Acquisition of local neighbor rules in the simulation of pedestrian flow by cellular automata

Abstract

Cellular Automata is applied to model the pedestrian flow, in which the local neighbor rules implemented to each person in the crowd are determined automatically in the process of simulation. The collision patterns in the visible area settled around each person is introduced, and each person walks around just avoiding the collision patterns. As the simulation proceeds along the time step, various collision patterns are observed. By being informed the collision patterns to all the pedestrians simultaneously, the pedestrian begins to flow quite smoothly and the column formation is observed. © Springer-Verlag 2004.