A CA-based Model for Simulating Vehicular-Pedestrian Mixed Flows in a Congested Network

In design of evacuation plans for major metropolises, it should be recognized that a potentially large number of evacuees depend either on transit or other modes for evacuation, or need to walk over a distance to access their cars. In the process of approaching some designated pick-up points or intermediate destinations, the massive number of pedestrians may incur tremendous burden to vehicles in the roadway network. Hence, development of a simulation tool, capable of replicating the realistic road condition for both the pedestrian and vehicle flows under the chaotic situation, is an imperative task. Such a simulation tool should be able to account for individual behaviors as well as all kinds of interactions, including conflicts between pedestrians, between vehicles, and between pedestrians and vehicles. Despite the increasing use of traffic simulation as the primary analysis tool, effective mechanisms to simulate the mixed vehicle-pedestrian flows under congested environments remains at its infancy. In this study, we attempt to address this vital subject with the Mixed-Cellular Automata (MCA) method. Our proposed simulation model has integrated the strengths of the CA method with some probabilistic functions, offering a realistic mechanism to reflect the competing and conflict interactions between vehicle and pedestrian flows. Although the development remains at its preliminary stage, our experimental results clearly indicate that failing to account for the impact of mixed flow interactions in a congested traffic system could result in far underestimate of the delay, travel time, and system throughput. Xin Zhang, Gang-Len Chang 3