Evacuation Simulation in Floor Field by Real-Coded Cellular Automata

Abstract

So far, Cellular Automata (CA) have been applied in a variety of scientific fields, including traffic problems and biological researches. Especially, as one of the excellent achievements, CA models for pedestrian dynamics have been developed. Since the pedestrian flows are caused by collective crowd behavior, it is difficult to model the pedestrian dynamics based on differential equations. Although the social force model has been proposed [1], CA approach could be more appropriate to describe pedestrian dynamics, because pedestrian flows are naturally emerged in a collective behavior of pedestrians [2-5].

In our previous study, we have proposed a new approach for pedestrian dynamics. We call it a Real-coded Cellular Automata (RCA) [6,7]. The idea is based on the Real-coded Lattice Gas (RLG) developed for fluid simulation [8]. We applied this scheme to the CA model for pedestrian dynamics. The numerical procedure is not explained here. Similar to RLG, the position and velocity of the pedestrian can be freely given, independent of grid points. It has been confirmed that the movement of pedestrians in an oblique direction is successfully described by RCA, which was not taken into account in the former CA models. Some benchmark simulations including room evacuation and lane formation in the street have been conducted.