Modeling Permissive Left-Turn Gap Acceptance Behavior at Signalized Intersections

The research presented in this thesis, studies driver gap acceptance behavior for permissive left turn movements at signalized intersections. The thesis attempts to model the gap acceptance behavior using three different approaches, a deterministic statistical approach, a stochastic approach, and a psycho-physical approach. First, the deterministic statistical modeling approach is conducted using logistic regression to characterize the impact of a number of variables on driver gap acceptance behavior. The variables studied are the gap duration, the driver's wait time in search of an acceptable gap, the time required to travel to clear the conflict point, and the rain intensity. Considering stochastic gap acceptance, two stochastic approaches are compared, namely: a Bayesian and a Bootstrap approach. The study develops a procedure to model stochastic gap acceptance behavior while capturing model parameter correlations without the need to store all parameter combinations. The model is then implemented to estimate stochastic opposed saturation flow rates. Finally, the third approach uses a psycho-physical modeling approach. The physical component captures the vehicle constraints on gap acceptance behavior using vehicle dynamics models while the psychological component models the driver deliberation and decision process. In general, the three proposed models capture gap acceptance behavior for different vehicle types, roadway surface conditions, weather effects and types of control which could affect the driver gap acceptance behavior. These findings can be used to develop weather responsive traffic signal timings and can also be integrated into emerging IntelliDrive systems.