A hybrid model for collective motion of pedestrians

Pedestrian crowds display elements of collective behaviour during emergency situations. Collective patterns are not restricted to humans, but have been observed in other biological systems that display herding, flocking, schooling and swarm intelligence. Previous research into animal dynamics has explored collective behaviour in various biological systems. Even in molecular/granular systems, pattern formation has been reported through extensive molecular dynamics research. This paper present a hybrid microscopic model of the collective motion of pedestrians based on animal dynamics and molecular dynamics. It is observed that even with a relatively simple mathematical formulation, the model is able to produce the collective motion of pedestrians. The model's performance is assessed by comparing its predictions with observed pedestrian traffic from experimental data. The capability of the model for addressing extreme case as in panic is also discussed.