A floor field cellular automaton for crowd evacuation considering different walking abilities

Abstract

It has been hard to model a crowd evacuation process considering different walking abilities using a synchronous cellular automaton. That is because the cross and the overlaps of routes have to be taken into consideration and the conflicts resolution between pedestrians is more complex. However, the desired velocities of evacuees might be quite different due to the discrepancies of the physiological function, including age, gender, physical state, and the psychological behavior, such as the perception and reflection to the dangers. Additionally, an evacuee might change his desired velocity constantly to adapt to the changing evacuation environment. Thus, a multi-velocities floor field cellular automaton model was established in this paper. Using little CPU time, a dense crowd evacuation simulation with tiny varied velocity can be conducted very well. Significant discrepancies between the single-velocity evacuation and the multi-velocities evacuation were observed. The plateaus, where the exit flow rate is rather low, can be well predicted by a dimensionless parameter describing the congestion level of the evacuation system. The crowd evacuation time almost depends on the low desired velocity evacuees, though the proportion is not high. We also observed that faster evacuees make the evacuation system easily approaching to the jam.