Agent-based modeling of a multi-room multi-floor building emergency evacuation

Panic during emergency building evacuation can cause crowd stampede, resulting in serious injuries and casualties. Agent-based methods have been successfully employed to investigate the collective human behavior during emergency evacuation in cases where the configurational space is extremely simple–usually one rectangular room–but not in evacuations of multi-room or multi-floor buildings. This implies that the effect of the complexity of building architecture on the collective behavior of the agents during evacuation has not been fully investigated. Here, we employ a system of self-moving particles whose motion is governed by the social-force model to investigate the effect of complex building architecture on the uncoordinated crowd motion during urgent evacuation. In particular, we study how the room door size, the size of the main exit, the desired speed and the friction coefficient affect the evacuation time and under what circumstances the evacuation efficiency improves.