Effects of clustering and a priori knowledge in panic dynamics

Abstract

We use an agent-based cellular automata model to study the effect of a priori knowledge of the location of exit and clustering in the throughput of panicking pedestrians. We found that the throughput decreases with the clustering coefficient when the a priori knowledge of the exit is fixed for all the pedestrians inside the room. When the clustering coefficient is dependent on the actual distance of the pedestrian from the exit, the throughput is proportional to the size of the area where exit location is known by the panicking pedestrians. When there is no a priori knowledge of the exit, there exists an optimal value of clustering coefficient that gives a maximum throughput.