Modeling the pedestrian behavior at conflicts with vehicles in multi-lane roundabouts (a cellular automata approach)

This paper discusses the conflicts between vehicle flows and crossing pedestrians at multi-lane roundabout. The conflicts between pedestrians and vehicle in recent years had been of great importance and attempts have been made for analyzing this conflict in different urban street segments such as midblock and signalized intersections. In addition, roundabouts were less emphasized in the related literature and the type of conflicts is quite different due to specific pattern of movements of vehicles. The present study aimed to analyze the pedestrian behavior pattern in an urban multi-lane roundabout. The required data were collected using the video images, along with the trajectory of 89 pedestrian movements. The cellular automata modeling was used to analyze the time and space of the possible conflicts between pedestrians and vehicles. Then, time-to-collision (TTC) and Post-Encroachment Time (PET), as two surrogate safety measures, were used to define the possible conflicts between pedestrian and vehicles. In addition, binary logit models were used for modeling the choice behavior of pedestrians encountering conflicts in the roundabout. The results suggested that middle-aged/elderly pedestrians compared to the teenager/young people are 7% less prone to conflict with vehicles when crossing the roundabout. In addition, the likelihood of conflicts reduced by 26% and 8% when the pedestrians and vehicles had lower speed, respectively. During the possible conflicts, pedestrians selected adjacent cells when the front cells are full, or stay in their place. Otherwise, there is the desire to select the front cells under different conditions. The results can be used to assess the safety of pedestrian in the multi-lane roundabouts and play a significant role in improve pedestrian safety and having better understanding of pedestrians' behaviors in roundabouts. (C) 2020 Elsevier B.V. All rights reserved.