Social Field Model to Simulate Bidirectional Pedestrian Flow Using Cellular Automata

Abstract

A collective phenomenon appearing in the simulation of bidirectional pedestrian flow in corridors is dynamic multi-lane (DML) flow. We present a cellular automata model that reproduces this behavior. We propose to incorporate a social distance emulating a territorial effect through a social field, similar to the dynamic floor field of Burstedde et al. (Physica A 295:507–525, 2001). This model also considers a vision field allowing a pedestrian to collect information from cells in front of him/her and to get the weighted social parameter as well; the importance of this parameter in the formation of dynamic lanes is that it helps a pedestrian to choose the lane that contains the highest concentration of persons walking in the same direction. We present numerical simulations in corridors with bidirectional flow and the fundamental diagram for unidirectional flow.