|Ph.D Student||Farah Haneen|
|Subject||Development of Models to Evaluate Safety of Two-Lane|
Highways Based on Infrastructure, Traffic
Characteristics and Drivers' Behavior
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Abishai Polus|
|Professor Shlomo Bekhor|
|Full Thesis text|
Two-lane rural highways constitute a substantial part of the roads in Israel and they are different from other types of highways because passing maneuvers on two-lane highways involve driving on the opposite lane direction. Passing maneuvers are considered in the literature as a complicated task that has direct implications on traffic flow and safety. Two important stages involved in the passing maneuver are studied: the decision to pass and the risk associated in completing a safe pass. Literature review revealed many gap-acceptance models for intersections, but very limited passing gap acceptance models for two-lane highways. Additionally, limited safety models capture the impact of the human factor together with the road and traffic characteristics simultaneously. This study investigates drivers' passing behavior and develops gap-acceptance and safety related models for two- lane rural highways.
Data related to passing maneuvers on two-lane highways are difficult to collect; therefore, the approach of this study in investigating passing behavior is a use of an interactive driving simulator in a laboratory environment. Drivers' personal data are collected by means of a questionnaire. To obtain additional and detailed information, two laboratory experiments were conducted for each driver: IOWA Gambling Task (IGT) and driving simulator session. The IGT focuses on the pattern of learning from experience when facing repeated decisions. The same participants were invited to drive in a driving simulator session.
The study findings with respect to drivers' decisions to pass reveal that the frequency of overtaking maneuvers on a driving simulator is associated with a flawed decision making in the IGT. In order to understand how drivers decide whether to pass or not, a gap acceptance model which incorporates variables that capture the infrastructure, traffic and drivers' individual characteristics is developed. The results indicate that all these of variables significantly affect passing behavior.
The study findings regarding the safety level as related to passing maneuvers revealed that drivers are not homogenous and they can be divided into different groups. Some drivers overestimate the probability to accept a passing-gap and others underestimate it. This idea is inspired from the concept of the Prospect Theory. According to the over and under estimation a measure of drivers' risk-taking level is developed. This risk-taking measure is found to be related to drivers' characteristics and their driving styles. It is also found that the road conditions impact drivers' risk taking.
A risk prediction model, which evaluates the risk associated with each passing maneuver, is developed. The risk measure was chosen to be the remaining distance from the opposing vehicle at the end of the passing maneuver.
The models developed in this study are a unique contribution to the state-of-the art, both at the methodological level and at the empirical level. The results presented enhance the knowledge about the factors that affect the decision to pass. They also show the remaining headway at the end of the passing process, depending on the same characteristics. A natural extension of this research could be in the implementation of the models in traffic micro- simulation models.