|Ph.D Student||Shmoueli Sitvanit|
|Subject||Development of Time Dependent Behavior Models for Gap|
Acceptance and Capacity at Roundabout
|Department||Department of Civil and Environmental Engineering||Supervisors||PROF. Abishai Polus|
|PROFESSOR EMERITUS Moshe Livneh (Deceased)|
This research analyzes critical gaps and entry capacities at roundabout intersections in order to evaluate the effect of waiting times on the approach leg, on drivers’ critical gaps, and consequently on entry capacity. The main hypothesis is that an increase in drivers’ average waiting times results in driver’s increased impatience and, consequently, in their willingness to accept smaller gaps. This willingness provides a significant addition in capacity compared to the capacity calculated on the assumption of a fixed critical gap, which is unaffected by the waiting times of drivers on the approach leg. This phenomenon has been observed and modeled at seven busy roundabouts with long queues and long waiting times.
.A logistic model for this change in the critical gap s-curving between the maximum and minimum values was developed. The parameters of each site studied, were found to correlate with the geometric complexity and traffic intensity at that particular location. The parameters included the outside diameter, pedestrian volume, and circulating flow of vehicles around the central island.
Another model described the driver’s behavior as a binary choice model with two alternatives: accepting or rejecting a gap. The model was calibrated in the maximum likelihood method using the Alogit program. It was found that an increase of the waiting time by 6.7 seconds, results in a reduction of 0.1 second in the critical gap.
A reduction in the critical gap results in a significant increase in the entry capacity. An new non-linear exponential entry capacity model was developed; The parameters included the circulating flow, outside diameter and critical gap. In the last part of the study, a simulation model in ARENA program was developed for comparing the driver’s behavior with constant critical gap and time dependent critical gap. It showed the increase of the capacity when using the time dependent critical gap. The thesis concludes with a comparison of the entry capacity between the basic HCM model, the new exponential model and the result of the simulation analysis.