erie
J 8] 125 & Hampden Ave-NB routing test
Mac Info
Map:
125 4 Hampden Ave-NB.rp
S]125 4Hamp.
01125 4 SO*
125 4 S Cote
125 4S Part
9jI254SPart
S]1254SYos<
01I25 45YOS.
125 4Urrvei
125 4Urvei.
TTC<1.0 Color
| ConSpuahyr CcrrVS Svmmafy EEtff Ttest Mg.
FIGURE 4.2 SCREEN CAPTURE. SSAM SCREEN MAP TAB VIEW AT
1-25 & E HAMPDEN AVE
74


4.1 Field Validation Test 1: Merging and Diverging Location Ranking
by Total Incidents
Validation Test 1 is a comparison of the ranking of merge/diverge locations
based on average hourly total conflict frequency versus the ranking of
merge/diverge locations based on average yearly total crash frequency. The
Spearman rank correlation coefficient (ps) values of ranking based on total
conflict comparison were 0.486 and 0.551 for merging and diverging
movements respectively, these values are significant at 95 percent level of
confidence.
As a basis of comparison, the merge/diverge locations were also ranked in
terms of total ADT values, and these were also compared with the ranking
based on average yearly total crash frequency. The Spearman rank
correlation coefficient (ps) values of this ranking comparison were 0.419
(significant at a 95 percent level of confidence) and 0.309 ( not significant at a
95 percent level of confidence) for merging and diverging movements
respectively, these values are shown in Table 4.2.
In general, the Spearman rank correlation tests showed that a higher
correlation was found for both merge and diverge movements with
merge/diverge locations rankings based on conflicts than ADT.
75


4.2 Field Validation Test 2: Merging and Diverging Location Ranking
by Specific Incident Types
Validation Test 2 repeats the same comparative ranking procedures as for
validation test 1 for the subsets of crash/conflict types: rear end and lane
changing. There were no crossing conflicts recorded to perform the ranking
comparison for crossing type incidents. Table 4.1 displays the distribution of
conflicts and crashes by incident type. There were 1 percent merging and 4
percent diverging lane changing conflicts were recorded, while the remaining
portion were rear end conflicts. However the proportion of crashes of the lane
changing and rear end maneuvers at merge and diverge, was 26 percent and
23 percent respectively.
There were significant differences between conflict distributions by type and
actual crash distributions by type. The ratios of conflicts-per-hour to crashes-
per-year for total incidents at merge and diverge were 117 and 137
respectively. The ratios of conflicts-per-hour to crashes-per-year for rear-end
and lane-change incidents at merge were 156 and 5 respectively. However
the ratios of conflicts-per-hour to crashes-per-year for rear-end and lane-
change incidents at diverge are 169 and 26 respectively. It seems plausible
that conflicts-to-crashes ratios are higher for less dangerous incidents. That
is, accepting that rear-ends are generally less severe than lane-change, there
is an abundance of these lower risk conflicts. In all the simulation scenarios,
rear-end conflict events made up the bulk of the total conflicts.
It was also observed that the total number of diverging conflicts were 42%
higher than that of the merging conflicts. Similarly, the total number of
diverging crashes are 22% higher than that of merging crashes This was
76


consistent with the study by Cillo (1967) and Lundy (1966) which showed that
the accident rates of off-ramps are higher than the accident rates of on-
ramps.
Figure 4.3 illustrates an example of a conflict, where some vehicles are
angling across lanes and have abruptly cut in front of other vehicles that must
decelerate to avoid collision.
%
i

%
FIGURE 4.3 SCREEN CAPTURE. VISSIM CONFLICT SCENARIO
RESULTING FROM LANE CHANGE MANEUVER
77


A higher proportion of rear-end conflict obtained from SSAM was consistent
with the study reported on crash prediction models on urban signalized
intersections. Guttman, Pu, Sayed, and Shelby (2008) modeled signalized
intersections in VISSIM and simulated under AM peak hour traffic conditions.
The conflict analysis was done using SSAM and found that the proportion of
crossing, rear-end, and lane-change conflicts were 1.7 percent, 91.0 percent,
and 7.3 percent respectively.
Moving on to the results of the ranking tests and looking into Table 4.2, there
is a significant correlation between conflicts and crashes when considered by
conflict type. The Spearman rank correlation coefficient (ps) value at merge
for rear end incident ranking comparison was 0.403, which is significant at a
95 percent level of confidence. Whereas, the Spearman rank correlation
coefficient (ps) value at diverge for rear end incident ranking comparison was
0.456, which is also significant at a 95 percent level of confidence. Moreover,
the Spearman rank correlation coefficient (ps) value at merge and diverge for
the lane changing incident ranking comparison were 0.540 (significant at a 95
percent level of confidence) and 0.550 (significant at a 95 percent level of
confidence) respectively.
For comparison, the merge/diverge locations were ranked on the basis of
total ADT values, and this was also compared with the rankings based on
rear-end and lane-change crashes. Rear-end ranking comparisons at merge
was not significant at 95 percent level of confidence. Moreover, the Spearman
rank correlation coefficient (ps) at diverge for rear-end incident ranking
comparison was not significant at a 95 percent level of confidence. However,
the Spearman rank correlation coefficient (ps) at both merge and diverge for
78


lane-changing ranking comparison were significant at a 95 percent level of
confidence.
In general, rank correlation testing showed a significant correlation between
rear-end conflicts and rear-end crashes and between lane-change conflicts
and lane-change crashes. Also, the rank correlation tests have shown that
stronger correlation was observed based on conflict than ADT.
Table 4.1 Distribution of crashes and conflicts by incident type
Incident Type
Merge Diverge
Rear Lane All Rear Lane All
End Change Type End Change Type
Average Hourly Conflicts 36,572 370 36,942 50,358 2,219 52,577
Percentage of Conflicts by Type 99% 1% 100% 96% 4% 100%
Average Yearly Crashes by Type 234 82 315 298 87 385
Percentage of Crashes by Type 74% 26% 100% 77% 23% 100%
Conflict-to-crash ratio 156 5 117 169 26 137
79


Table 4.2 Correlation coefficients by incident type
Incident Type
Ranking based on Merge Diverge
Rear Lane All Rear Lane All
End Change Type End Change Type
Crashes and Conflicts 0.403 0.540 0.486 0.456 0.550 0.551
Crashes and ADT 0.318 0.533 0.419 0.217 0.551 0.309
4.3 Field Validation Test 3: Crash and Conflict Prediction Regression
Model Paired Comparison
Validation Test 3 assessed the correlation between conflicts and crashes by
using regression to construct a conflicts-based model to predict crash
frequencies at merging and diverging locations. Additionally, the capabilities
of conflict-based crash-prediction were compared to a traditional volume-
based crash-prediction model.
A standard generalized linear modeling (GLM) approach was used to
establish a model of the expected crashes as a function of mainline, merging,
and diverging ADT. Crashes in these models are expressed in terms of
average yearly crash frequency, as a function of the model variables which
are the ADT volumes of the mainline, merging, and diverging (in vehicles per
80


day) or ADTma\n\\ne, ADTmerge, and ,4D7diVerge, respectively. The estimates of
parameters for these models are shown in Table 4.3 and Table 4.4. However,
conflicts are expressed in terms of average hourly conflicts (conflicts per
hour).
Table 4.3 and Table 4.4 show the estimates of the parameters of the total
crash model at merging and diverging locations respectively. Furthermore,
the table shows that both the Pearson chi-squared and the scaled deviance
values were not significant at the 90- percent confidence level, indicating a
good fit. Moreover, these models were compared using the R-squared
goodness-of-fit test (Miao, 1996), and the results conform to those of the
Pearson chi-squared and the scaled deviance. Moreover, graphs
corresponding to Model (1) and Model (2) were provided in Figure 4.4 and
Figure 4.5 respectively.
81


Table 4.3 Prediction model of total crashes as a function of mainline ADT
and merging ADT
Model (1) Total Crash/yr merge= 2.1 2E-04 X ADTmainline 1773 y ADT -209 a r\\-j i merge
Degree of R-Squared Scaled Pearson y2 x 0.1,30 Shape
Freedom R2 Deviance X2 Parameter
30 0.54 32.24 33.61 40.26 5.08
Variable Coefficient
Constant 2.12E-04
ADT mainline 0.773
ADT merge 0.209
Table 4.4 Prediction model of total crashes as a function of mainline ADT
and diverging ADT
Model (2)
Total Crash/yr diverge = 0.061 X ADT mainline X ADTdiverge
Degrees of R-Squared Scaled Pearson x20.i,3i Shape
Freedom R2 Deviance X2 Parameter
31 0.82 34.72 34.39 41.42 30.21
Variable
Constant
ADT mainline
ADT diverge
Coefficient
0.061
0.058
0.478
82


Total Crash per Year
Merging ADT
FIGURE 4.4 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT
AND MERGING ADT
83
Mainline ADT


Full Text

PAGE 1

SAFETY ASSESSMENT OF FREEWAY MERGING AND DIVERGING INFLUENCE AREAS BASED ON CONFLICT ANALYSIS OF SIMULATED TRAFFIC By Markos Alito Atamo B.Sc., Addis Ababa University, Ethiopia, 2002 M.Sc., Addis Ababa University, Ethiopia, 2005 A thesis submitted to the University of Colorado Denver in partial fulfillment of the requirements for the degree of Doctor of Philosophy Civil Engineering 2012

PAGE 2

2012 by Markos Alito Atamo All rights reserved

PAGE 4

Atamo, Markos Alito (Ph.D., Civil Engineering) Safety Assessment of Freeway Merging and Diverging Influence Areas Based on Conflict Analysis of Simulated Traffic Thesis directed by Professor Bruce N. Janson. ABSTRACT The safety of merging and diverging influence areas, intersections, interchanges, and other traffic facilities is asses sed by tracking and analyzing police-reported motor vehicle crash records. Since the nature of road crashes is random and infrequent, this process is slow to r eveal the need for remediation of the roadway design and traffic contr ol strategy. Moreover, this process is not applicable to assess new designs tha t have not yet been built, or deployed in the real world. This study summarizes a technique combining mi cro-simulation and automated conflict analysis to assess the safety of traffic facilities without waiting for a statistically above-normal number of crashes to occur. The technique is also valuable in assessing the relativ e performance of one design versus another. Traffic Conflict Technique ( TCT) is among the most common surrogate measures to study the safety of ro adway facilities. The software referred to as Surrogate Safety Assessment Model (SSAM) was developed by FHWA and was used in this study. The S SAM software

PAGE 5

application was designed to perform statistical ana lysis of vehicle trajectory data. Trajectory data is the output from microscopi c traffic simulation models. Among the various traffic simulation modeling tools VISSIM was chosen because of its compatibility with SSAM, its versati lity for analyzing networks of big sizes and its ability to provide users with the capability to model any type of geometric configurations. Forty-two merging and forty-two diverging influence areas in Colorado, USA, were modeled in VISSM and conflict analysis was per formed using SSAM under AM-peak traffic conditions. Five field valida tion tests were conducted. Conflicts predicted by SSAM approach were compared with actual crash records at merging and diverging influence areas in all statistical validation tests. The results of the validation effort of safe ty assessment based on conflict analysis of simulated traffic suggested th at this technique is recommended for safety assessment at merging and di verging locations. This abstract accurately represents the content of the candidateÂ’s dissertation. I recommend its publication. Signed Bruce N. Janson

PAGE 6

DEDICATION I dedicate this thesis to my loving and devoted wif e, Azeb, for her unfaltering support and understanding while I was completing it I also dedicate this to my daughters Yohana, Heldana, and Yosabet and to my parents who gave me an appreciation of learning and taught me the va lue of perseverance and resolve.

PAGE 7

ACKNOWLEDGEMENT My thanks to each of the members of my doctoral com mittee for their support during the completion of this thesis. Very special thanks to my advisor, Professor Janson for his wise guidance, instruction and unlimited support. I am very gratef ul for having the opportunity to learn from him.

PAGE 8

TABLE OF CONTENTS 1. IntroductionÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 1 1.1 BackgroundÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 1 1.2 Traffic Safety OverviewÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 2 1.3 Statement of the ProblemÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 4 1.4 Research Objectives and Contributions to the Transportation IndustryÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 5 1.5 Glossary of TermsQuick Reference GuideÂ…Â…Â…Â…Â…Â…Â… .. 7 1.6 Study Flow StructureÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 11 2. Review of the LiteratureÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 13 2.1 Traffic Conflict TechniqueÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 13 2.2 Merging and Diverging Movements on FreewaysÂ…Â…Â…Â… Â…. 15 2.3 Survey of Modeling DataÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 16 2.4 Interchanges and RampsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 17 Figures xiii Tables xv Chapter

PAGE 9

2.5 Background of Safety on Freeway RampsÂ…Â…Â…Â…Â…Â…Â…Â…. 22 2.6 Log-linear Regression ModelsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 23 2.6.1 Poisson Regression ModelÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 25 2.6.2 Negative Binomial Regression ModelÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 27 2.6.3 Goodness of Fit MeasureÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…... 29 2.7 Traffic Crash Prediction ModelÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 32 2.8 2.8.1 2.8.2 Empirical Bayes TechniqueÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…... Empirical Bayes Safety EstimateÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. Bayesian Identification of Accident-Prone Locations Â…Â…Â…. 35 36 38 2.9 Simulation ModelingÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 42 2.9.1 Choice of a Simulation ToolÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 42 2.9.2 VISSIM Micro-Simulation ToolÂ…Â…Â…Â…Â…Â…Â…Â…Â….Â…Â…Â…Â…. 4 3 2.10 SSAM Conflict Analysis ToolÂ…Â…Â…Â…Â…Â…Â…Â…Â….Â…Â…Â…Â…Â… 44 2.11 SAS/STAT Software and the GENMOD ProcedureÂ…Â…Â…Â… .. 49 3. Research Approach and MethodologyÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 51 3.1 Field ValidationÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 51 3.2 PurposeÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 51 3.3 Data Assembly and Geometric DefinitionÂ…Â…Â…Â…Â…Â…Â…Â…Â… 51 3.4 Simulation Modeling of Interchanges and Modeling AssumptionsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 55

PAGE 10

3.5 Identification and Removal of OutliersÂ…Â…Â…Â…Â…Â…Â…Â… Â…Â… 59 3.6 MethodologyÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 60 3.6.1 Field Validation Test 1: Merging and Divergin g Location Ranking by Total IncidentsÂ…Â….Â…Â…Â…Â…Â…Â…Â…Â…... 60 3.6.1.1 Conflict RankingÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 61 3.6.1.2 Crash RankingÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 61 3.6.1.3 Ranking ComparisonÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 62 3.6.2 Field Validation Test 2: Merging and Divergin g Location Ranking by Specific Incident TypesÂ…Â…Â…Â…Â…Â….. 63 3.6.3 Field Validation Test 3: Crash and Conflict Prediction Regression Model Paired ComparisonÂ…Â…Â…Â…Â… 64 3.6.4 Field Validation Test 4: Crash and Conflict Prediction Regression Model Comparative Analysis For Total IncidentsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 65 3.6.5 Field Validation Test 4: Crash and Conflict Prediction Regression Model Comparative Analysis For Specific Incident TypesÂ…Â…Â…Â….Â…Â….Â…Â…Â….. 69 3.7 Goodness of Fit MeasureÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 69 4. Test Results and DiscussionÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 72

PAGE 11

4.1 Field Validation Test 1: Merging and Diverging Location Ranking by Total IncidentsÂ…Â…Â…Â…Â…Â….Â…Â…Â…Â…... 75 4.2 Field Validation Test 2: Merging and Diverging Loca tion Ranking by Specific Incident TypesÂ…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â… 76 4.3 Field Validation Test 3: Crash and Conflict Prediction Regression Model Paired ComparisonÂ…Â…Â…Â…Â…. 80 4.3.1 IllustrationÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 91 4.4 Field Validation Test 4: Crash and Conflict Pre diction Regression Model Comparative Analysis for Total IncidentsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 93 4.4.1 Development of Conflict Prediction Regression Models for Total IncidentsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 93 4.4.2 Identification of Accident Prone Locations (A PL)Â…Â…Â…Â…Â…... 95 4.4.3 IllustrationÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 96 4.4.4 Ranking LocationsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 97 4.5 Field Validation Test 5: Crash and Conflict Prediction Regression Model Comparative Analysis For Specific Incident TypesÂ….Â…Â…Â….Â…Â…Â…Â…Â….. 101 4.5.1 Development of Conflict Prediction Regression Models for Specific Incident TypesÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…... 101

PAGE 12

4.5.2 Identification of Accident Prone Locations (A PL) to Rear End IncidentsÂ…Â…Â…Â…Â…Â…Â…Â…..Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 104 4.5.3 Ranking LocationsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 104 5. Summary, Conclusions, and RecommendationsÂ…Â…Â…Â…Â… 1 08 5.1 Summary and ConclusionsÂ…Â…Â…Â…Â…Â…..Â…Â…Â…Â…Â…Â…Â…Â….. 102 5.2 RecommendationsÂ…Â…Â…Â…Â…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 113 Appendix A Interchange traffic flow informationÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 116 B Average 3 year crashes recorded at merge and diverge locations Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 117 C Average yearly crashes recorded at merge and diverge locationsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 118 D Conflicts of five replications at merge and div erge locations recorded by SSAM softwareÂ…Â…Â…Â…...Â…Â…Â…Â…Â…. 119 E Average hourly conflicts at merge and diverge locations recorded by SSAM softwareÂ…............... ................... 120 BibliographyÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 121

PAGE 13

LIST OF FIGURES Figure 1.1 STUDY WORKFLOW STRUCTURE……….................... ............. 12 2.1 TYPICAL INTERCHANGE CONFIGURATIONS…………………. 19 2.2 TYPICAL RAMP CONFIGURATIONS..…………………………… 21 2.3 EVENT-FILE BASED INFORMATION FLOW DIAGRAM……..................................... ............................. 45 2.4 ILLUSTRATION OF CONFLICT ANGLE DIAGRAM…………… 49 3.1 ILLUSTRATION OF MERGE INFLUENCE AREA………………. 54 3.2 ILLUSTRATION OF DIVERGE INFLUENCE AREA…………….. 54 3.3 SCREEN CAPTURE. VISSIM GEOMETRIC MODEL OF I-25 & E HAMPDEN AVE ……………………………. 57 3.4 SCREEN CAPTURE. VISSIM TRAFFIC MODEL OF I-25 & E HAMPDEN AVE…………………………… 58 4.1 SCREEN CAPTURE. SSAM SCREEN – CONFIGURATION TAB…………………………………………… 73

PAGE 14

4.2 4.3 SCREEN CAPTURE. SSAM SCREEN MAP TAB VIEW AT I-25 & E HAMPDEN AVEÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… SCREEN CAPTURE. VISSIM CONFLICT SCENARIO RESULTING FROM LANE CHANGE MANEUVERÂ…Â…Â…Â…Â…. 74 77 4.4 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT AND MERGING ADTÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 83 4.5 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT AND DIVERGING ADTÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….. 84 4.6 TOTAL CRASH AS A FUNCTION OF TOTAL CONFLICT AT MERGEÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…... 88 4.7 TOTAL CRASH AS A FUNCTION OF TOTAL CONFLICT AT DIVERGEÂ…..Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 89

PAGE 15

LIST OF TABLES Table 4.1 Distribution of crashes and conflicts by incide nt typeÂ…Â…Â…Â….. 79 4.2 Correlation coefficients by incident typeÂ…Â…Â…Â…Â…Â… Â…Â…Â…Â… 80 4.3 Prediction model of total crashes as a function of mainline ADT and merging ADTÂ…Â…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â…Â….. 82 4.4 Prediction model of total crashes as a function of mainline ADT and diverging ADTÂ…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â…Â…. 82 4.5 Prediction model of total crashes as a function of conflicts at mergeÂ…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…... 86 4.6 Prediction model of total crashes as a function of conflicts at diverge.Â…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â… 86 4.7 Prediction model of conflicts as a function of mainline PHV and merging PHVÂ….Â…Â…Â…Â…Â…Â…Â….Â…Â…Â…Â… 94 4.8 Prediction model of conflicts as a function of mainline PHV and diverging PHVÂ…Â…Â…Â…Â…Â….....Â…Â…Â…Â…Â…... 94

PAGE 16

4.9 Comparison between total crash and total confli ct prone locations at mergeÂ…Â…Â…Â…Â…Â…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â….. 98 4.10 Comparison between total crash and total confl ict prone locations at divergeÂ…Â…Â…Â…Â…..Â…Â…Â…Â….Â…Â…Â…Â…Â…Â….. 99 4.11 Summary of APLs of total incidentsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…. 100 4.12 4.13 Average total crashes and total conflicts based on incident proneness of locationsÂ…Â…Â…Â…Â…Â…Â…Â…Â… Prediction model of rear-end crashes as a function of mainline ADT and merging ADT Â…Â…Â…Â…Â…Â…Â…Â….. 100 102 4.14 Prediction model of rear-end crashes as a function of mainline PHV and merging PHV.. Â…Â…Â…Â…Â…Â…Â…Â… 102 4.15 Prediction model of rear-end crashes as a function of mainline ADT and diverging ADT.Â…Â…Â…Â…Â…Â…Â…. 103 4.16 Prediction model of rear-end crashes as a function of mainline PHV and diverging PHV.Â…Â…Â…Â…Â…Â….Â… 103 4.17 Comparison between rear-end crash and rear-end conflict prone locations at mergeÂ….Â…Â…Â…Â…Â….Â…Â…Â…Â…Â…Â…Â… 105 4.18 Comparison between rear-end crash and rear-end conflict prone locations at divergeÂ…Â…Â…Â…Â…Â….Â…Â…Â…Â…Â…Â…... 106

PAGE 17

4.19 4.20 Summary of APLs of rear-end incidentsÂ…Â…Â…Â…Â…Â…Â…Â…Â….. Average rear-end crashes and rear-end conflicts based on incident proneness of locationsÂ…Â…Â…Â…Â…Â…Â…Â…Â….. 107 107

PAGE 18

1. Introduction 1.1 Background According to the National Highway Traffic Safety Ad ministration (NHTSA 2006) report, motor vehicle travel is the primary m eans of transportation in the United States, providing an unprecedented degree of mobility. Despite all its advantages, deaths and injuries resulting from moto r vehicle crashes are the leading cause of death for people of every age from 2 through 34. Traffic fatalities accounted for more than 90 percent of tr ansportation-related fatalities. National Highway Traffic Safety Administration (NHT SA 2009) stated that in the United States more than 5.5million police-repor ted motor vehicle crashes occurred. Of these 1.5million (28%) crashes resulte d in an injury, 30,797 (fewer than 1%) resulted in a death and 3.4million (73%) involved in property damage only. The fatality rate per 100 million VMT was 1.03. There were 8.6million vehicles involved in single a nd two-vehicle crashes. The number of vehicles involved in a merging/lane c hanging maneuver were 312,000 (3.6%). Resulting from these maneuvers were 769 fatalities, 48,000 injury, and 263,000 property damage only crashes. S imilarly, the number of vehicles involved in a going straight maneuver was 4.3million (50.4%). Resulting from these were 27, 124 fatalities 1.25mi llion injury, and 3.0million property damage only.

PAGE 19

The World Health Organization (2004) projected that between 2004 and 2030 global deaths increase by 28% which is predominantl y due to the increasing number of road traffic accident deaths. Road traffi c accident deaths are projected to increase from 1.3 million in 2004 to 2 .4 million in 2030, primarily due to the increased motor vehicle ownership and us e associated with economic growth in lowand middle-income countries Road traffic accidents are projected to rise from the ninth leading cause of death globally in 2004 to the fifth in 2030. According to the United Nations General Assembly (2008), by 2020 road traffic deaths and injuries will excee d HIV/AIDS as a burden of death and disability. 1.2 Traffic Safety Overview A traffic accident occurs when a road vehicle colli des with any physical object and results in injury, property damage, and death. Factors contribute to the risk of collision include; vehicle design, speed of operation, road design, and driver impairment. Worldwide motor vehicle collisio ns lead to significant death and disability as well as significant financial cos ts to both society and the individual. For the last two decades, several legislations have been making safety a central, explicit, comprehensive, and integrated pa rt of transportation planning. As a result of this, safety management sy stems have advanced to a great extent. Traffic safety data and analytical to ols have been improved and refined, and the effects of countermeasures have be come better understood.

PAGE 20

Since 1991 there have been three major Acts announc ed in the Unites States of America: (1) the 1991 Intermodal Surface Transpo rtation Efficiency Act (ISTEA); (2) the 1998 Transportation Equity Act for the 21st Century (TEA-21), and (3) the 2005 Safe, Accountable, Flexible, Effic ient Transportation Equity Act (SAFETEA-LU.) The Intermodal Surface Transportation Efficiency Ac t (ISTEA) of 1991 moved the historical focus of highway and transit program s away from construction, capacity, and congestion. The Act changed the empha sis towards mobility and access, system performance, and consideration f or the environment and quality of life. The Act did not specifically mention safety as part of the planning process but mandated six comprehensive management systems inclu ding a Safety Management System (SMS) as a prerequisite for fundi ng. The SMS was part of the strategy to improve the management, operatio ns, and safety of the highway system through improved data analysis and c ollection, through improved coordination, cooperation, and communicati on among agencies, and through the development of collaborative strate gic plans (Depue 2003). In 1998, the Transportation Equity Act for the 21st Century (TEA-21) called for comprehensive safety consciousness. The Act req uired state DOTs (and MPOs) to increase the safety and security of the tr ansportation system for motorized and non-motorized users. This was the fir st time that safety became an explicit part of transportation plans. Pr ior to TEA-21, safety was sometimes a prominent factor in project development and design, but this legislation calls for safety consciousness in a mor e comprehensive, systemwide, multimodal context, (FHWA 2001a).

PAGE 21

The Safe, Accountable, Flexible, Efficient Transpor tation Equity Act (SAFETEA-LU) became law in 2005. SAFETEA-LU has a s trong focus on integrated and comprehensive safety planning. It wa s built on previous legislation in giving specific and increasing recog nition to safety issues. The Act establishes the Highway Safety Improvement Program (HSIP) as a core program and nearly doubles the funds available for infrastructure safety and comprehensive strategic highway safety planning The purpose of the HSIP is to reduce fatal and serious/life changing c rashes. The program includes planning, implementation, and evaluation o f safety programs and projects. (Bahar, G., and Morris, N. (2007)). Generally, these legislations have encouraged or in sisted on giving safety priority through better data, better analysis, bett er reporting systems, and the adoption of a structured comprehensive approach. A detailed study of the safety of highways would help to understand safety and help to make safety legislation a success. 1.3 Statement of the Problem Until recently the safety of merging and diverging influence areas, intersections, interchanges, and other traffic faci lities is most often assessed by tracking and analyzing police-reported motor vehicle crashes over time. Since the nature of crashes is random an d infrequent, the process of analysis of safety based on the police r eport is slow to reveal the need for remediation of either the roadway desi gn or the flow-control strategy. This process is also not applicable to as sess new designs that

PAGE 22

have yet to be built, or to assess new flow control strategies before they are employed on-site. This study summarizes a technique combining m icro-simulation and automated conflict analysis to assess the safety of traffic facilities without waiting for a statistically above-normal number of crashes to occur. The technique is also valuable in assessing the relativ e performance of one design versus another in comparing alternative desi gns. 1.4 Research Objectives and Contributions to the Tr ansportation Industry The objective of this research was to assess the sa fety of freeway merging and diverging influence areas using conflict analys is technique and compare the results with the actual crash experience. SSAM conflict analysis software was used for predicting safety performances at the predefined locations. The Surrogate Safety Assessment Model (SSAM) is a softw are application designed to perform statistical analysis of vehicle trajectory data output from microscopic traffic simulation model, in this study VISSIM. VISSIM is a traffic simulation modeling tool which was chosen because of its compatibility with SSAM, its versatility for analyz ing networks of big sizes and its ability to provide users with the capability to model any type of geometric configurations or unique operation/driver behavior encountered within the transportation system. It is a time-based microscop ic simulation tool that uses various driver behavior and vehicle performance mod els to accurately

PAGE 23

n represent urban traffic and public transport operat ions. This study identified areas in need of further research. The study will make the following contributions to the industry: Helps in assessing the relative safety performance of one design versus another. Provide transportation agencies, planners, enginee rs, researchers with information about the type, frequency and relative location of conflicts. Provide solution that improves safety, access, and mobility of the merge/diverge influence area. Provide the safety analyst with a fast report to r eveal the need for remediation of the roadway design.

PAGE 24

1.5 Glossary of Terms – Quick Reference Guide Accident (traffic) : Event between road-users that results in injury, fatality or property damage. Accident causation : Underlying reasons that pre-empt a traffic accide nt, most usually involving an unforeseen chain-of-event s. Accident causation is often attributed to one of the three main component s of the traffic system: road-user, vehicle or roadway, or a combination of thereof. Accident outcome : Result of an accident in terms of injury severity fatality and in some cases also property damage. Accident rate : Number of accidents in accordance with a measure of exposure. Accident severity : Level of injury sustained in a traffic accident: usually categorized as minor, serious or fatal. Calibration : Process used in Traffic Simulation to (statistica lly) ensure that the functioning and behavior of a particular model and/or sub-model corresponds with observed empirical measurements or predetermined values. Car-following : Term used to describe the status of a vehicle tha t has a time/distance gap or headway less than a predetermi ned maximum value. Collision : Impact event between two or more road-users/vehic les, or a roaduser (vehicle) and stationary object.

PAGE 25

Collision course : Existence of a common projected conflict point in time and space for two (or more) road-users/vehicles, usuall y based on momentary measures of trajectory, speed and distance. Conflict : An observable situation in which two or more road users approach each other in time and space to such an extent that there is risk of collision if their movements remain unchanged. Conflict distance : A momentary measurement of (spatial) distance to a common conflict point for a road-user/vehicle in a conflict situation. Conflict point : Common spatial location of projected trajectories given momentary measures of speed and distance for two or more roadusers/vehicles. Conflict severity : Seriousness of a potential collision or near-acci dent measured by temporal or spatial proximity. Conflict speed : Momentary measurement of velocity for a road-user (vehicle) in a conflict situation. Conflict zone : Common area used by road-users/vehicles approachi ng from different trajectories. Crash : Term that is sometimes preferred to (traffic) acc ident due to the fact that it implies an element of causality rather than an unforeseen random occurrence.

PAGE 26

r Driver behavior : Largely misused and over-simplified term used in traffic engineering that is used to describe the actions an d/or variability of drivers in different driving situations. It should relate to t he study of individual behavioral processes that underlie driver actions (performance ). Exposure : Measure of spatial or temporal duration in the tr affic system in relation to the number of dynamic system objects ro ad-users, vehicles (axles), etc. Fatality : Death resulting from a traffic accident (usually within a 30 day period after the accident occurrence). Gap-acceptance : Process that describes and measures interaction b etween prioritized and non-prioritized road-users. General ly involves spatial or temporal measurement of gaps or lags in prioritized streams that are accepted or rejected in relation to a particular yi elding maneuver. Incident : the term used in a more abstract sense to refer to either crashes or conflicts. Injury accidents : Traffic accidents that result in minor or serious injury to one or more parties. Macroscopic (macro-) simulation : Simulation at a less detailed (aggregated, macroscopic) level. Microscopic (micro-) simulation : Simulation at a very detailed (microscopic) level.

PAGE 27

PET : It is the minimum post-encroachment time observed during the conflict. Post encroachment time is the time between when the first vehicle last occupied a position and the time when the second ve hicle subsequently arrived to the same position. A value of zero indic ates a collision. A postencroachment time is associated with each time step during a conflict. A conflict event is concluded when the final PET valu e is recorded at the last location where a time-to-collision value was still below the critical threshold value. This value is recorded in seconds. Police reported accidents : Accidents that are reported by the police and are recorded in the accident database of accident. Safety : Freedom from accident or loss. Simulation (traffic) : Abstract imitation of the operation of a real-wo rld process or system over time and event occurrence. T raffic simulation is concerned with the modeling of processes in the tra ffic system and can be conducted at different levels of abstraction depend ing on the purpose of the study. TTC : It is the minimum time-to-collision value observe d during the conflict. This estimate is based on the current location, spe ed, and future trajectory of two vehicles at a given instant. A TTC value is def ined for each time step during the conflict event. A conflict event is conc luded after the TTC value rises back above the critical threshold value. This value is recorded in seconds.

PAGE 28

1.6 Study Workflow Structure Figure 1.1 illustrates Study Workflow Structure. Ac cordingly, this research is divided into three Parts. Part 1 included introduct ion, information and data search. It consists of Chapter 1 and Chapter 2. Cha pter 1: Introduction, it provided an introduction to the study by presenting the background information, traffic safety overview, statement of the problem, study objectives and contributions to the transportation industry; a nd glossary of terms. Chapter 2: Literature Review: a comprehensive liter ature review on the safety performance of merging and diverging influence area s from published information was provided. It presented the overview of the merging and diverging movements on freeways, the survey and col lection of available modeling data, background of safety on freeway spee d change locations, regression models, Bayesian accident-prone location s, and simulation and conflict analysis tools. Also preliminary qualitati ve analysis of the data was carried out to identify the general validity of the selected variables. Part 2 included data analysis. It consists of Chapter 3: R esearch approach and methodology, this section described the study metho dology used in the field validation test and the steps taken to complete the study. Part 3 consists of Chapter 4 and Chapter 5. Chapter 4: Test results an d discussion, this section validated the actual crash data and discussed the r esults of the analysis. Chapter 5: Presented the summary, conclusions and r ecommendations drawn from the results of the study.

PAGE 29

FIGURE 1.1 STUDY WORKFLOW STRUCTURE Definition of Study Objectives Literature Review Design of Study Framework Theoretical C onsiderations Survey of Available Modeling Data Data Collection Refinement of Crash/Conflict Data Simulation Modeling using VISSIM Import of TRJ files from VISSIM, Conflict analysis using SSAM Development of Regression Model SAS/SAT, GENMOD Procedure Comparative Analysis of Actual Crashes and Predicted Conflicts Test Results, Discussion, Conclusion, and Recommendation Part 1: Introduction, Information and Data Search (CH.1 &CH.2) Part 3: Test Results and Conclusion (CH.4 & CH.5) Part 2: Data Analysis. (CH.3) Research Methodology/ Field Validation Test

PAGE 30

2. Review of the Literature 2.1 Traffic Conflict Technique According to Amundsen and Hyden (1977), a conflict is defined as an observable situation in which two or more road user s approach each other in time and space to such an extent that there is risk of collision if their movements remain unchanged. The traffic conflict technique method initially dev eloped at the Detroit General Motors laboratory in the late 1960Â’s for identifyin g safety problems (Perkins and Harris, 1968). It was based on observer judgmen ts using time-lapse filming, and proves a costly and time consuming tec hnique. They correlated conflict patterns to accident types. Spicer (1973) studied about the safety of six intersections and concluded that there was good correlation between serious conflicts and reported injury accidents. Sa yed (1997) conducted a study to estimate the safety of unsignalized inte rsections using traffic safety technique. A computer simulation model was used to study traffic conflicts with time-to-collision as the critical traffic even t in simulating driver behavior. Using the data collected from 30 conflict surveys, he established traffic conflict frequency and severity standards for un-si gnalized intersections. These standards allow the relative comparison of th e conflict risk at various intersections. Miglez et al. (1985) developed a methodology to predict traffic accidents from observed conflicts. Also, statistical procedures were developed to determine conflict rate values that co uld be considered "abnormally" high. The study concluded that traffic conflicts are good

PAGE 31

surrogates of accidents in that they produce estima tes of average accident rates nearly as accurate as those produced from his torical accident data. Gettman, Pu, Sayed, and Shelby (2008), studied a me thod of safety assessment utilizing a traffic conflicts analysis t echnique applied to simulation models of intersections, interchanges, and roundabo uts. They developed a software application to automate the task of traffi c conflicts analysis and conducted validation testing to gauge the efficacy of the assessment method. Several surrogate traffic conflict measures have be en developed to assess the safety of highway facility. The common surrogat e conflict measures are gap time, encroachment time, deceleration rate, pro portion of stopping distance, post-encroachment time, initially attempt ed post-encroachment time, and time-to-collision. There is general consensus that higher rates of tra ffic conflicts can indicate lower levels of safety for a particular facility, g iven the fact that conflicts generally result from a lack or misunderstanding of communication between the different road users (Risser 1985 and Archer 20 00) In this study, traffic conflicts occur with adequat e frequency to overcome the statistical challenges posed by infrequent crashes. Also, since adequate conflict data can be collected in a relatively shor t time, conflict analysis is not subject to the problem of changing of the underlyin g conditions (e.g., traffic volumes, geometry) that affect long-term crash reco rds.

PAGE 32

2.2 Merging and Diverging Movements on Freeways According to Roess, R., Prassas, E., and McShane W. (2004), merging occurs when two separate traffic streams join to fo rm a single stream. Merging can occur at on-ramps to freeways or multilane highways, or when two significant facilities join to form a single tr affic stream. Merging vehicles often make lane changes to align themselves in lane s appropriate to their desired movement. Diverging occurs when one traffic stream separates to form two separate traffic streams. This occurs at o ff-ramps from freeways and multilane highways, but can also occur when a major facility splits to form two separate facilities. Again, diverging vehicles must properly align themselves in appropriate lanes, thus indicating lane-changing; n on-diverging vehicles also make lane changes to avoid the turbulence created b y diverge maneuvers. Therefore, a merge area would always follow an on-r amp whereas a diverge area would always precede an off-ramp. Road users i n such areas are required to select and adjust speeds without causin g undue hazards to themselves or to other road users. Such hazards wou ld normally increase in terms of frequency and intensity with increasing vo lumes of vehicles entering and exiting a specific segment on the freeway. This increase may result from the combination of increasing number of required la ne changes and decreasing probability of a suitable gap between mo ving vehicles to perform a merge or diverge operation (Sarhan, M.; Hassan, Y.; and Abd El Halim, A.O. 2008).

PAGE 33

n 2.3 Survey of Modeling Data Since the nature of road crashes is random, the cho ice of the analysis period has a significant impact on the accuracy and reliab ility of the safety assessment. Accident counts that are recorded for o verly extended time period may introduce biases in the analysis when cu rrent traffic and geometric conditions differ from those prevailing w hen the crashes occurred. Similarly, an overly short period reduces the numbe r of crashes considered and hence the statistical accuracy of the model wou ld be affected. According to the Permanent International Association of Road Congresses PIARC (2003) Road Safety Manual, the minimum accepted ana lysis period is 3 years. For this study, a three year data from year 2008 through 2010 was used which was obtained from the Colorado Departmen t of Transportation (CDOT) accident database. The greatest statistical challenge of the safety as sessment of a facility in accurately pinpointing an underlying crash rate is the nature of motor vehicle crashes. Crashes are so infrequent and exhibiting v ariable hourly and yearly crash counts. The problem is increasingly difficult for areas with lower (more infrequent) crash rates. There is possible bias in comparing peak-hour conflicts to annual crashes. The statistical models that correlate the peakhour crash with the peak hour conflict on the mergi ng and diverging locations may not converge as there are too few peak-hour cra shes (which are less than 25% of the total crashes). Field validation th rough estimation with a larger data set would result in a more significant parameters estimates and more accurate crash prediction.

PAGE 34

Twenty-one interchanges, forty-two merging and fort y-two diverging influence areas were selected. A preliminary assessment was m ade of the types of interchange elements that were in sufficient number s and had sufficient data available for statistical modeling of accidents to be conducted. Accordingly diamond interchange configurations were selected. Using each interchangeÂ’s milepost as a common ident ifier, modeling data was collected for each interchange taking the following into consideration: 1. Interchange layout: number of lanes on mainline and ramps, ramp length, and ramp configuration. 2. Traffic volumes: average daily traffic (ADT), d esign hourly volume (DHV), percent trucks, and on-ramp hourly traf fic volume recorded by ramp meters. 3. Data on crash frequency on mainline and ramps. 4. Location of crash occurred on the highway. Only crashes at the merge and diverge influence areas of the interchange were considered. 5. Accident Type: accidents were regrouped into th ree: rear-end, crossing and lane-changing. Only vehicle-to-vehicle acc idents were considered. 6. Direction: the direction of traffic flow at the time of crash: North, Northeast, Northwest, East, West, South, South east, Southwest. 2.4 Interchanges and Ramps Grade separation of intersecting roadways reduces c rashes caused by crossing and turning movements. An interchange is a system of interconnecting roadways in conjunction with one or more grade separations that provides for the movement of traffic between t wo or more roadways or highways on different levels. Access control is a h ighly desirable feature

PAGE 35

along the crossroad at an interchange to provide ef ficient traffic operations and safety. The most common roadways connections at an interchange consist of freeway-to-freeway (access-controlled fr eeways), freeway-toarterial and arterial-to-arterial. Freeway-to-arte rial and arterial-to-arterial connections are non-access-controlled highways. Thi s research focused exclusively on a freeway-to-arterial interchanges. There are several basic interchange configurations to accommodate turning movements at a grade separation and it is determine d by number of intersection legs, expected volumes of through and turning movements, composition and type of traffic, topography, desi gn control, proper signing, and physical constraints such as existing rivers, r ailroads, and roadways. All freeway interchanges with non-access-controlled highways should provide ramps to serve all basic directions to prevent wron g-way movements. According to AASHTO 2004, the term “ramp” includes all types, arrangements, and sizes of turning roadways that co nnect two or more legs at an interchange. The different ramp patterns of an i nterchange (i.e. the different types of interchange configurations) are made up of various combinations of ramps. Figure 2.1 illustrates typi cal interchange configurations: from the simplest full-diamond inte rchange to complex, multilevel directional interchanges. Many variations of each of these basic interchange configurations arepossible.

PAGE 36

r FIGURE 2.1 TYPICAL INTERCHANGE CONFIGURATIONS (BAUER, M. AND HARWOOD, W., 1998)

PAGE 37

A one way road that leaves a mainline highway facil ity is called an off-ramp or exit ramp. A one way road that joins a mainline hig hway facility is called as an on-ramp or entrance ramp. This distinction is impor tant because according to Cirillo (1967) & Lundy (1966), the accident rate of on-ramps was consistently lower than off-ramp accident rates. Accordingly an on-ramp accident of 0.59 accident per MVMT and off-ramp accident of 0.95 acc ident per MVTM have been confirmed. Vehicles typically travel along off -ramps at higher speeds than along on-ramps, so that accidents are more lik ely to occur on off-ramps. Figure 2.2, illustrates a number of typical ramp co nfigurations. Each of the ramps of the ramp configurations illustrated traffi c exiting from a mainline freeway, but an analogous ramp configuration for tr affic entering the mainline freeway also exists.

PAGE 38

FIGURE 2.2 TYPICAL RAMP CONFIGURATIONS (BAUER, M. AND HARWOOD, W., 1998)

PAGE 39

2.5 Background of Safety on Freeway Ramps Numerous studies have studied the safety performanc e of freeway ramps during the past several decades. Bared, J., Giering G., Warren, D., (1999) developed the statistical model of accidents to est imate accident frequencies for entire ramps as a function of speed change lane length among other variables. According to the accident model develope d in that study, the longer speed change lane showed the less accident frequenc y. The result of the study by Lord, D., Bonneson, J., (2005), showed that the exit ramps are more dangerous and the non-free-flow type ramp experience twice as many accidents as other types of ramps. Previous researchers have developed several crash prediction models to relate crash frequency at ramp sections to different explanatory variables such as traffic volumes and ramp design elements. Bauer, M, Harwood, W. (1998), studied the relations hip between traffic crashes and highway geometric design elements and t raffic volumes for interchange ramps and speed change lanes. The stati stical modeling approaches used in that research included Poisson a nd negative binomial regression. Several models were developed to predic t crashes on ramp sections and speed change lanes. The variables that were included in crash models included mainline freeway AADT, ramp AADT, a rea type (rural/urban), and ramp type (on/off), ramp configu ration, right shoulder width, and lengths of ramp and speed-change lane. The regr ession model is presented in Section 2.6 of this study. Accident rates decrease as length of weaving area, or acceleration and deceleration lane increases (Cirillo 1967, 1968, 19 70). Cirillo developed statistical model and provided a relationship betwe en the length of weaving

PAGE 40

and acceleration and deceleration lanes and traffic volume levels and percentage of merging and diverging traffic. Increa se in traffic volume was associated with an increase in crash. The effect of acceleration lane length on accident rate was significant when merging traffic percentage exceeded 6% of the main line traffic volume. As the percentage of merging traffic increase beyond this volume, the additional length of accele ration lane provides a significant reduction in accidents. The effect was not as great for deceleration lanes compared to acceleration lanes. Studies suggested that a good portion of the accide nts associated with ramps occur at the entrance and exit of ramps. Mullins an d Keese (1961) found that 23% of all through-lane accidents occur near entran ce terminals. 52% of the on-ramp accidents according to Lundy (1966) occur a t entrance terminal. McCartt, A., Northrup, V., Retting, R. (2004) studi ed about the types and characteristics of ramp-related motor vehicle crash es on urban interstate roadways in Northern Virginia. The study found that 48% accidents occurred in the process of exiting freeway, 36% occurred whe n the traffic entering and 16% occurred in the midpoint. 2.6 Log-linear Regression Models Bauer, M. and Harwood, W. (1998) developed statisti cal model for interchange. These models are Log-linear regression models. They included Poisson and negative binomial regression models. St atistical background on both the Poisson and negative binomial models is pr ovided next.

PAGE 41

The following notations are used: n = interchange elements (e.g., ramp proper segme nts, entire ramps). = is a set of q parameters n (e.g. the geometric design, traffic volume, and other related chara cteristics of that element. = the number of accidents occurring at the r element during a specific period (say 3-year): where = the actual observation of during the 3-year period, where = the expected number of accidents at the r element and the q parameters, Note: In section 2.5 of this study all logari thms are natural logarithms and are denoted by log. The main objective of a statistical model is to dev elop a relationship and q parameters, !!!". This relationship can be formulated through a gener al linear model (GLM) of the form: #$%&' ()*(!*+*("!"

PAGE 42

where: ()(((" are the regression coefficients and estimated from the data. The estimation procedure used to obtain the regress ion coefficients is dependent on the assumption made about the distribu tion of the 2.6.1 Poisson Regression Model The assumption of a log-normal distribution [i.e., the assumption that ,'follows a normal distribution] is not valid when th e average number of accidents at a ramp is small. The Poisson model the n becomes a natural choice as it models the occurrence of rare discrete events well. The Poisson distribution has a mean and variance, both equal to According to the model formulated by Bauer and Harw ood (1998): ,'()*/(0!0 01 Assuming the number of accidents,, follows a Poisson distribution with mean the probability that a ramp with accidents can be expressed as

PAGE 43

n 2 3 4567859 : where: 9 denotes the factorial of Maximum likelihood estimation method is used to est imate the Poisson regression coefficients ()((" Maximization of the likelihood function is performed by taking the derivative of the likelihoo d function with respect to the parameter value. However, it is convenient to take the natural log of the likelihood function before differentiating. Since t he natural log is a monotonic function, maximizing the log-likelihood is equivale nt to maximizing the likelihood. So, the likelihood is: ; <<=> ?@ A> B7=C>DEF+FDG<9HH<=9 = 1 I The log-likelihood over all n interchange elements in the category of elements is therefore: JKL ; M>* <*+*<= ,'> M,'<9HHA=9 N

PAGE 44

Which is equivalent to: ,'O / P ,'M M,'9 QR 1 S 2.6.2 Negative Binomial Regression Model As it has been discussed in Section 2.5.1, in Poiss on distribution the mean and the variance of the distribution are equal. How ever, previous traffic accident research has shown that this is not always the case. Paternoster and Brame, (1997); Osgood, (2000) sugge sted the negative binomial distribution as an alternative to the Pois son when there is evidence of overdisperson. Overdispersion is a phenomenon that implies there i s more variability around the model's fitted values than is consistent with a Poisson formulation. Suppose a Poisson model is used for modeling accide nts and if the variance (or dispersion) of the data exceeds the estimated m ean of the accident data distribution, then the data are said to be overdisp ersed, and the underlying assumption of the variance being equal to the mean for the Poisson distribution is violated (Bauer, M. and Harwood, W. (1998)). The negative binomial is proposed as a means to correct for this problem (Osgood, 2000). Negative binomial distribution has two parameters, unlike the Poisson distribution.

PAGE 45

For negative binomial distribution the relationship between the expected number of accidents occurring at the TU element and the q parameters is same as Equation (2.1). The probability that a ramp accidents can be expressed as: V < WX<* Y Z W <* WX Y Z Y> D *Y> DF [ 3<\ The mean and variance of the negative binomial dist ribution of accident frequency are given by: ]6 ^ _`%6_` *a> b where: Y the negative binomial dispersion parameter The model regression coefficients ()(("are estimated by the method of maximum likelihood minimizing the negative of th e log likelihood. Bauer, M. and Harwood, W (1998) developed the log l ikelihood for the negative binomial distribution and is given by:

PAGE 46

r ,'O / R 1,'-c d d* eMa,'*d *f$%&''fa Modifying Equation (2.10): ,'O / R 1,'-c a a* eM Y ,'* a *f$%&''fa Which is equivalent to: ,'-g X Y Z h()*(!*(!*+*("!" 2.6.3 Goodness of Fit Measure The scaled deviance and PearsonÂ’s chi-square statis tic are helpful in assessing the goodness of fit of a givengeneralized linear model.According to Nelder and Wedderburn (1972), for a fixed value of the dispersion parameter i scaled deviance is defined to be twice the differe nce between the maximum achievable log likelihood and the log l ikelihood at the maximum likelihood estimates of the regression parameters.

PAGE 47

Let J A> the log-likelihood function, where: > the mean of the distribution < the response values, Then the scaled deviance is defined by: jk <> JJ << MJ <> : jk <> k <> i I where: k <> model (residual) deviance i= the generalized linear model dispersion parameter Accordingly, scaled deviance for the Poisson distri bution is given by Equation (2.15): lm/,n oM/ M R 1 R 1p N

PAGE 48

Also,scaled deviance for the negative binomial distribut ion is given as: k/ q oMn<* r oJKL s <* Y >* r tu S = 1 Y the negative binomial dispersion parameter Since i is one for both Poisson distribution and negative binomial distributions the scaled deviance is equal to the d eviance. PearsonÂ’s chi-square statistics is given by: v/ w < \= 1 and the scaled PearsonÂ’s chi-square is defined as: vi ^ where: >the mean of < w < the variance function of the distribution.

PAGE 49

In both Equation (2.16) and (2.17) the weight to ea ch interchange is to be one. Since i is one for both Poisson distribution and negative binomial distributions the scaled PearsonÂ’s chi-square is eq ual to the PearsonÂ’s chisquare statistics. 2.7 Traffic Crash Prediction Model Bauer, M. and Harwood, W. (1998) developed models t o predict accidents. Initially accidents on ramps and speed-change lanes were modeled separately with the thought that accident predictio ns from the separate models could be added together to determine the com bined safety performance of a ramp and its adjacent speed-change lane. The study proved that the separate models, by themselves, didn't pro vide an adequate fit to the data. Therefore, they developed the best models of the safety performance of interchange ramps and speed-change lanes by combini ng the accident experience of ramps and their adjacent speed-change lanes into a single model. Independent variables included in the model include : Ramp AADT Mainline freeway AADT Area type (rural/urban) Ramp type (off/on) Ramp configuration Length of speed-change lane Ramp length

PAGE 50

It was determined that these seven independent vari ables are all statistically significant at the 20-percent significance level in the model for total accidents, and all of these independent variables, except ramp configuration, are significant in the model for fatal and injury accid ents. To predict the average accident frequency for a ram p including the adjacent speed-change lane, the regression coefficients x)xxxn are replaced in equation (2.19) by the estimated values of the coef ficients and the variables n are replaced by their appropriate values or levels @yz{K > BA| x) }}k~€‚ƒ„BA| x ……BA|xnnb For example, the expected 3-year total accident fre quency can be estimated as B7†† )†‡ )ˆBA| INˆ BA| \^‰ BA| MŠ BA| Sb‹ BA| M:\† BA| :\‡ BA| MNbŒ BA| S) where: expected number of total accidents in a 3-year per iod on entire ramp plus adjacent speed-change lane ramp AADT (veh/day) mainline freeway AADT for the direction of travel i n which the ramp is located (veh/day)

PAGE 51

ˆ 1 if the ramp is a diamond ramp; 0 otherwise ‰ 1 if the ramp is a parclo loop ramp; 0 otherwise Š 1 if the ramp is a free-flow loop ramp; 0 otherwis e ‹ 1 if the ramp is an outer connection ramp; 0 other wise † 1 if the area type is rural; 0 otherwise ‡1 if the ramp is an off-ramp; 0 otherwise Œ speed-change lane length (mi) ) ramp length (mi) The estimate of dispersion after fitting, as measur ed by the deviance and Pearson’s chi-square divided by the degrees of free dom, are 1.0 and 0.95 respectively for total accidents which are well wit hin the acceptable range. Also the R2 goodness of fit is approximately 0.38. Among these variables, it was found that mainline a nd ramp AADT to be the most important predictor of accident on ramps. The increase in crash frequency was associated with higher accidents. The effect of acceleration length was more complicated since on one hand longe r lengths provide safer traffic maneuvering and on the other hand correspon ding increased length may provide longer exposure and more accidents. It was noted that the 3-year accident prediction mo dels were divided by 3 to obtain the corresponding accidents per ramp per yea r.

PAGE 52

2.8 Empirical Bayes Technique Bayes theorem is a probability theory that shows ho w new information can be used to update or revise an existing set of probabi lity. The prior probability may have to be estimated based on very little infor mation. These probabilities would be improved as more information becomes avail able. When it comes to traffic safety, Empirical Bayes (E B) is a theoretical framework that combines the actual crash count of a location with an estimated crash frequency determined from the safet y performance function of the same location. Sayed (1995) stated that the expected number of accidents at a location is a random variable that f luctuates around some unknown mean. This randomness is reason that histor ical accident data at a location does not always accurately reflect its lon g-term accident characteristics. Higle and Witkowski, (1988) state d that a location that has low accident frequency during long periods of time may have had high accident rates during portions of this period and v ice versa and confirmed that Empirical Bayes approach accounts for random variat ions of accident rate. The technique is based on Bayes’ theorem which can be mathematically described as: V iA V A  i HVi Ž VAiHVi

PAGE 53

n Where: i = a parameter such as the number of accidents at a location, V i the prior distribution of i V Ai the probability of making A observation for a specific value of i(observation distribution), and V iA the posterior distribution of iwhich represents the resolution of the prior distribution given the observ ations. According to Calvin (1990) the distribution of the observation is assumed to be a Poisson or Binomial distribution and the prior distribution will be a gamma or beta distribution. In the Empirical Bayes approach, the parameters are estimated using a sample of observations from p opulation of similar locations. 2.8.1 Empirical Bayes Safety Estimate Hauer (1992), Brude and Larsson (1988) stated that the safety of a location provide clues about its traffic and road characteri stics, and its historical accident data. The Empirical Bayes (EB) approach ma kes use of both kinds of clues and is used to refine the estimate of the expected number of accidents at a location by combining the observed n umber of accidents at the location with the predicted number of accidents obt ained from the GLM model to yield a more accurate, location-specific safety estimated. Equation 2.31 confirmed this. Brude and Larsson, (1988) showed th at EB method significantly reduces the regression to the mean ef fects that are inherent in observed accidents count. Regression to the mean is a statistical phenomenon that refers to the tendency of extreme e vents (high number of

PAGE 54

accidents) to be followed by less extreme values (a lower number of accidents) even if no treatment is applied to the f ield configuration that generates the accidents at that location. According to Hauer (1992), the EB safety estimate c an be calculates by. ‘€’“TD“‘T€T“”|•B–* M” zKy{ where, ” —• |•B– |•B– where, count = observed number of accidents at the location pred = predicted number of accidents as estimated from the GLIM model var(pred) = the variance of the GLM estimates According to (Sayed et.al., 1998), —• |•B– |•B– r

PAGE 55

Where: rshape parameter Then, Equation (2.22) can be rearranged as: ‘€’“TD“‘T€T“n r r*|•B– o|•B–*n |•B– r*|•B– ozKy{: In addition, the variance of the EB estimate can be calculated using (Kulmala, 1995): —•‘€’“TD“‘T€T“ r|•B– r*|•B– *n |•B– r*|•B– ozKy{I 2.8.2 Bayesian Identification of Accident-Prone Loc ations Higle and Witkowski (1988) described that Empirical Bayes procedure is a useful technique to identify accident-prone locatio ns. The following notations are used: ™ š accident rate at a location (treated as a random variable), ›number of accidents at location during the period of time in question, wnumber of vehicles passing through location during the period of time in question, @ ™›w probability density function associated with the ac cident rate at location given the observations › and w (Posterior distribution)

PAGE 56

r @œ ™ probability density function associated with the a ccident rate across the region (Posterior distribution) and ”x= parameters of gamma distribution. Two assumptions were made: 1. At any given location, the actual number of acci dents › follows a Poisson distribution such that at any given loc ation, where the accident rate is known ™ š™ and the expected value is given by ™w then the observation probability distribution is given by the following: V›™ š™w ™w=9 B7Ÿ E N 2. The probability distribution of the regional acc ident rate (the prior distribution), @œ ™ follows a gamma distribution and is given by: @œ ™ x¡W ” ™¡7B7ƒŸ S The parameters ” and x of the prior distribution are estimated using the method of moments estimates (MME), and are chosen s o that the mean and variance associated with the gamma distribution are equal to the mean A and variance ¢ of the sample.

PAGE 57

If A and ¢ are the sample mean and variance of the observed a ccident rates respectively, and m is the number of location s under observations, then: A £ / ›w  1 \ ¢ £M /n ›w MA o 1 ^ Morris (1988) and Berger (1985) modified the param eters of gamma distribution to avoid the bias and improve the esti mation. Accordingly, the probability density function associated with the ac cident rate at a location is given by: @ ™›w x ¡EW ” ™¡E7B7ƒEŸ b And thus, location will be identified as accident-prone if there is a significant probability that the location’s accident rate,™ exceeds the observed regional accident rate, œ.

PAGE 58

Thus, location is identified as accident prone if: V¤™ œ›w¦§ : or equivalently if: ¨M x ¡EW ” ™¡E7B7ƒEŸ–™)§ : where: § the confidence level desired, such as 0.95 or 0.99 and The value of the regional accident rate, œ, is calculated using: œ Ž › 7Ž w 7 : Sayed (1995) modified the Higle and Witkowski’s met hod by substituting ”*™Ewfor ” in Equation (2.31) and is given by: ¨M x X¡FŸ E EZW”*™Ew )™X¡FŸ E E7ZB 7ƒEŸ –™§ ::

PAGE 59

2.9 Simulation Modeling 2.9.1 Choice of a Simulation Tool A need for simulation modeling has been increasing to understand the behavior of traffic as transportation systems have become more complex. Many researches in the past have studied different similarities and differences of simulation tools and their capacity to model dif ferent combinations of traffic, highway type, geometric configuration, etc. of the system. Gettman D., and Head L., (2003) conducted a study t o evaluate the various simulation modelsÂ’ capabilities for producing measu res of intersection safety and specify algorithms for calculating the measures The nine multipurpose micro simulation software includes CORSIM, SIMTRAFF IC, VISSIM, HUTSIM, PARAMICS, TEXAS, AIMSUN, WATSIM, and INTEGR ATION. It was noted that VISSIM, AIMSUN, PARAMICS, and TEXAS outputs the proper TRJ files required by SSAM. It was concluded that there are a number of advantages that VISSIM possesses over other micr o-simulation tools. VISSIM appears to be a full-featured microscopic si mulation model with the ability to obtain detailed state variable informati on on each vehicle on time scales with better than second-by-second accuracy. Moreover it has been interfaced to other external codes such as hardware signal controllers, thus the developers have experience in development colla boration. The priority rules feature of VISSIM appears to allow complex mo deling of junction behavior, including friendly merging (situations wh ere following vehicles will slow for merging vehicles to create a gap), as it o ccurs in the real world. It is not apparent that other simulation models are able to represent such behavior. VISSIM allows multilane merging behavior; it is typical for vehicles entering the mainline flow to cross the path of an oncoming vehicle traveling

PAGE 60

in the same direction as the intended direction of travel of the entering vehicle and start accelerating in the adjacent lane. In thi s way, the oncoming vehicle can continue at its current speed without having to break for the turning vehicle. VISSIM appears to support most of the mode ling features required for obtaining surrogate measures at a reasonable le vel of fidelity. Because of its competitive advantages, combined wit h its capability to analyze networks of all sizes, VISSIM was chosen to perform the simulation modeling of traffic. 2.9.2 VISSIM Micro-Simulation Tool According to VISSM User Manual (2009), VISSIM is a microscopic, time step and behavior-based simulation model developed to mo del urban traffic and public transport operations. The program can analyz e private and public transport operations under constraints such as lane configuration, traffic composition, traffic signals, etc., thus making it a useful tool for the evaluation of various alternatives based on transportation eng ineering and planning measures of effectiveness. The model was developed at the University of Karlsr uhe, Germany during the early 1979. PTV America, Inc., a subsidiary of PTV AG located in Karlsruhe, Germany, is the North American distributor and deve loper of PTV Vision software products. VISSIM version 5.20 was used in this study. Unlike many other simulation tools, it is not stric tly node-link based, it is based on link-connector geometry. In VISSIM the user buil ds a network based on

PAGE 61

aerial photographs at the exact point where they ar e needed to have the desired effect on road-users. Traffic flow in VISSIM is simulated by moving drive r-vehicle-units through a network. Every driver is stochastically assigned to a specific vehicle. Consequently, the driverÂ’s behavior will be adapted to the vehicle performance characteristics. Attributes characteriz ing each driver-vehicle-unit can be classified in to three components: (1) techn ical specification of the vehicle (e.g. length, engine power, weight, maximum speed, potential acceleration actual position in the network, and ac tual speed and acceleration), (2) behavior of driver-vehicle units (e.g acceleration based on current speed and driverÂ’s desired speed, (3) inter dependence of drivervehicle units (e.g. reference to leading and follow ing vehicles on own and adjacent travel lanes) 2.10 SSAM Conflict Analysis Tool The Surrogate Safety Assessment Model (SSAM) is own ed by FHWA. According to Lili Pu and Rahul Joshi (2008), SSAM o perates by processing data describing the trajectories of vehicles drivin g through a traffic facility and identifying conflicts. The vehicle trajectory input data for SSAM are generated by traffic simulation software in a trajectory file format (where files are labeled with a .trj file extension), specially designed for SSAM. SSAM calculates surrogate measures of safety corresponding to each vehicle-to-vehicle interaction and determines whether or not each inte raction satisfies the criteria to be deemed an official conflict. A table of all identified conflicts and their corresponding surrogate safety measures is th en presented to the user.

PAGE 62

Figure 2.3 illustrates the workflow for using SSAM. It was reported that the users of the SSAM software would include traffic sa fety engineers, traffic simulation analysts, and traffic researchers. FIGURE 2.3 EVENT-FILE BASED INFORMATION FLOW DIAGRAM ( Gettman D., and Head L., (2003)). Surrogate Safety Assessment (SSAM) software uses tw o threshold values for surrogate measures of safety to delineate which veh icle-to-vehicle interactions are identified as conflicts. These two thresholds are Time-toCollision (TTC) and Post-Encroachment Time (PET). S SAM utilizes a default maximum TTC value of 1.5 seconds as suggested by (S ayed, Brown, Navis, 1994) and (Hayward, 1972) and the maximum PET value of 5 seconds as suggested by Hyden (1987).SSAM filters out conflict s that do not fall in the specified rangesThese default values were used in this study.

PAGE 63

n Gettman, Pu, Sayed, and Shelby (2008) developed con flict identification algorithm of SSAM and described as follows: First, SSAM determines the dimensions of the analys is area based on the TRJ file and constructs a zone grid to cover the en tire rectangular analysis area. By dividing the region into zones, the number of vehicle-to-vehicle comparisons necessary to identify potential conflic ts is reduced considerably. Second, SSAM analyzes a single time step of a traje ctory file. For each vehicle in the analysis region, it projects that ve hicleÂ’s expected location as a function of its current speed, if it were to contin ue traveling along its (future) path for up to the duration of the configured timeto-collision (TTC) value. Third, for each vehicle, SSAM calculates the rectan gular perimeter delineating the location and orientation of that ve hicle at its projected future position. Any time a vehicle is added into a zone t hat currently contains one or more other vehicles, SSAM checks for overlap of the new vehicle (rectangle) with each of the other vehicles (rectan gles) in that zone. It is possible that two vehicles may partially occupy the same zone without overlapping. However, two overlapping rectangles in dicate that a future collision is projected for this pair of vehicles, a nd therefore, a potential conflict has been identified. SSAM maintains a list of all c onflicting vehicle pairs (all conflict events) for the current time-step. Each ti me-step, the list is prepopulated with all conflicting vehicle-pairs fro m the prior time-step. If the current vehicle being added to the zone grid overla ps with any other vehicle, that vehicle-pair is added to the conflict list for the current time-step (if not already in the list).

PAGE 64

Four, SSAM continues to perform a more detailed pro cessing of each conflicting vehicle-pair in the list for the curren t time-step as follows: (a) The TTC of the vehicle-pair is updated by itera tively shortening the future projection timeline by a tenth of second and re-pro jecting both vehicles as before over successively short distances until the pair of vehicles no longer overlaps in their projected locations. In this way, a more accurate TTC value is established for this time-step. Instead of the l arge overlap, the vehicles can have just barely come into contact. Note that if th e projection timeline reduces to zero seconds and the vehicles still overlap, the n this is a crash. (b) The minimum TTC (taking the minimum of the curr ent TTC value and that of the prior time-step, if applicable) are calculat ed and updated. Also, the current (actual) positions of both vehicles are rec orded for post-encroachment analysis. (c) If it was found that the vehicle-pair does not overlap over any projection time between zero and maximum TTC, then this vehicl e-pair has made its way into the conflict event list by virtue of being in the list during the prior time-step. In this case, the event remains in the l ist, watching for the one vehicle (the trailing vehicle) to eventually occupy (or encroach on) on a position formerly held by the other vehicle (the le ading vehicle). The time differential between when the leading vehicle occup ied this location and the trailing vehicle arrived is the post-encroachment t ime (PET). If a postencroachment was observed, then the minimum PET is updated, and this conflict event remains in the list, as the post-enc roachment could potentially reduce as the vehicle trajectories progress over ti me.

PAGE 65

(d) If a vehicle-pair in the conflict event list is no longer on an imminent collision course, and it is clear that PET to any p rior positions could not further reduce the minimum PET, or the maximum PET has elap sed, then this vehicle-pair is identified for removal from the con flict event list. Prior to removal, all final surrogate measures are computed, including conflict starting and end points, and conflict angles. After SSAM identifies the conflicts, it classifies them in to four types: rear end, lane change, crossing and unclassified conflicts. T he link and lane information is utilized for classification in the case that the vehicles both occupy the same link at either the start or end of the conflict eve nt. The simulation model (VISSIM) that produced the vehicle trajectory data can generally provide link and lane information for both vehicles. If both ve hicles occupy the same lane at the start and end of the event, then it is class ified as a rear-end event. If either vehicle ends the conflict event in a differe nt lane than it started (while having not changed links), then the event is classi fied as a lane change. If either of the vehicles changes links over the cours e of the event, then the conflict angle determines the classification using the threshold values mentioned next. The event type is classified based on the absolute value of the Conflict Angle. A conflict is classified as unclassified if the con flict angle is unknown, rear-end if ||Conflict Angle|| < 30, crossing if ||Conflict Angle|| > 85, lane change if 30 || conflict angle || 85. Figure 2.4 illustrates the conflict angle.The threshold angles (30 and 85) were determined by e xperimentation.

PAGE 66

r FIGURE 2.4 : ILLUSTRATION OF CONFLICT ANGLE DIAGRAM (SSAM SOFTWARE) 2.11 SAS/STAT Software and the GENMOD Procedure According to the SAS/STAT software user manual (201 1), the software provides comprehensive statistical tools for a wide range of statistical analyses, including analysis of variance, categoric al data analysis, cluster analysis, multiple imputation, multivariate analysi s, nonparametric analysis, power and sample size computations, psychometric an alysis, regression,

PAGE 67

survey data analysis, and survival analysis. A few examples include nonlinear mixed models, generalized linear models, correspond ence analysis, and robust regression. The current release of SAS softw are, SAS 9.3 version, was used for this study. A component of SAS software called GENMOD procedure was used. The GENMOD procedure fits generalized linear models, as defined by Nelder and Wedderburn (1972). The class of generalized linear models is an extension of traditional linear models that allows the mean of a population to depend on a linear predictor through a nonlinear link function and allows the response probability distribution to be any member of an exp onential family of distributions. Many widely used statistical models are generalized linear models. These include classical linear models with normal errors, logistic and probit models for binary data, and log-linear model s for multinomial data. Many other useful statistical models can be formula ted as generalized linear models by the selection of an appropriate link func tion and response probability distribution. The GENMOD procedure fits a generalized linear mode l to the data by maximum likelihood estimation of the parameter vect or. There is, in general, no closed form solution for the maximum likelihood estimates of the parameters. The GENMOD procedure estimates the para meters of the model numerically through an iterative fitting process. T he dispersion parameter is also estimated by maximum likelihood or, optionally by the residual deviance or by PearsonÂ’s chi-square divided by the degrees o f freedom. Covariances, standard errors, and |-values are computed for the estimated parameters based on the maximum likelihood estimators.

PAGE 68

3. Research Approach and Methodology 3.1 Field Validation The field validation effort for the study sites wer e assessed with SSAM, and the results were compared to actual crash histories Also, surrogate safety estimates were compared with traditional volume-bas ed models for crash prediction. The field validation testing was based solely on modeling with the VISSIM simulation. 3.2 Purpose The main purpose of the field validation effort was to compare the predictive safety performance capabilities of the SSAM approac h with actual crash experience of merging and diverging freeway section s. This effort consisted of a series of statistical tests to assess the corr elation between actual crash frequencies at a series of merge/diverge sections a nd the corresponding frequency of conflicts observed in simulation model s of these sections. Traditional volume-based crash prediction models we re used as a basis for comparison. 3.3 Data Assembly and Geometric Definition Merging and diverging influence areas along freeway s at interchanges were selected based on the following guidelines: (1) a s ufficient number of

PAGE 69

interchanges were needed to obtain the range of par ameters necessary to test the methodology, and to establish statistical measure of significance. (2) interchanges were not selected based on their crash performance (e.g., most dangerous interchanges in an area) as this wil l lead to the regression-to-the-mean bias (Hauer, E. 1980). The r egression-to-the-mean bias is the tendency of a randomly high accident fr equency occurring at a location during a specific time period to be follow ed by a smaller accident frequency during a consecutive period of equal dura tion, even if no treatment is applied to the field configuration that generate s the accidents at that location. Appendix A provided a summary of the twenty-one int erchanges used in this study. It shows, for each interchange: the highway names, cross street names, truck percentage, average daily traffic (ADT ) and peak hour volume (PHV) for the mainline, on-ramp and off-ramp moveme nts. Prior to fitting the model all merge/diverge relate d accidents were isolated from the CDOT accident database. This eliminated th e influence of nonmerge/diverge accidents from the analysis. Appendi x C presents the crash data in terms of average yearly crash counts for ea ch merge/diverge locations. The crash counts were derived by filteri ng through all interchange crash records to include only two (or more) vehicle crashes. Thus, singlevehicle crashes, such as run-off-road crashes, fixe d-object crashes, and animal, pedestrian, or bicycle related crashes, wer e excluded. In addition to the crash count, properly defining t he interchange geometry was an important part of the modeling process. Traffic exiting a highway is required to reduce the speed. The speed-change lane that enables the driver

PAGE 70

to reduce the speed between the highway and the tur ning roadway in a safe and comfortable manner is called deceleration lane. On the other hand traffic entering a highway is required to increase the spee d. The speed-change lane that enables the driver to increase the speed betwe en the highway and the turning roadway in a safe and comfortable manner is called acceleration lane.Roess and Ulerio ( 1993) studied the on-ramp and of f-ramp junctions and showed that lane changing and turbulence extend to 1500 feet downstream from the tip of the merge gore and 1500 feet upstre am from the tip of the diverge gore. Moreover, the mainline traffic starts to change lane upstream of the gore at merge. Similarly at diverging location it is common to see some traffic to change lanes downstream of the gore for adjustment. As a result of these, at both merge/diverge locations crashes at a distance of 1500ft on either sides of the gore (total 3000ft) were compar ed with the corresponding conflicts recorded for the same distance (3000ft). Figure 3.1 and Figure 3.2 illustrates the model geometric configuration of me rge and diverge influence areas respectively. Collection of traffic counts fr om CDOT crash database and conflict analysis by SSAM were based on the def inition of these Figures. The model considered only right-hand traffic flow c onditions for both merge/diverge influence areas.

PAGE 71

1500 feet traffic movement number of lanes vary 1500 feet traffic movement number of lanes vary acceleration lane(s) on-ramp deceleration lane(s) off-ramp merge influence area diverge influence area 1500 feet gore 1500 feet gore FIGURE 3.1 ILLUSTRATION OF MERGE INFLUENCE AREA FIGURE 3.2 ILLUSTRATION OF DIVERGE INFLUENCE AREA

PAGE 72

3.4 Simulation Modeling of Interchanges and Modelin g Assumptions Forty two merging and diverging locations were code d in VISSIM simulation modeling tool. The process of modeling of interchan ges in VISSIM was started by tracing an aerial photo of the proper ge ometric configuration by defining the number, width, and length of lanes. Af ter the geometric configuration of the interchange was defined, other modeling parameters such as traffic flow (peak-hour volume), traffic co mposition, percent truck, routing decision, etc., were allocated. Several modeling parameters were defined during the process including: speed profiles, vehicle type, traffic composition, routing decisions and other aspects. The traffic composition was assumed to be composed of passenger cars and trucks. The percentage of trucks was coded at each interchange. The desired speed profiles were assumed to range fr om 50mph to 75mph for mainline traffic and 30mph to 45mph for on-ramp and off-ramp traffic. Carfollowing and lane-changing behavior were set to mo del freeway (free lane selection) traffic flow using the Wiedemann 99 mode l with all default parameters used (VISSIM 5.2 User Manual, 2009). A simulation resolution of 5 time steps per simulation second was used. Among the different types of routing decisions a “static” routing option was use d which defines the vehicles routes from a start point to any of the defined des tinations using a static percentage for each destination. The lane-changing algorithm in VISSIM allows for tw o types of lane changes: necessary lane changing and free lane changing. A f ree lane changing option was chosen. In a free lane change, VISSIM checks fo r the desired speed

PAGE 73

n safety distance of the trailing vehicle on the new lane. This safety distance depends on the vehicle speeds. Following nominal verification, each interchange wa s simulated five times for a period of 3600 simulation seconds (1 hour). Inste ad of running multiple simulation runs with different random seeds, a mult irun option was used. This option lets VISSIM automatically run the requested five successive runs at each location. When run this way, VISSIM automatica lly uses a new random seed number at the starts of each run. The modeled interchanges were tested for realistic and reasonable vehicle behaviors. The TRJ output files from VISSIM run mod el were imported into the SSAM application to identify traffic conflicts and calculate the corresponding surrogate safety measures. The output from SSAM was formatted and analyzed by SAS GENMOD statistical so ftware to produce safety performance functions. The traffic flow inputs to VISSIM were based on pea k hour volumes, as recorded by the CDOT traffic database. The volumes used for simulation are included in Appendix A.1. Figure 3.3 and Figure 3.4 are screen captures of geometric layout and flow of traffic modeled in VIS SIM respectively.

PAGE 74

FIGURE 3.3 SCREEN CAPTURE. VISSIM GEOMETRIC MODEL OF I-25 & E HAMPDEN AVE Traffic Direction

PAGE 75

FIGURE 3.4 SCREEN CAPTURE. VISSIM TRAFFIC MODEL OF I-25 & E HAMPDEN AVE

PAGE 76

r 3.5 Identification and Removal of Outliers Before any statistical inferences were applied, the outliers were identified and removed from the dataset. Outliers are observations or subsets of observations which appear to be inconsistent with t he remainder of the data (Barnett & Lewis, 1978). There are three ways for o utliers to arise in a sample (Barnett and Lewis (1994, pp. 33-34)): (1) data due to a reading error, a recording error, or a calculation error in the data ; (2) a data value that incorrectly included in the data set; (3) a correct ly recorded data value that belongs in the data set. Since these causes of outl iers are possible to occur in the data, due attention were given to identify and remove them. The Standardized Residual Criterion (z-score method ) was used to identify outliers. The standardized residual, the STUDENT st atistic, is the residual expressed in units of standard deviation from the m ean value. Thus, a standardized residual with an absolute value of 3 i s 3 standard deviations above the mean value. Observations with standardize d residual with an absolute value of 3 or greater are potential outlie rs and were removed from the analysis. AMA : where: = It denotes the number of standard deviations a da ta value A is from the mean, A.

PAGE 77

n 3.6 Methodology The methodology used for the field validation task was based on relating surrogate safety measures that SSAM derived from si mulation models with actual crash occurrence observed on the field at me rging and diverging locations. The field validation effort involve the analysis of 21 interchanges (42 merging and 42 diverging) locations, modeled wi th the VISSIM simulation. Gettman, Pu, Sayed, and Shelby (2008) provided five statistical field validation tests for the analysis of intersections. These tests were modified and used in this study for merging and diverging lo cations: 1. Field Validation Test 1: Merging and Diverging S ection Ranking by Total Incidents 2. Field Validation Test 2: Merging and Diverging S ection Ranking by Specific Incident Types 3. Field Validation Test 3: Crash and Conflict Pred iction Regression Model Paired Comparison 4. Field Validation Test 4: Crash and Conflict Pred iction Regression Model Comparative Analysis of Total Incidents 5. Field Validation Test 5: Crash and Conflict Pred iction Regression Model Comparative Analysis of Specific Incident Types 3.6.1 Field Validation Test 1: Merging and Divergin g Location Ranking by Total Incidents In this field validation test the ranking of mergin g and diverging locations from SSAM according to predicted total conflicts was com pared to the same

PAGE 78

n locations using actual crash frequency. Three steps were performed in this test: 3.6.1.1 Conflict Ranking In this step, SSAM was used to predict the expected total number of conflicts at each merging and diverging locations. Each locat ion was simulated for five replications, each lasting 3600 simulation seconds (1 hour). Thus, the sum of all conflicts recorded over all five replications w as divided by 5 to determine the average hourly conflict frequency in terms of c onflicts per hour. The merging and diverging sections were then ranked bas ed on their average hourly conflict frequency in descending order. 3.6.1.2 Crash Ranking Crash ranking in this step was performed in such a way that the average yearly crash frequency for each merging and divergi ng locations was first determined by dividing the total number of crashes over the observation period by the number of years in the observation pe riod. The 3 years observation period were used for all study location s. Merging and diverging locations were then ranked based on their average y early crash frequency in descending order.

PAGE 79

n 3.6.1.3 Ranking Comparison The ranking derived from SSAM’s conflict prediction (section 3.6.1.1) was compared to the rankings based on average yearly cr ash frequency (section 3.6.1.2). The Spearman rank correlation coefficient was used to determine the level of agreement between the two rankings. Th e Spearman rank correlation coefficient is often used as a nonparam etric alternative to a traditional coefficient of correlation and can be a pplied under general conditions. The Spearman rank correlation coefficie nt ‘ is calculated as shown in Equation 3.2. A score of 1.0 represents pe rfect correlation and a score of 0 indicates no correlation. An advantage o f using ‘ is that when testing for correlation between two sets of data, i t is not necessary to make assumptions about the nature of the populations sam pled. ‘M S Ž – M : Where: –the difference between two rankings for pair the number of paired ranked. For hypothesis test, the Spearman rank correlation coefficient‘ was compared with the critical value‘T€. For : ‘T€ can be read from the critical values table

PAGE 80

n For :use: ‘T€ M :: Where: corresponds to the level of significance. For example, if ”N, then bS.If the absolute value of the test statistics ‘ exceeds the positive critical value, then reject ‘ and conclude that there is a correlation. 3.6.2 Field Validation Test 2: Merging and Divergin g Location Ranking by Specific Incident Types Comparative ranking procedures for Rear-end, crossi ng and lane-changing crash/conflict types were repeated in Field Validat ion Test 2 the same as Field Validation Test 1. These three conflict types were classified and counted by SSAM for each type. The average hourly conflict frequency for each conflict type was computed as in Field Validation T est 1, and these results were used to rank the merging and diverging section s for each conflict type. To provide type-specific crash frequencies, all cra sh reports from CDOT crash database were reviewed to determine whether i t was a rear-end incident, crossing incident, or lane-changing incid ent. If the classification based on the crash type was not obvious, engineerin g judgment was applied to determine the most representative type. Average yearly crash frequencies was tabulated for each incident type, rank ordered, and compared using the

PAGE 81

n SpearmanÂ’s rank correlation test, as in Field Valid ation Test 1. The results of this step demonstrated the capability of SSAM to ac curately identify and rank merging and diverging sections 3.6.3 Field Validation Test 3: Crash and Conflict P rediction Regression Model Paired Comparison. In this field validation test, the correlation betw een conflicts obtained from SSAM and crashes were developed. Regression equatio ns were established to estimate the average yearly crash frequencies at merging and diverging locations as a function of the average hourly confl ict frequencies. The R2 Pearson chi-square and scaled deviance goodness-offit testing were used to determine the strength of the relationship between conflicts and crashes. Regression equations to predict crashes based on vo lume were developed and compared with the conflict-based crash predicti on model to study the relative capabilities of surrogate safety assessmen t. Generalized linear modeling GLM technique (Hauer, 1997; Sayed and Rodr iguez, 1999) were used to calculate the expected crash frequency at e ach merging and diverging location, using the GENMOD procedure in t he SAS 9.3 statistical software package.

PAGE 82

n 3.6.4 Field Validation Test 4: Crash and Conflict P rediction Regression Model Comparative Analysis of Total Incide nts This Field Validation Test consists of four steps a nd the details of the test are as follows: Step A: Conflict prediction models for merging and diverging locations were developed using standard GLM procedures relating co nflicts calculated by SSAM as a function of hourly traffic volume. Step B: Crash prediction models for merging and div erging locations were developed using standard GLM procedures relating ac tual crash data obtained from CDOT accident database as a function of daily traffic volume. Step C: The two prediction models were compared to determine whether or not the conflict prediction models can predict risk in a manner similar to the crash prediction model for merging and diverging lo cations with the same characteristics. The comparison consists of identif ication and ranking of crash/conflict prone locations. A crash/conflict pr one location is defined as any location that exhibits a significantly higher n umber of crashes/conflicts as compared to a specific normal value. The normal val ue was provided by the volume-based crash/conflict prediction models. As s tated previously, the Empirical Bayes (EB) technique improves the locatio n-specific prediction and thus was used to identify hazardous locations. The EB refinement method identifies problem sites according to the following four-step process: 1. Estimate the predicted number of crashes/conflic ts and its variance using the crash/conflict prediction model. Thi s prediction can be

PAGE 83

nn assumed to follow a gamma distribution (the pri or distribution) with parameters ” and x, where: x  w• r  –”x r :I Where: w• the variance of the predicted crashes/conflicts.  predicted accident frequency r shape parameter 2. Determine the appropriate point of comparison ba sed on the mean and variance values obtained in step 1. Usually the 50th percentile VŠ) or the mean is used as a point of comparison. VŠ)is calculated such that Equation 3.5 is satisfied. n r  o™7B7 n o Ÿ r )–™N :N 3. Calculate the EB safety estimate and the varianc e by rearranging Equations 2.23 and 2.24 respectively as shown i n Equation 3.6 and Equation 3.7.

PAGE 84

n ‘€’“TD“‘T€T“X F Z *X F Z zKy{ :S where: zKy{ observed accident frequency. w•‘€’“TD“‘T€T“  r* r *  r* zKy{ :\ This is also a gamma distribution (the posterior di stribution) with parameters ” and x defined as follows: x  w•  r  *–”xr*zKy{:^ Then, the probability density function of th e posterior distribution is given by the following: @ ™ n r  *o F=T ™F=T7B7 n F o Ÿ r*zKy{ :b 4. Identify the location as crash/conflict-prone if there is significant probability that the location’s safety estimate exceeds the VŠ) value (or the mean).

PAGE 85

n The location is prone if: ¨M X FZ G ŸG„“n „o F=T –™)§: Where: § is the desired confidence level (usually selected at 0.95). The merging and diverging locations were then ranke d in terms of priority for treatment once crash/conflict-prone sites were iden tified. Ranking problem sites enables the road authority to establish an ef fective road safety program, ensuring the efficient use of the limited funding a vailable for road safety. Two techniques were suggested (Sayed and Rodriguez, 199 9) that reflect different priority objectives for a road authority. The first ranking criterion is to calculate the ratio between the EB estimate and the predicted frequency as obtained from the GLM model (a risk-minimization objective). The ratio represents the level of deviation that the merge/diverge is away f rom a “normal” safety performance value, with the higher ratio representi ng a more hazardous location. The second criterion, the cost effectiven ess objective, is a “potential for improvement” (PFI) criterion, which is calculat ed as the difference between the observed crash/conflict frequency and t he volume-based estimate of “normal” crash/conflict frequency. Step D: Two comparisons were undertaken. The first compared the locations identified as crash-prone to merging and diverging locations identified as conflict-prone. The second compared both the “risk ratio” and PFI merging

PAGE 86

nr and diverging location rankings obtained using cras h data to the corresponding rankings obtained using conflict data 3.6.5 Field Validation Test 5: Crash and Conflict P rediction Regression Model Comparative Analysis of Specific Type Incidents Field Validation Test 5 repeated the same comparati ve analysis as Field Validation Test 4 for the rear end, crossing and la ne change crash/conflict types. 3.7 Goodness of Fit Measure Three statistical measures were used to assess the goodness-of-fit of the regression models to the actual crash data to the m erging and diverging locations. The first measure was the Pearson chi-squared and c omputed by the following model: 2B•¢Kv/ P
PAGE 87

< the actual crash frequency of merging or diverging location the number of merging or diverging location The second statistical measure used was the scaled deviance. This is the likelihood ratio test statistic measuring twice the difference between the loglikelihood of the data under the developed model an d its log-likelihood under the full or saturated model. The scaled deviance is computed by the following model: jk/c
PAGE 88

Where: ”the overdispersion parameter estimated in the model ”€" the overdispersion parameter estimated in the nega tive binomial model, the model with only a constant t erm and an overdispersion parameter.

PAGE 89

4. Test Results and Discussion This section presents the results of the field vali dation testing effort. The experimental procedure, prior to statistical testin g, is summarized as follows. A set of twenty-one interchanges, selected from fie ld sites in Colorado, North America, and were modeled in VISSIM. Each merge/di verge model was simulated for five replications for 1 hour of simul ated time. The corresponding five output files (i.e., TRJ files) from VISSIM wer e then imported into SSAM for identification of conflicts and computation of the surrogate measures of safety for each conflict event. SSAM was configured to use conflict identification threshold values. The maximum time-t o-collision (TTC) and maximum post-encroachment time (PET) values used we re 1.5 seconds and 5.0 seconds, respectively. The results of the SSAM analysis consists of the number of total conflicts and the number of conflic ts of each type of vehiclevehicle interaction: crossing, rear end, and lane c hanging. Before any statistical inferences were applied, the outliers w ere identified and removed from the dataset. Figure 4.1 shows five files added to the case files section. These files were obtained from a model simulated for five replicatio ns. Note that the files selected have a .trj extension. The default maximum TTC and maximum PET values were not changed. Also the threshold angles (30 and 85 ) which were determined by experimentation were not changed

PAGE 90

FIGURE 4.1 SCREEN CAPTURE. SSAM SCREEN—CONFIGURATION TAB SSAM contains a built-in feature to visualize confl icts. The conflicts can be visualized within a Map tab or imported to SSAM as VISSIM .inp file. Typically, the Map file is the graphic file of the simulated network for the case document. The VISSIM .inp file is the VISSIM networ k file that has been simulated and has generated the vehicle trajectory file that was used for

PAGE 91

SSAM analysis.Figure 4.2 shows the display of conflicts in the Ma p view section. The average hourly conflict counts for eac h merge/diverge locations is provided in Appendix E. FIGURE 4.2 SCREEN CAPTURE. SSAM SCREEN – MAP TAB VIEW AT I-25 & E HAMPDEN AVE Diverging conflict Merging conflict

PAGE 92

4.1 Field Validation Test 1: Merging and Diverging Location Ranking by Total Incidents Validation Test 1 is a comparison of the ranking of merge/diverge locations based on average hourly total conflict frequency ve rsus the ranking of merge/diverge locations based on average yearly tot al crash frequency. The Spearman rank correlation coefficient (s) values of ranking based on total conflict comparison were 0.486 and 0.551 for mergin g and diverging movements respectively, these values are significan t at 95 percent level of confidence. As a basis of comparison, the merge/diverge locatio ns were also ranked in terms of total ADT values, and these were also comp ared with the ranking based on average yearly total crash frequency. The Spearman rank correlation coefficient (s) values of this ranking comparison were 0.419 (significant at a 95 percent level of confidence) a nd 0.309 ( not significant at a 95 percent level of confidence) for merging and div erging movements respectively, these values are shown in Table 4.2. In general, the Spearman rank correlation tests sho wed that a higher correlation was found for both merge and diverge mo vements with merge/diverge locations rankings based on conflicts than ADT.

PAGE 93

n 4.2 Field Validation Test 2: Merging and Diverging Location Ranking by Specific Incident Types Validation Test 2 repeats the same comparative rank ing procedures as for validation test 1 for the subsets of crash/conflict types: rear end and lane changing. There were no crossing conflicts recorded to perform the ranking comparison for crossing type incidents. Table 4.1 d isplays the distribution of conflicts and crashes by incident type. There were 1 percent merging and 4 percent diverging lane changing conflicts were reco rded, while the remaining portion were rear end conflicts. However the propor tion of crashes of the lane changing and rear end maneuvers at merge and diverg e, was 26 percent and 23 percent respectively. There were significant differences between conflict distributions by type and actual crash distributions by type. The ratios of c onflicts-per-hour to crashesper-year for total incidents at merge and diverge w ere 117 and 137 respectively. The ratios of conflicts-per-hour to c rashes-per-year for rear-end and lane-change incidents at merge were 156 and 5 r espectively. However the ratios of conflicts-per-hour to crashes-per-yea r for rear-end and lanechange incidents at diverge are 169 and 26 respecti vely. It seems plausible that conflicts-to-crashes ratios are higher for les s dangerous incidents. That is, accepting that rear-ends are generally less sev ere than lane-change, there is an abundance of these “lower risk” conflicts. In all the simulation scenarios, rear-end conflict events made up the bulk of the to tal conflicts. It was also observed that the total number of diver ging conflicts were 42% higher than that of the merging conflicts. Similarl y, the total number of diverging crashes are 22% higher than that of mergi ng crashes This was

PAGE 94

consistent with the study by Cillo (1967) and Lundy (1966) which showed that the accident rates of off-ramps are higher than the accident rates of onramps. Figure 4.3 illustrates an example of a conflict, w here some vehicles are angling across lanes and have abruptly cut in front of other vehicles that must decelerate to avoid collision. FIGURE 4.3 SCREEN CAPTURE. VISSIM CONFLICT SCENARIO RESULTING FROM LANE CHANGE MANEUVE R conflict scenario

PAGE 95

A higher proportion of rear-end conflict obtained f rom SSAM was consistent with the study reported on crash prediction models on urban signalized intersections. Guttman, Pu, Sayed, and Shelby (2008 ) modeled signalized intersections in VISSIM and simulated under AM peak hour traffic conditions. The conflict analysis was done using SSAM and found that the proportion of crossing, rear-end, and lane-change conflicts were 1.7 percent, 91.0 percent, and 7.3 percent respectively. Moving on to the results of the ranking tests and l ooking into Table 4.2, there is a significant correlation between conflicts and crashes when considered by conflict type. The Spearman rank correlation coeffi cient (s) value at merge for rear end incident ranking comparison was 0.403, which is significant at a 95 percent level of confidence. Whereas, the Spearm an rank correlation coefficient (s) value at diverge for rear end incident ranking co mparison was 0.456, which is also significant at a 95 percent le vel of confidence. Moreover, the Spearman rank correlation coefficient (s) value at merge and diverge for the lane changing incident ranking comparison were 0.540 (significant at a 95 percent level of confidence) and 0.550 (significant at a 95 percent level of confidence) respectively. For comparison, the merge/diverge locations were ra nked on the basis of total ADT values, and this was also compared with t he rankings based on rear-end and lane-change crashes. Rear-end ranking comparisons at merge was not significant at 95 percent level of confiden ce. Moreover, the Spearman rank correlation coefficient (s) at diverge for rear-end incident ranking comparison was not significant at a 95 percent leve l of confidence. However, the Spearman rank correlation coefficient (s) at both merge and diverge for

PAGE 96

r lane-changing ranking comparison were significant a t a 95 percent level of confidence. In general, rank correlation testing showed a signi ficant correlation between rear-end conflicts and rear-end crashes and between lane-change conflicts and lane-change crashes. Also, the rank correlation tests have shown that stronger correlation was observed based on conflict than ADT. Table 4.1 Distribution of crashes and conflicts by incident type Incident Type Merge Diverge Rear End Lane Change All Type Rear End Lane Change All Type Average Hourly Conflicts 36,572 370 36,942 50,358 2,219 52,577 Percentage of Conflicts by Type 99% 1% 100% 96% 4% 100% Average Yearly Crashes by Type 234 82 315 298 87 385 Percentage of Crashes by Type 74% 26% 100% 77% 23% 100% Conflict-to-crash ratio 156 5 117 169 26 137

PAGE 97

Table 4.2 Correlation coefficients by incident type Ranking based on Incident Type Merge Diverge Rear End Lane Change All Type Rear End Lane Change All Type Crashes and Conflicts 0.403 0.540 0.486 0.456 0.550 0.551 Crashes and ADT 0.318 0.533 0.419 0.217 0.551 0.309 4.3 Field Validation Test 3: Crash and Conflict Pre diction Regression Model Paired Comparison Validation Test 3 assessed the correlation between conflicts and crashes by using regression to construct a conflicts-based mod el to predict crash frequencies at merging and diverging locations. Add itionally, the capabilities of conflict-based crash-prediction were compared to a traditional volumebased crash-prediction model. A standard generalized linear modeling (GLM) approa ch was used to establish a model of the expected crashes as a func tion of mainline, merging, and diverging ADT. Crashes in these models are expressed in terms of average yearly crash frequency, as a function of th e model variables which are the ADT volumes of the mainline, merging, and d iverging (in vehicles per

PAGE 98

day) or ADTmainline, ADTmerge, and ADTdiverge, respectively. The estimates of parameters for these models are shown in Table 4.3 and Table 4.4. However, conflicts are expressed in terms of average hourly conflicts (conflicts per hour). Table 4.3 and Table 4.4 show the estimates of the p arameters of the total crash model at merging and diverging locations resp ectively. Furthermore, the table shows that both the Pearson chi-squared a nd the scaled deviance values were not significant at the 90percent conf idence level, indicating a good fit. Moreover, these models were compared usin g the R-squared goodness-of-fit test (Miao, 1996), and the results conform to those of the Pearson chi-squared and the scaled deviance. Moreo ver, graphs corresponding to Model (1) and Model (2) were provi ded in Figure 4.4 and Figure 4.5 respectively.

PAGE 99

Table 4.3 Prediction model of total crashes as a function of mainline ADT and merging ADT Total Crash/yr merge= 2.12E-04 x ADTmainline 0.773 x ADTmerge 0.209 Degree of R-Squared Scaled Pearson 2 0.1,30 Shape Freedom R 2 Deviance 2 Parameter 30 0.54 32.24 33.61 40.26 5.08 Variable Coefficient Constant 2.12E-04 ADT mainline 0.773 ADT merge 0.209 Table 4.4 Prediction model of total crashes as a function of mainline ADT and diverging ADT Total Crash/yr diverge = 0.061 x ADTmainline 0.058 x ADTdiverge 0.478 Degrees of R-Squared Scaled Pearson 2 0.1,31 Shape Freedom R 2 Deviance 2 Parameter 31 0.82 34.72 34.39 41.42 30.21 Variable Coefficient Constant 0.061 ADT mainline 0.058 ADT diverge 0.478 Model ( 2 ) Model ( 1 )

PAGE 100

FIGURE 4.4 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT AND MERGING ADT Total C rash per Year Mainline ADT Merging ADT n n r n r nnrrn

PAGE 101

n n r n n nrrnrn nn FIGURE 4.5 TOTAL CRASH AS A FUNCTION OF MAINLINE ADT AND DIVERGING ADT Total C rash per Year Mainline ADT Diverging ADT

PAGE 102

However, another regression technique was then empl oyed to relate the actual crashes frequency to the conflict frequency predicted by SSAM. It was assumed that both real life crashes and real-life c onflicts were discrete random events. Therefore, the validation test techn ique assumed that both crashes and simulated conflicts follow a negative b inomial distribution. Table 4.5 and Table 4.6 show the estimate of the paramete rs of these regression equations. These parameter estimates were obtained using a SAS GENMOD statistical tool that was developed to iteratively update the likelihood equations. Also, graphs corresponding to Model (3) and Model (4) were provided in Figure 4.6 and Figure 4.7 respectively. The scaled deviance and the Pearson chi-squared val ues for both models were not significant at the 90-percent confidence l evel, indicating good fit of the regression equation. Moreover, the result of th e R-squared goodness-offit test conforms to those of the Pearson chi-squar ed and the scaled deviance and found to have values of 0.57 and 0.90 for mergi ng and diverging maneuvers respectively.

PAGE 103

n Table 4.5 Prediction model of total crashes as a function of conflicts at merge Total Crash/yr merge = 1.072 x Total Conflicts merge 0.373 Degree of R-Squared Scaled Pearson 2 0.1,31 Shape Freedom R 2 Deviance 2 Parameter 31 0.57 31.91 31.00 41.42 5.40 Variable Coefficient Constant 1.072 Conflicts merge 0.373 Table 4.6 Prediction model of total crashes as a function of conflicts at diverge Total Crash/yr diverge = 2.617 x Total Conflicts diverge 0.204 Degree of R-Squared Scaled Pearson 2 0.1,32 Shape Freedom R 2 Deviance 2 Parameter 32 0.90 35.33 36.19 42.58 52.36 Variable Coefficient Constant 2.617 Conflicts diverge 0.204 Model ( 3 ) Model ( 4 )

PAGE 104

In the previous studies with similar traffic facili ties, traditional multiple regression models based on the negative binomial di stribution were performed to determine the relationships between ac cidents and the geometric design and traffic volume characteristics of interchanges. These models explained between 10 percent and 42 percent of the variability in the accident data. For example, in FHWA-RD-97-106, Stat istical Models of Accidents on Interchange Ramps and Speed Change Lan es —Report linear regression models were applied to fit 3 years of cr ash data for a set of the 551 ramps, 276 deceleration lanes, and 192 accelera tion lanes, yielding an R-squared value of 0.38 for the total crash frequen cy The conflicts-based model and volume-based model we re compared using the R-squared goodness-of-fit test. For crash predi ction model based on volumes, the R -squared values were recorded to be 0.54 and 0.82 f or merge and diverge approach respectively. Similarly, for c rash prediction model based on simulated conflict, the R -squared values were recorded to be 0.57 and 0.90 for merge and diverge approach respectivel y. Thus, the conflict based model exhibited a better correlation than the volume-based model. However, both the volume-based model and the confli ct based model had a better correlation than the traditional multiple re gression model.

PAGE 105

n r r n FIGURE 4.6 TOTAL CRASH AS A FUNCTION OF TOTAL CONFLICT AT MERGE Total Crashes per Y ear at Merge Total Conflicts per H our at Merge

PAGE 106

r FIGURE 4.7 TOTAL CRASH AS A FUNCTION OF TOTAL CONFLICT AT DIVERGE n Total Conflicts per H our at Diverge Total Crashes per Y ear at Diverge

PAGE 107

r Figures 4.4 through Figure 4.7 helped estimate cras hes as a function of ADT and conflict frequency at both merge and diverge se ction of freeways. In the ADT approach (Model 1 and Model 2), merge an d diverge locations that have the same ADT would have the same crash fr equencies, which is not always true. On the contrary, looking at crash as a function of conflict (Model 3 and Model 4), it explained the operational differ ences and vehicle dynamics of locations. The conflict analysis considered the dynamics of simulate traffic systems on a vehicle-by-vehicle basis by updating p osition, speed, acceleration, lane position, and other variables on time steps, such as on a seconds basis, as the vehicles interact with other vehicles and roadway geometrics. In the evaluation of surrogate safety m easures, microscopic simulations model the key driver behaviors (car fol lowing, gap acceptance, lane changing) that affect the occurrence of traffi c crashes. At merging locations simulation models allow vehicles to merge safely by accelerating in the adjacent lane, which occurs in the real world. At diverging location simulation models allow vehicles to diverge safely by deceleration in the adjacent lane, which occurs in the real world as we ll. In general, conflict models (SSAM) can be used to a ssess the relative safety of a pairs of merge/diverge design alternatives. Un der equivalent traffic conditions (e.g., traffic volumes and truck percent age), for both merge/diverge design alternatives, SSAM could exhibit statistical ly significant differences in the total number of conflicts, the number of confli cts by type (lane-change, or rear-end incidents). This helps to conclude that on e design exhibited a higher frequency of conflicts, than the alternative design This type of assessment helps the decision-making process about which desig n is the safer of the two.

PAGE 108

r Therefore, models that relate total crash and total conflict (Model 3 and Model 4) provided a more realistic output to represent th e real world crash frequencies. 4.3.1 Illustration After identifying the locations that exhibited abno rmal crash/conflict frequencies, they were examined for a possible expl anation as to why that happened. To illustrate, a corridor on I-25 from E. Dry Creek Rd to E. Hampden Ave was taken. After recording crash frequencies at the sel ected data points, high number of merging and diverging crashes were observ ed both in the north and south bound directions, see Appendix D. Looking at each crash, congestion builds throughout the rush hour even tho ugh the volumes are not that much different. As congestion builds so does the chance for something to go wrong which in this case could be the crash c ounts. Entering and exiting volumes, especially with I-225 (a high volu me interstate freeway) located at the north end can have an effect as larg e volume introduced from/to I-225. Looking at the same corridor, the number of conflic ts recorded by SSAM at merge both at E. Dry Creek Rd and E. Belleview Ave were high. This happened because the acceleration lengths at these locations were very short to allow traffic to change lane and merge wit h the through traffic. According to AASHTO (2004), acceleration lane shoul d have sufficient length to permit adjustments in speeds of both through and entering vehicles so that

PAGE 109

r the driver of the entering vehicle can position him self in a safe and comfortable manner in a through traffic stream and maneuver into it before reaching the end of the acceleration lane. In driving north the same corridor, but studying th e recorded conflict at diverging locations, E. Dry Creek Rd exhibited the highest number of conflict whereas Belleview exhibited the lowest number of co nflicts. The main reason for this was the discrepancy of the composition of traffic at these locations. The percentage of trucks at E. Dry Creek Rd and E. Belleview Ave were 6.6% and 5.1% respectively, while geometric and other tr affic data remain the same. Therefore, the high proportion of truck contr ibuted to the increased number of conflicts. Moreover, there was evidence to suggest that the o ccurrence of conflicts and crashes may be somewhat distinct while still signif icantly related. For example, for the same corridor defined above, at E. Dry Creek Rd in the north bound direction the number of recorded crashes were the lowest while the corresponding number of conflicts recorded by SSAM were the highest. This is true both at the merge and diverge locations.

PAGE 110

r 4.4 Field Validation Test 4: Crash and Conflict Pre diction Regression Model Comparative Analysis for Total Incidents 4.4.1 Development of Conflict Prediction Regressi on Models for Total Incident The first step of this validation test included dev eloping a volume-based conflict prediction model and a volume-based crashprediction model. The volume-based crash prediction model was previously developed in Validation Test 3, Model (1) and Model (2). The parameters of those models are shown in Table 4.3 and Table 4.4 respectively. A volume-based conflict-prediction model was develo ped using the same procedure but relating conflicts to traffic volumes on the mainline, merge and diverge areas. Note that the conflicts identified b y SSAM were based on the peak-hour volumes used for simulation. The variable s used in the conflictprediction model were: PHVmainline vehicle per hour (VPH) on the mainline approach, PHVmerge, vehicle per hour (VPH) on the merge location, and PHVdiverge, vehicle per hour (VPH) on the diverge location. T able 4.7 (Model 5) and Table 4.8 (Model 6) shows the estimates of the parameters for total conflict prediction model at merge and diverge loca tion respectively. The R squared values were assessed and recorded to be 0.2 2 and 0.29 for merging and diverging locations respectively.

PAGE 111

r Table 4.7 Prediction model of conflicts as a function of mai nline PHV and merging PHV Total Conflict/hr merge = 0.071 x PHV mainline 0.659 x PHV merge 0.394 Degrees of R-Squared Scaled Pearson 2 0.1,30 Shape Freedom R 2 Deviance 2 Parameter 30 0.22 35.71 46.90 40.26 1.97 Variable Coefficient Constant 0.071 PHV mainline 0.659 PHV merge 0.394 Table 4.8 Prediction model of conflicts as a function of mai nline PHV and diverging PHV Total Conflict/hr diverge = 1.51E-05 x PHV mainline 1.264 x PHV merge 0.965 Degrees of R-Squared Scaled Pearson 2 0.1,29 Shape Freedom R 2 Deviance 2 Parameter 29 0.29 34.99 53.51 39.09 1.70 Variable Coefficient Constant 1.51E-05 VPH mainline 1.264 VPH merge 0.965 Model ( 5 ) Model ( 6 )

PAGE 112

r 4.4.2 Identification of Accident Prone Locations ( APL) The second step of this validation test was the ide ntification of crash-prone locations and conflict-prone locations. This test d etermined whether the conflict prediction equation can predict risk in a manner similar to the crash prediction equation for interchanges with the same characteristics, by identifying accident prone locations (APL). A cras h/conflict prone location is defined as any location that exhibits a significant ly higher number of crashes/conflicts as compared to a specific normal value, which in this test is provided by the volume-based crash/conflict predict ion models (Sayed and Rodriguez, 1999) Empirical Bayes (EB) method combines the informatio n contained in crash counts of a specific highway segment with an estima te of its safety performance derived from safety performance functio n obtained from GLM models to increase the precision of estimation. Als o, the EB method corrects for the regression-to-mean bias. Comparison between crash and conflict prone locatio ns at merge and diverge locations are shown in Tables 4.9 and 4.10 respecti vely and a summary was displayed in Table 4.11. Accordingly fewer location s were identified as accident prone by regression Models (1) and (2) bas ed on actual crash than those identified by the regression Model (5) and (6 ) based on simulated conflicts records. Table 4.11 also showed that 5 APL and 2 APLs were c ommon between the total crash/conflict models at merging and divergin g location respectively. This indicates a poor agreement between the total c rash and total conflict

PAGE 113

rn models in identifying incident prone locations at b oth merging and diverging location. These 5APL merging locations can thus be considered by decisionmakers or safety analysts as candidate for remedial treatments before crashes occur. Table 4.12 shows the average total c rashes and total conflicts based on incident proneness of locations. 4.4.3 Illustration Consider a location where traffic merging from E. H ampden Ave onto I-25 in the NB direction. This location has an observed tot al crash rate of 13.0 crashes/ year, applying the model of the safety per formance function developed from GLM, the expected crash value is 12. 4 crashes/year with a variance of 15.0 (crashes/year)2. The P50 value for the Gamma Distribution was calculated as 8.2 crashes/year. The EB safety a nd its variance are 12.8 crashes/year and 6.5 (crashes/year)2 respectively. It was determined that the probability of having crashes less than P50 is 2.39%. This means that there is a significant probability (97.61%) of exceeding the P50 value and this location can be considered as accident-prone. Also consider a location where traffic diverging fr om I-25 on the SB direction to S. University Blvd. This location has a simulate d total conflict rate of 687 conflicts/hour, applying the model of the safety pe rformance function developed from GLM, the expected conflict value was 493 conflicts/hour with a variance of 404,040 (conflicts/hour)2. The P50 value for the Gamma Distribution was calculated as 673 conflicts/hour. The EB safety and its variance were 686 conflicts/hour and 684 (conflicts /hour)2 respectively. It was determined that the probability of having crashes l ess than P50 is 31.3%. This means that this location will not be considered as accident-prone, so the

PAGE 114

r engineering improvements may not be cost effective to reduce the number of accidents related to diverging maneuvers. 4.4.4 Ranking Locations Once crash/conflict-prone sites were identified, th e next step of this test was to rank the locations in terms of priority for trea tment. Two techniques were used: (1) the ratio between the EB estimate and the predicted frequency obtained from the GLM regression equations (a riskminimization objective) and, (2) the PFI (a cost-effectiveness objective). Table 4.9 and Table 4.10 showed the Spearman rank-correlations for both the Ratio and PFI rankings. At merge, the Spearman rank correlation coefficient (s) values of the Ratio and PFI ranking comparisons were 0.118 and 0.045, respectively, indicating an insignificant correlation between locations with “excessive crashes” and locations with “excessive conflicts,” relative to values estimated by volumebased crashand conflict-prediction models respect ively. At diverge, the Spearman rank correlation coefficie nt (s) values of the Ratio and PFI ranking comparisons were -0.196 and -0.110 respectively, indicating an insignificant correlation between loc ations with “excessive crashes” and locations with “excessive conflicts,” relative to values estimated by volume-based crashand conflict-prediction mode ls respectively.

PAGE 115

r ID Actual Total Crash Simulated Total Conflict Model Estimate (E(X)) EB Estimate Crash Prone Location Conflict Prone Location Crash Conflict Crash Conflict 1 4.7 70 5.7 206 5.1 71 no no 1a 3.7 117 5.3 182 4.5 118 no no 2 7.3 111 6.7 271 7.0 112 no no 2a 5.0 158 6.5 266 5.7 159 no no 3 3.3 53 10.4 416 5.6 55 no no 4 5.3 721 11.6 534 7.2 720 no yes 4a 10.7 663 11.4 517 10.9 662 no yes 5 19.7 693 11.7 512 17.3 692 yes yes 5a 7.0 365 11.4 456 8.4 365 no yes 6 17.7 353 11.9 546 15.9 354 yes yes 6a 5.0 268 10.6 441 6.8 269 no no 7 21.0 719 12.1 456 18.4 718 yes yes 7a 6.7 702 12.0 471 8.2 701 no yes 8 13.0 366 12.4 544 12.8 367 yes yes 8a 6.7 331 12.0 530 8.3 332 no yes 9 7.0 113 8.7 316 7.6 114 no no 9a 2.0 207 9.9 368 4.7 208 no no 10a 7.7 253 10.0 367 8.5 254 no no 11 11.0 445 10.4 387 10.8 445 no yes 12 16.0 292 9.8 349 13.9 292 yes no 12a 22.0 376 9.1 302 17.4 376 yes yes 13 5.7 21 5.9 205 5.8 23 no no 14a 4.7 156 7.1 260 5.7 157 no no 15 10.0 256 5.6 190 7.9 255 no no 15a 2.3 143 4.5 128 3.5 143 no no 16a 7.0 1240 7.0 259 7.0 1233 no yes 18 2.7 746 7.3 266 4.6 742 no yes 18a 16.7 312 7.2 262 12.8 312 no no 19 3.0 54 7.1 255 4.7 56 no no 19a 14.3 240 7.2 269 11.4 240 no no 20 8.7 428 8.0 327 8.4 427 no yes 20a 5.3 90 6.7 227 5.9 91 no no 21 4.3 135 4.8 157 4.6 135 no no Spearman rank correlation based on Ratio ranking me thod = 0.118 Spearman rank correlation based on PFI ranking meth od = 0.045 Table 4.9 Comparison between total crash and total conflict prone locations at merge.

PAGE 116

rr ID Actual Total Crash Simulated Total Conflict Model Estimate (E(X)) EB Estimate Crash Prone Location Conflict Prone Location Crash Conflict Crash Conflict 1 7.0 127 8.2 245 7.9 128 no no 1a 4.0 75 7.1 181 6.5 76 no no 2 5.0 194 9.3 455 8.3 195 no no 2a 7.7 251 8.8 433 8.5 252 no no 3a 7.3 28 10.7 929 9.8 30 no no 4 7.0 2291 10.0 1799 9.2 2291 no yes 4a 10.0 2557 9.6 1659 9.7 2556 no yes 5 10.7 1279 11.9 1751 11.6 1279 no yes 5a 13.3 870 11.2 1317 11.8 871 no yes 6 12.7 1593 11.5 2012 11.8 1593 no yes 6a 7.0 739 8.9 1197 8.4 740 non yes 7 15.7 1253 10.9 1276 12.2 1253 yes yes 7a 8.7 1678 10.7 1375 10.2 1678 no yes 8 14.0 1624 14.5 2054 14.3 1624 yes yes 8a 8.3 1287 13.6 1927 12.0 1288 no yes 9 8.0 336 6.9 548 7.1 337 no no 9a 9.0 435 9.2 801 9.1 436 no no 10 15.3 1278 10.6 934 11.8 1277 no yes 10a 5.7 844 10.1 806 9.0 844 no yes 11 12.0 2118 10.8 912 11.1 2116 no yes 11a 13.7 849 9.9 476 10.9 848 no yes 12a 7.3 687 7.7 493 7.6 686 no no 13a 13.7 110 7.1 223 8.4 111 no no 14b 3.3 177 9.1 400 7.8 178 no no 15 9.3 119 7.1 195 7.5 120 no no 15a 3.3 75 4.4 75 4.3 75 no no 16a 16.0 2054 9.3 399 10.8 2047 no yes 18 4.3 186 9.1 416 8.0 187 no no 18a 12.0 749 8.9 399 9.6 748 no yes 19 4.3 236 8.8 379 7.8 237 no no 19a 15.3 1630 9.4 430 10.8 1625 no yes 20 15.3 225 11.9 695 12.9 226 yes no 20a 9.7 194 7.7 285 8.1 195 no no 21a 3.7 43 6.0 127 5.6 44 no no Table 4.10 Comparison between total crash and total conflict prone locations at diverge Spearman rank correlation based on Ratio ranking me thod = -0.196 Spearman rank correlation based on PFI ranking meth od = -0.110

PAGE 117

Table 4.11 Summary of APLs of total incidents Incident location merging diverging Number of APLs identified based on total crash 6 3 Number of APLs identified based on total conflict 14 17 Common number of APLs identified based on both total crash and total conflict 5 2 Table 4.12 Average total crashes and total conflicts based on incident proneness of locations Location Crash Prone Conflict Prone Yes No Yes No Merge 18.2 6.6 582 160 Diverge 15.0 8.9 1453 206

PAGE 118

4.5 Validation Test 5: Crash and Conflict Predictio n Regression Model Comparative Analysis for Specific Type of Inc idents. 4.5.1 Development of Conflict Prediction Models for Specific Incident Types Validation test 5 repeats the same process for vali dation Test 4 but for each incident type (rear-end, lane-changing, and crossin g). Due to zero number of observed crossing incidents, crossing-type incident s were excluded from this analysis. In addition, lane change type incidents were also e xcluded from the analysis because the models didnÂ’t converge to a finite value. This phenomenon is due to the condition that the relative Hessian conv ergence criterion has been exceeded. This criterion detects the occasional cas e where the change in parameter convergence is satisfied in the GENMOD procedure by the SAS statistical software package. This is perhaps due t o the relatively low number of lane-changing incidents (both crashes and confli cts). Table 4.13 through 4.16 show the estimates of the p arameters for the rearend incidents, Model (7) through (10) respectively based on ADT and PHV values. The model parameters were assessed and reco rded.

PAGE 119

Table 4.13 Prediction model of rear-end crashes as a function of mainline ADT and merging ADT Table 4.14 Prediction model of rear-end conflicts as a functi on of mainline PHV and merging PHV Rear End Conflicts/hr merge = 0.063 x PHV mainline 0.680 x PHV merge 0.385 Degrees of R-Squared Scaled Pearson 2 0.1,30 Shape Freedom R 2 Deviance 2 Parameter 30 0.21 35.72 45.93 40.26 1.97 Variable Coefficient Constant 0.063 ADT mainline 0.680 ADT merge 0.385 Rear-End Crashes/yr merge = 9.64E-05 x ADTmainline 0.827 x ADTmerge 0.195 Degree of R-Squared Scaled Pearson 2 0.1,30 Shape Freedom R 2 Deviance 2 Parameter 30 0.43 33.33 33.03 40.26 2.61 Variable Coefficient Constant 9.64E-05 ADT mainline 0.827 ADT merge 0.195 Model ( 8 ) Model ( 7 )

PAGE 120

Table 4.15 Prediction model of rear-end crashes as a function of mainline ADT and diverging ADT Rear-End Crashes/yr diverge = 0.098 x ADTmainline -0.038 x ADTdiverge 0.515 Degrees of R-Squared Scaled Pearson 2 0.1,34 Shape Freedom R 2 Deviance 2 Parameter 31 0.61 35.77 33.14 41.42 9.40 Variable Coefficient Constant 0.098 ADT mainline -0.038 ADT diverge 0.515 Table 4.16 Prediction model of rear-end conflicts as a functi on of mainline PHV and diverging PHV Rear-End Conflicts/hr diverge = 3.68E-05 x PHV mainline 1.163 x PHV diverge 0.954 Degrees of R-Squared Scaled Pearson 2 0.1,29 Shape Freedom R 2 Deviance 2 Parameter 29 0.27 35.05 53.05 39.09 1.65 Variable Coefficient Constant 3.68E-05 VPH mainline 1.163 VPH diverge 0.954 Model ( 9 ) Model ( 10 )

PAGE 121

4.5.2 Identification of Location Prone to Rear-End Incidents Once the parameters of the prediction models were e stimated, the models were then used to identify the crashes/conflict pro ne locations. Comparison between rear-end crash and rear-end conf lict prone locations at merge and diverge are shown in Tables 4.17 and Tabl e 4.18 respectively and a summary was displayed in Table 4.19. Accordingly fewer locations were identified as accident prone by regression Models ( 7) and (9) based on actual crash than those identified by the regression Model (8) and (10) based on simulated conflicts records. Table 4.19 also showed that 5 APL was common between the crash/conflict models at merging and 2 at diverging location. This indicates a poor agreement between the rear-end cra sh and rear-end conflict models in identifying incident prone locations at b oth merging and diverging location. Table 4.20 shows the average rear-end crashes and rear-end conflicts based on incident proneness of locations. 4.5.3 Ranking Locations The PFI and Ratio rankings (as explained in test 4) were then obtained using the rear-end crash prediction models, Models (7) an d Model (9), and the corresponding rear-end conflict-prediction models, Model (8) and Model (10). Table (16) and Table (17) show the Spearman rank-co rrelations for both the Ratio and PFI rankings. These results indicated an insignificant correlation between diverging locations with “excessive rear-en d incidents” and locations with “excessive rear end conflicts.” Moreover, the discussion of incidentprone ranking comparisons of Validation Test 4 is a pplicable here as well.

PAGE 122

Table 4.17 Comparison between rear-end crash and rear-end con flict prone locations at merge ID Actual Rear-End Crash Simulated Rear-End Conflict Model Estimate (E(X)) EB Estimate Crash Prone Location Conflict Prone Location Crash Conflict Crash Conflict 1 4.0 70 4.1 203 4.0 71 no no 1a 2.0 117 3.8 180 2.7 118 no no 2 6.0 111 4.8 266 5.6 112 no no 2a 3.7 158 4.7 261 4.0 159 no no 3 0.7 50 7.6 410 2.4 52 no no 4 2.7 721 8.7 531 4.1 720 no yes 4a 7.3 663 8.6 514 7.6 662 no yes 5 12.0 687 8.7 506 11.2 686 yes yes 5a 4.3 365 8.5 452 5.3 365 no yes 6 14.3 341 8.8 539 13.1 342 yes yes 6a 3.7 268 8.0 438 4.7 269 no no 7 17.7 707 9.0 454 15.7 706 yes yes 7a 4.0 701 8.9 468 5.1 700 no yes 8 10.7 366 9.1 536 10.3 367 yes yes 8a 2.0 329 8.9 522 3.6 330 no yes 9 5.7 113 6.5 314 5.9 114 no no 9a 1.7 206 7.3 366 3.2 207 no no 10a 3.7 252 7.4 364 4.6 253 no no 11 8.7 445 7.7 384 8.4 445 no yes 12 13.3 292 7.3 347 11.8 292 yes no 12a 20.7 376 6.8 301 16.9 376 yes yes 13 4.0 19 4.3 202 4.1 21 no no 14a 3.3 156 5.1 256 3.9 157 no no 15 7.7 255 4.0 187 6.2 254 no no 15a 1.0 143 3.3 127 2.0 143 no no 16a 5.7 1209 5.1 255 5.5 1202 no yes 18 0.7 738 5.3 263 2.2 734 no yes 18a 14.3 310 5.3 258 11.3 310 yes no 19 1.7 54 5.1 252 2.8 56 no no 19a 12.3 238 5.3 265 10.0 238 no no 20 7.3 421 5.8 321 6.9 420 no yes 20a 3.0 90 4.9 224 3.6 91 no no 21 1.7 133 3.5 155 2.4 133 no no Spearman rank correlation based on Ratio ranking me thod = 0.041 Spearman rank correlation based on PFI ranking meth od = 0.026

PAGE 123

n Table 4.18 Comparison between rear-end crash and rear-end con flict prone locations at diverge ID Actual Total Crash Simulated Total Conflict Model Estimate (E(X)) EB Estimate Crash Prone Location Conflict Prone Location Crash Conflict Crash Conflict 1 5.3 127 6.5 244 6.0 128 no no 1a 3.0 75 5.5 181 4.6 76 no no 2 4.3 194 7.3 445 6.0 195 no no 2a 5.0 251 6.9 423 6.1 252 no no 3a 6.7 28 8.2 865 7.5 30 no no 4 5.3 2291 7.4 1612 6.5 2290 no yes 4a 6.3 2557 7.1 1489 6.8 2556 no yes 5 8.0 1279 9.1 1599 8.5 1279 no yes 5a 9.3 870 8.5 1207 8.9 870 no yes 6 10.7 1593 8.6 1828 9.6 1593 no yes 6a 5.3 739 6.5 1093 6.0 740 no yes 7 11.7 1253 8.1 1160 9.8 1253 no yes 7a 5.0 1678 8.0 1249 6.6 1677 no yes 8 11.3 1624 11.2 1879 11.3 1624 yes yes 8a 4.7 1287 10.5 1764 7.4 1287 no yes 9 6.7 336 5.1 510 5.6 337 no no 9a 6.0 435 6.9 743 6.5 436 no no 10 12.7 1278 8.1 867 10.2 1277 no yes 10a 2.7 844 7.7 750 5.4 844 no yes 11 9.0 2118 8.2 847 8.6 2116 no yes 11a 11.3 849 7.5 445 9.2 847 no yes 12a 6.0 687 5.7 459 5.8 686 no yes 13a 11.7 110 5.5 220 7.8 111 no no 14b 2.0 177 7.1 388 4.9 178 no no 15 7.7 119 5.5 194 6.3 120 no no 15a 2.3 75 3.3 75 3.0 75 no no 16a 14.7 2054 7.3 388 10.5 2047 yes yes 18 1.7 186 7.1 402 4.7 187 no no 18a 10.7 749 6.9 385 8.5 747 no yes 19 1.7 236 6.9 367 4.7 237 no no 19a 13.0 1630 7.3 416 9.8 1625 no yes 20 12.0 225 9.5 670 10.8 226 yes no 20a 8.3 194 6.0 277 6.9 194 no no 21a 1.3 43 4.7 128 3.6 44 no no Spearman rank correlation based on Ratio ranking me thod = -0.159 Spearman rank correlation based on PFI ranking meth od = -0.040

PAGE 124

Table 4.19 Summery of APLs of rear-end incidents Table 4.20 Average rear-end crashes and rear-end conflicts bas ed on incident proneness of locations Location Crash Prone Conflict Prone Yes No Yes No Merge 14.7 4.2 576 160 Diverge 12.7 6.6 1410 176 Incident Location merge diverge Number of APLs identified based on rear-end crash 7 3 Number of APLs identified based on rear-end conflict 14 18 Common number of APLs identified based on both rear-end crash and conflict 5 2

PAGE 125

5. Summary, Conclusions, and Recommendations 5.1 Summary and Conclusions This study has provided the results of the field va lidation effort conducted to assess the safety of freeway merging and diverging locations by comparing conflicts predicted by SSAM with the actual crash e xperience. Twenty-two interchanges, forty-two merging and forty-two diver ging locations for the field validation were selected. These locations were mode led in the VISSIM simulation system, and the simulation results were imported into SSAM to obtain a record of traffic conflicts based on the r equirements on TTC and PET values. Several modeling assumptions were made incl uding assumptions related to intersection geometry, speed profiles, v ehicle type characteristics, traffic composition, and others. Crash records were obtained from Colorado Department of Transportation (CDOT) accident databa se. Five statistical tests were applied to the data to quantify the correlation between interchange conflicts (from simulation) and interchange crashes (from CDOT database). The tests evaluated total inc ident (crash/conflict) frequencies and incident frequencies by maneuver ty pe ( lane change, or rear end). The crossing maneuver types were under-repres ented in the simulation due to the fact that the flow of traffic was in the same direction. The tests included the following: Field Validation Test 1: Merging and diverging loc ation ranking by total incidents.

PAGE 126

r Field Validation Test 2: Merging and diverging loc ation ranking by specific incident types. Field Validation Test 3: Crash and conflict predic tion regression model paired comparison. Field Validation Test 4: Crash and conflict predict ion regression model comparative analysis for total incidents Field Validation Test 4: Crash and conflict predict ion regression model comparative analysis for specific type of incidents In the first two field validation tests, the rankin g derived from SSAMÂ’s conflict prediction model was compared to the ranking from t he actual accident risk. The SpearmanÂ’s rank correlation test was used to de termine the level of agreement between the two ranks at the 90% confiden ce level. The Spearman correlation testing showed a significant c orrelation between total crashes and total conflicts, rear-end conflicts and rear-end crashes, between lane-change conflicts and lane-change crashes. Also the rank correlation tests have shown correlation between ADT and the th ree incident types. There were 1 percent merging and 4 percent divergin g lane changing conflicts were recorded, while the remaining portio n were rear end conflicts. In all the simulation scenarios, rear-end conflict events made up the bulk of the total conflicts which was consistent with study reported on crash prediction models on urban signalized intersections Guttman, Pu, Sayed, and Shelby (2008). The crash and conflict prediction regression model paired comparison involved fitting a regression equation that relates actual crash to the SSAM predicted conflicts. The goodness-of-fit measure w as tested using Scaled

PAGE 127

deviance, Pearson chi-square, and the R-square valu e. Both the Pearson chi-squared and the scaled deviance values were not significant at the 90percent confidence level, indicating a good fit. Mo reover, these models were compared using the R-squared goodness-of-fit test ( Miao, 1996), and the results conform to those of the Pearson chi-squared and the Scaled deviance. The relationship between total conflicts and total crashes exhibited a correlation (R-squared) of 0.57 and 0.90 for mergin g and diverging locations respectively. However, as a benchmark basis for comparison, tradi tional safety assessments based on average daily traffic volumes were compared to the volume-based model and conflicts-based safety asses sment. For crash prediction model based on volumes, the R -squared values were recorded to be 0.54 and 0.82 for merge and diverge approach res pectively. Thus, both the volume-based model and the conflict based model exhibited comparable correlation. Moreover, both the volume-based model and the conflict based model had a better correlation than the traditional regression model in the previous researches. In the volume-based model (ADT approach), merge and diverge locations that have the same ADT would have the same crash fr equencies, which is not always true. On the contrary, looking at crash as a function of conflict, it explained the operational differences and vehicle d ynamics of locations. Conflict models (SSAM) can be used to assess the re lative safety of a pairs of merge/diverge design alternatives. Under equivalent traffic conditions (e.g., traffic volumes and truck percentage), for both mer ge/diverge design alternatives, SSAM could exhibit statistically sign ificant differences in the total number of conflicts, the number of conflicts by typ e (crossing, lane-change, or

PAGE 128

rear-end incidents). This helps to conclude that on e design exhibited a higher frequency of conflicts, than the alternative design This type of assessment helps the decision-making process about which desig n is the safer of the two. Due to the fact that the increase in traffic volume is proportional to the occurrence of conflicts and crashes, both conflicts and crashes were correlated with interchange traffic volume. This ef fort found that the correlation between simulated conflicts and actual crashes were significant. Crash and conflict prediction regression model comp arative analysis for total incidents and for specific type of incidents involv ed the development of two prediction models relating the conflicts predicted by SSAM and the real-world crashes to traffic volume characteristics both at t he merge and diverge locations. Due to zero number of observed crossing incidents, crossing-type incidents were excluded from this analysis. In addi tion, lane change type incidents were also excluded from the analysis beca use the models didnÂ’t converge to a finite value. This was perhaps due to the relatively low number of lane-changing crashes and conflicts. The identification of crash-prone locations and con flict-prone locations were studied. This test determined whether the conflict prediction equation can predict risk in a manner similar to the crash predi ction equation for interchanges with the same characteristics, by iden tifying accident prone locations (APL). Empirical Bayes (EB) method combin es the information contained in crash counts of a specific highway seg ment with an estimate of its safety performance derived from safety performa nce function obtained from GLM models to increase the precision of estima tion. Also, the EB method corrects for the regression-to-mean bias.

PAGE 129

Comparison between crash prone locations and confli ct prone locations at merge and diverge were compared, accordingly fewer locations were identified as accident prone by regression models b ased on actual crash than those identified by the regression model based on s imulated conflicts records. It was showed that 5 APL and 2 APLs were common bet ween the total crash/conflict models at merging and diverging loca tion respectively. This indicates a poor agreement between the total crash and total conflict models in identifying incident prone locations at both mer ging and diverging location. The same conclusion is true for the rear-end and la ne changing maneuvers. The merging and diverging locations were then ranke d in terms of priority for treatment once crash/conflict-prone sites were iden tified. Two techniques were used. The first ranking criterion, Ratio crite rion, was calculated the ratio between the EB estimate and the predicted frequency as obtained from the GLM model (a risk-minimization objective). The seco nd criterion, the cost effectiveness objective, was a “potential for impro vement” (PFI) criterion, which was calculated as the difference between the observed crash/conflict frequency and the volume-based estimate of crash/co nflict frequency. The Spearman rank-correlations for both the PFI and rat io rankings were very poor and non-significant at the 90% confidence leve l for total crash/conflict and specific types of incidents. In general, the results of the validation effort of safety assessment based conflict analysis of simulated traffic at merging a nd diverging locations suggested that this technique showed great potentia l for interchange safety prediction.

PAGE 130

5.2 Recommendations For the last two decades, several legislations have been making safety a central, explicit, comprehensive, and integrated pa rt of transportation planning. These legislations have encouraged or ins isted on giving safety priority through better data, better analysis, bett er reporting systems, and the adoption of a structured comprehensive approach. A detailed study of the safety of highways would help to understand safety and help to make safety legislation a success. The SSAM approach demonstrated significant correlat ion with actual crash date. Project planners and engineers of current pro jects use a well-developed safety performance functions provided in this disse rtation to get realistic estimates of expected safety performance of merge/d iverge locations. SSAM provides a new option to assess the safety of traf fic facilities that have not yet been constructed and traffic control policies that have not yet been enacted in the field. It is also applicable to facilities wher e traditional, volume based crash-prediction models (and norms) have not been e stablished thus the SSAM approach exhibited great promise. Understandin g of the safety and design of future projects would eliminate wasting t ime and money on improvements that soon will become inadequate or in appropriate. During the modeling process, it was observed that t here was abrupt lanechanging behavior. To reduce this behavior certain measures were taken such as extending the lengths of each approach (mod el link) which provides vehicles with significantly more time to decide on their route decision. Moreover, VISSIMÂ’s lane changing algorithm needs to be revised to eliminate the abrupt lane-changing behavior that occurred dur ing traffic simulation

PAGE 131

modeling. Such an abrupt behavior can increase the number of conflicts unrealistically. The field validation test exhibited a distribution of conflicts by type that lean more heavily toward rear-end incident (less dangero us events) than the distribution of events found in actual crash record s. It is difficult to explain whether this discrepancy is due to shortcomings in the underlying simulations model (VISSIM), the conflict analysis tool (SSAM), or if the conflict occurrence in the field does indeed differ from crash occurren ce. Therefore, in order to reconcile these facts much remains to be done in th e future to study, analyze and quantify the shortcomings of simulated conflict data versus real-world crash data including any potential shortcomings of the current traffic safety analysis method. It is hoped that this dissertation will serve as a basis for future research in this direction.

PAGE 132

APPENDIX Dataset A. Interchange traffic flow information B. Average 3 year crashes recorded at merge and div erge locations C. Average yearly crashes recorded at merge and div erge locations D. Conflicts of five replications at merge and dive rge locations recorded by SSAM E. Average hourly conflicts at merge and diverge lo cations recorded by SSAM

PAGE 133

n Appendix A Interchange traffic flow information HWY NO Interchange ADT PHV ID Name Traffic Mainline On/Off Ramp Mainline On/Off Ramp Direction I-25 1 Ridgegate Pkwy NB 47,500 7,650 3,800 612 1a (truck=8.2%) SB 47,500 5,600 3,800 448 2 E Lincoln Ave NB 55,000 9,720 4,400 960 2a (truck=7.1%) SB 55,000 8,650 4,400 910 3 E County Line Rd NB 86,500 15,100 6,920 1,330 3a (truck=7.1%) SB 86,500 12,560 6,920 1,110 4 E Dry Creek Rd NB 110,500 10,430 9,945 1,370 4a (truck=6.6%) SB 110,500 9,640 9,945 1,260 5 E Arapahoe Ave NB 100,500 15,400 8,040 1,760 5a (truck=6.1%) SB 100,500 13,500 8,040 1,310 6 E Orchard Rd NB 105,000 14,060 8,400 1,920 6a (truck=5.8%) SB 105,000 8,205 8,400 1,120 7 E Belleview Ave NB 110,500 12,542 8,840 1,120 7a (truck=5.1%) SB 110,500 12,058 8,840 1,210 8 E Hampden Ave NB 96,500 23,200 7,720 2,190 8a (truck=3.9%) SB 96,500 20,360 7,720 2,050 9 E Yale Ave NB 93,500 4,930 7,480 580 9a (truck=4.5%) SB 93,500 8,941 7,480 860 10 E Evans Ave NB 90,000 12,150 7,200 1,060 10a (truck=3.9%) SB 90,000 11,000 7,200 910 11 S Colorado Blvd NB 91,000 12,650 7,280 1,020 11a (truck=6.1%) SB 91,000 10,600 7,280 520 12 University Blvd NB 93,500 8,769 7,480 750 12a (truck=6.3%) SB 93,500 6,190 7,480 520 I-225 13 E Yosemite Ave NB 53,000 6,037 4,240 500 13a (truck=5.9%) SB 53,000 5,640 4,240 480 14 S Parker Rd NB 60,000 10,030 4,800 802 14a (truck=5.9%) SB 60,000 9,350 4,800 748 15 E Iliff Ave NB 50,500 5,572 4,040 446 15a (truck=7.1%) SB 50,500 2,057 4,040 165 16 E Mississippi Ave NB 58,500 9,570 4,680 766 16a (truck=6.5%) SB 58,500 9,640 4,680 771 17 E Alameda Ave NB 60,000 12,571 4,800 1,030 17a (truck=6.5%) SB 60,000 10,681 4,800 910 18 6th Ave NB 63,000 9,140 5,040 731 18a (truck=7.2%) SB 63,000 8,750 5,040 700 19 E Colfax Ave NB 61,000 8,650 4,880 692 19a (truck=7.0%) SB 61,000 9,862 4,880 789 I-70 20 Kipling St EB 61,000 16,245 4,880 1,300 20a (truck=9.4%) WB 61,000 6,576 4,880 526 21 Sheridan Blvd EB 46,300 3,945 3,700 316 21a (truck=4.5%) WB 46,300 4,000 3,700 320

PAGE 134

#$$% "#! $! !" $ %&$$"'& ( $!)*+,++-+&$.$!.$! .$!$!.$!/#0 RELCTotalRELCTotalRELCTotalRELCTotal 1 Ridgegate Pkwy 192.07122146511165219312 2 E Lincoln Ave 192.9918422114151321515823 3 E County Line Rd 195.132810------20222 4 E Dry Creek Rd 196.14881622103216521191130 5 E Arapahoe Ave 197.193623591382124832281240 6 E Orchard Rd 198.29431053114153263816521 7 E Belleview Ave 199.3953106312820351247151126 8 E Hampden Ave 201.58327396142034842141125 9 E Yale Ave 202.64174215162042418927 10 E Evans Ave 203.54---111223388468917 11 S Colorado Blvd 204.0426733---2793634741 12University Blvd205.0640848624665616721842213 E Yosemite Ave 1.3312517------35641 14 S Parker Rd 3.951752210414314356410 15 E Iliff Ave 5.3723730347235287310 16 E Mississippi Ave 6.8926733174212943344448 17 E Alameda Ave 7.85195245161552033437 18 6th Ave 8.9626843750581332436 19E Colfax Ave9.9054937643581339746 I-7020 Kipling St 267.4122426971636104625429 (EB:WB)21Sheridan Blvd270.505813------4711 418142560283103386455127582439135574 19.96.826.713.54.918.421.76.027.720.96.427.3 7525100732710078221007624100 Three Year Crash Record IDNameMile Post merge diverge NB/EB SB/WB Interchange NB/EB SB/WB I-25 I-225 HWY NO Sum Average Yearly Crashes by Type Percentage of Crashes by Type

PAGE 135

%&$$"'& ( $!)*+,++-+&$.$!.$! .$!$!.$!/#0 #$$ ## ! $! !" $ RELCTotalRELCTotalRELCTotalRELCTotal 1 Ridgegate Pkwy 192.074.00.74.72.01.73.75.31.77.03.01.04.0 2 E Lincoln Ave 192.996.01.37.33.71.35.04.30.75.05.02.77.7 3 E County Line Rd 195.130.72.73.3------6.70.77.3 4 E Dry Creek Rd 196.142.72.75.37.33.310.75.31.77.06.33.710.0 5 E Arapahoe Ave 197.1912.07.719.74.32.77.08.02.710.79.34.013.3 6 E Orchard Rd 198.2914.33.317.73.71.35.010.72.012.75.31.77.0 7 E Belleview Ave 199.3917.73.321.04.02.76.711.74.015.75.03.78.7 8 E Hampden Ave 201.5810.72.313.02.04.76.711.32.714.04.73.78.3 9 E Yale Ave 202.645.71.37.01.70.32.06.71.38.06.03.09.0 10 E Evans Ave 203.54---3.74.07.712.72.715.32.73.05.7 11 S Colorado Blvd 204.048.72.311.0---9.03.012.011.32.313.7 12University Blvd205.0613.32.716.020.71.322.018.75. 324.06.01.37.3 13 E Yosemite Ave 1.334.01.75.7------11.72.013.7 14 S Parker Rd 3.955.71.77.33.31.34.710.31.311.72.01.33.3 15 E Iliff Ave 5.377.72.310.01.01.32.37.71.79.32.31.03.3 16 E Mississippi Ave 6.898.72.311.05.71.37.09.71.311.014.71.316.0 17 E Alameda Ave 7.856.31.78.01.70.32.05.01.76.711.01.312.3 18 6th Ave 8.960.72.02.714.32.316.71.72.74.310.71.312.0 19E Colfax Ave9.901.71.33.012.32.014.31.72.74.313.0 2.315.3 20 Kipling St 267.417.31.38.73.02.35.312.03.315.38.31.39.7 21Sheridan Blvd270.501.72.74.3------1.32.33.7 1394718794341291524219414645191 6.62.38.94.51.66.17.22.09.27.02.19.17525100732710078221007624100 Sum Average Yearly Crashes by Type Percentage of Crashes by Type NB/EB SB/WB NB/EB I-25 I-225 HWY NO Average Yearly Crash IDNameMile Post merge diverge I-70 SB/WB Interchange

PAGE 136

r %&$$"'& ( $!)*+,++-+&$.$!.$! .$!$!.$!/#0 #$$& $111/$ $! !"$!!.#,,2 RELCTotalRELCTotalRELCTotalRELCTotal 1 Ridgegate Pkwy 192.073490349585058563706373750375 2 E Lincoln Ave 192.9955605567910791648216691,158951,253 3 E County Line Rd 195.1325216268------12614140 4 E Dry Creek Rd 196.143,60303,6033,31403,31410,0241,43211,45611,765 1,02212,787 5 E Arapahoe Ave 197.193,435313,4661,82501,8255,5918016,3923,8604914 ,351 6 E Orchard Rd 198.291,706621,7681,34201,3426,9051,0587,9633,11557 83,693 7 E Belleview Ave 199.393,535603,5953,50363,5095,7025666,2687,9114838 ,394 8 E Hampden Ave 201.581,82901,8291,645111,6568,109118,1205,9774596, 436 9 E Yale Ave 202.6456505651,02951,0341,592891,6812,0651092,174 10 E Evans Ave 203.54---1,26051,2656,0083816,3893,7484714,219 11 S Colorado Blvd 204.042,22502,225---9,75084010,5904,0142364,250 12University Blvd205.061,46201,4621,87801,8782,7812 613,0423,2651703,435 13 E Yosemite Ave 1.33959104------5510551 14 S Parker Rd 3.9523,77527124,046780078044,71530445,0198850885 15 E Iliff Ave 5.371,27651,2817150715581165973740374 16 E Mississippi Ave 6.8926,03350926,5426,0451556,20027,23020127,43110,0 1225410,266 17 E Alameda Ave 7.8537,99529238,28739,53731639,85324,84524225,08721 ,23511621,351 18 6th Ave 8.963,690413,7311,551101,561904269303,656923,748 19E Colfax Ave9.9027102711,188101,1981,095851,1808, 090598,149 I-7020 Kipling St 267.412,104352,13945204521,035901,12593931970 (EB:WB)21Sheridan Blvd270.5066310673------2170217 115,4191,341116,76067,44051867,955158,4556,421164,8 7593,3384,68098,010 5,771675,8383,967303,9978,8033579,1604,4452234,667 991100991100964100955100 SB/WB HWY NO Interchange Number of Conflicts in Five Replications IDNameMile Post merge diverge NB/EB SB/WB NB/EB I-25 I-225 Sum Average (per merge/diverge location) Percentage (by type)

PAGE 137

%&$$"'& ( $!)*+,++-+&$.$!.$! .$!$!.$!/#0 #$$ "# $1 "$!!"$ !!.#,,2 RELCTotalRELCTotalRELCTotalRELCTotal 1 Ridgegate Pkwy 192.07700701170117127012775075 2 E Lincoln Ave 192.9911101111580158190419423219251 3 E County Line Rd 195.1350353------25328 4 E Dry Creek Rd 196.14721072166306632,0052862,2912,3532042,557 5 E Arapahoe Ave 197.19687669336503651,1191601,27977298870 6 E Orchard Rd 198.293411235326802681,3812121,593623116739 7 E Belleview Ave 199.397071271970117021,1401131,2531,582961,678 8 E Hampden Ave 201.58366036632923311,62221,6241,195921,287 9 E Yale Ave 202.64113011320612073181833641322435 10 E Evans Ave 203.54---25212531,202761,27875094844 11 S Colorado Blvd 204.044450445---1,9501682,11880346849 12University Blvd205.062920292376037655652608653346 87 13 E Yosemite Ave 1.3319221------1100110 14 S Parker Rd 3.954,755544,80915601568,943619,0041770177 15 E Iliff Ave 5.3725512561430143116311975075 16 E Mississippi Ave 6.895,2071025,3091,209311,2405,446405,4862,002522,0 54 17 E Alameda Ave 7.857,599587,6577,907637,9704,969495,0184,247234,27 0 18 6th Ave 8.9673887463102312181518673118749 19E Colfax Ave9.90540542382240219172361,618121,630 I-7020 Kipling St 267.41421742890090207182251886194 (EB:WB)21Sheridan Blvd270.501332135------43043 23,08426723,35113,48810313,59131,6911,28432,97518,6 6793519,602 1,154131,16879367991,761711,83288945933 991100991100964100955100 SB/WB NB/EB SB/WB Average Hourly Conflicts IDNameMile Post merge diverge Average (per merge/diverge location) Percentage (by type) NB/EB I-25 HWY NO Interchange I-225 Sum

PAGE 138

BIBLIOGRAPHY A policy on geometric design of highways and street s. (2004). America Association of State Highway and Transportation Officials (AASHTO), Washington, D.C. Amundsen, F. and Hyden, C.(1977). Proceedings of first workshop on traffic conflicts Oslo: Institute of Transport Economics. Archer, J. (2000 ). Developing the potential of microsimulation modelin g for traffic safety assessment. Proceedings of 13th Annual International Cooperation on Theories and Concepts on Traffic Saf ety Workshop Corfu, Greece. Bahar, G., and Morris, N. (2007): Is a strong safety culture taking root in our highway agencies? AAA Foundation for Traffic Safety. iTRANS Consulting. Bared, J., Giering G., Warren, D., (May 1999). Safety of evaluation of acceleration and deceleration lane lengths Institute of Transportation Engineers Journal, 50-54. Barnett, V. and Lewis, T. (1978). Outlier in statis tical data. John Wiley & Sons Inc., New York. Barnett, V. and Lewis, T. (1994). Outliers in statistical data ( 3rd ed.). New York, Wiley, 33-34. Bauer, K.M. and Harwood, D.W. (1998). Statistical models of accidents on interchange ramps and speed-change lanes FHWA-RD-97-106, FHWA, U.S. Department of Transportation.

PAGE 139

Brude, U., and Larsson, J. (1988). The use of prediction models for eliminating effects due to regression-to the mea n in road accident data. Accident Analysis and Prevention, 20(4), 299-310. Calvin, M., (1990). Bayesian analysis Handbook of statistical methods for engineers and scientists, Wadsworth H. Editor, M cgraw-Hill. Cirillo, J.A. 1967. Interstate system accident research study II An interim report, Part I, Highway Research Record, no. 188. Cirillo, .A. (1968). Interstate system accident research study II An interim report, Part II, Public Road. Cirillo, J.A. 1970. The relationship of Accident to Length of Speed-Cha nge Lanes and Weaving Area on Interstate Highways Highway Research Record 312. Depue, L. (2003). Safety management systems: A synthesis of highway practice NCHRP Synthesis 322. National Cooperative Highwa y Research Program. Washington, DC: Transportatio n Research Board. Federal Highway Administration (FHWA). 2001a. Transportation safety planning: Safety-conscious planning Transportation E-Circular E-C025 Gettman, D. and Head, L. (2003). Surrogate measures of safety from traffic simulation models Report No. FHWA-RD-03-050. Federal Highway Administration (FHWA): Washington, D.C

PAGE 140

Gettman, D., Pu,L., Sayed, T., Shelby, S. (2008). Surrogate safety assessment model and validation: Final Report. Report No. FHWAHRT-08-051. Federal Highway Administration (FHW A): Washington, D.C. Hauer, E. (1980). Bias-by-selection. Accident Analysis and Prevention, 12, 113–117. Hauer, E. (1992). Empirical bayes approach to the estimation of ‘unsafety’: The multivariate regression method Accident Analysis and Prevention 5(5), 457-477. Hauer, E. (1997). Observational before-after studies in road safety Pergamon Press. Hayward, J.C. (1972). Near-miss determination through use of a scale of danger Report No. HRR 384. Highway Research Board, 24–35 Higle, L. and Witkowski, M., (1988). Bayesian identification of hazardious locations Transportation Research Record, 1185, 24-31. Hyden, C. (1987). The development of a method for traffic safety evaluation: The Swedish Conflicts Technique Department of Traffic Planning and Engineering,Lund University: Sweden. Institute of Transportation Engineers, California B order Section, Highway Capacity Task Force (2004). A Report on the use of Traffic Simulation Models in the San Diego Region Kulmala, R., (1995). Safety at rural threeand four-arm junctions. Development of accident models. Technical Research Center of Finland, VTT 233.

PAGE 141

Lord, D., Bonneson, J., (2005). Calibration of pred ictive models for estimating safety of ramp design configurations. Transportation Research Record 1908, 88-95. Lundy, A. (1966) The effect of ramp type and geometry on accidents (2nd ed.). Traffic Department, Division of Highways, D epartment of Public Works, State of California. McCartt, A., Northrup, V., Retting, R. (2004). Types and characteristics of ramp-related motor vehicle crashes on urban inte rstate roadways in Northern Virginia Journal of Safety Research, 35, 107-114. Miaou, P. (1996). Measuring the goodness-of-fit of accident predictio n models Report No. FHWA-RD-96-040. Federal Highway Administration (FHWA): McLean, VA. Migletz, D. et al. (1985). Relationships between tr affic conflicts and accidents. Report No. FHWA-RD-84-042, Federal Hi ghway Administration, Washington, DC. Mullins, K., and Keese, J. (1961). Freeway traffic accident analysis and safety study. Highway Research Board, 291, 26-78. Nelder, A. and Wedderburn, M. (1972). Generalized linear models. Journal of the Royal Statistical Society, Series A, 135(3), 370. Osgood, W. (2000). Poisson-based regression analysis of aggregate crime rates Journal of Quantitative Criminology, 16: 21-43. Parker, R. and Zegeer, V. (1989), Traffic conflict techniques for safety and operations--Observers Manual Report No. FHWA-IP-88-027. Federal Highway Administration (FHWA): Washington D.C., 40.

PAGE 142

Paternoster, R., and Brame, R. (1997 ). Multiple routes to delinquency? A test of developmental and general theories of crime Criminology,35(1), 45-84. Perkins, R., and Harris, I. (1968). Traffic conflic t characteristics: Accident potential at intersections. Highway Res earch Record, 225, 45-143. Highway Research Board, Washington DC Pu L.,and Joshi R. (2008). Surrogate safety assessment model (SSAM): Software User Manual Report No. FHWA-HRT-08-050. Federal Highway Administration (FHWA): Washington, D.C. Risser, R. (1985). Behavior in traffic conflict sit uations. Accident Analysis and Prevention, 17(2), 179-197. Roess, R., Prassas, E., and McShane, W. (2004). Tr affic engineering. (3rd ed.). New Jersey: Pearson Education, Inc. Roess, P., and Ulerio, M. (1993). Capacity and level of service at rampfreeway junctions Final Report. NCHRP project 3-37, Polytechnic university, Brooklyn, N.Y. Sarhan, M., Hassan, Y., and Abd El Halim, A.O. (200 8). Safety performance of freeway sections and relation to length of speed change lanes. Canadian Journal of Civil Engineering 34(5), 531-541. SAS Institute Inc., SAS/STAT Software: The GENMOD Procedure Release 6.09, SAS Technical Report P-243, Car y, NC, 88 pp, 1993. SAS Institute Inc., SAS/STAT Software.(2011). The GENMOD Procedure Release 9.3.

PAGE 143

n Sayed, T. (1995). A highway safety expert system: A new approach to safety programs Unpublished doctoral thesis, University of Britis h Colombia. Sayed, T (1997). Estimating the safety of unsignali zed intersections using traffic conflicts. Third international symposium on intersections without traffic signals. Oregon: Transportation Research Board. Sayed, T., Brown, G., and Navis, F.(1994). Simulation of traffic conflicts at unsignalized intersections with TSC-Sim. Accident Analysis and Prevention, 26(5), 593–607 Sayed, T., Feng, C., and Rodriguez, F. (1998). Accident prediction models for signalized intersections in British Colombia Paper presented at the 68th Institute of Transportation Engineers Annual Meeti ng. Toronto, Canada. Sayed, T. and Rodriguez, F. (1999). Accident prediction models for urban unsignalized intersections in British Columbia. Transportation Research Record, 1665, 93– 99. United Nations General Assembly Verbotim Report meeting 38 session 60 page 6, Mr. Al-Hinai Oman on 26 October 2005 (retrieved 2008-0709) U.S. Department of Transportation. (2006). Traffic Safety Facts (early ed.) National Highway Traffic Safety Administration (NHTSA). (DOT HS 810 809 Updated March 2008) U.S. Department of Transportation. (2009). Traffic Safety Facts (early ed.). National Highway Traffic Safety Administr ation (NHTSA).

PAGE 144

VISSIM 5.2 User Manual. Karlsruhe, Germany: PTV Pla nung Transport Verker AG, 2009. World Health Organization. The Global Burden of Disease 2004 Update.