Abstract:

In order to make the evacuation of personnel in the multi-exit scenario more consistent with reality, a dynamic exit selection model considering the replacement of target exits was established based on CA. Comprehensively considering the distance from pedestrians to the exit and the impact of congestion before the exit on their selection, the hesitation and randomness of pedestrians in choosing an exit by selecting ambiguity and exit retention rate were simulated. Then, according to the change of the evacuation environment, the probability of each exit being selected was calculated, and the target exit of pedestrians was updated. The differences between dynamic egress selection policies and static egress selection strategies were compared. The results show that the dynamic exit selection strategy has two stages: invalid at low density and effective at high density. In addition, selecting small ambiguity at high density is more conducive to evacuation. Whether the increase in the exit retention rate can improve the efficiency of evacuation is related to the ambiguity of the selection. What's more, the increase in the radius of the exit area can reduce the evacuation time to a certain extent. Compared with the static exit selection strategy, this strategy can improve the export utilization rate, alleviate the imbalance of the density of each outlet, and improve evacuation efficiency.

Key words: multi-exit scenarios, dynamic exit choice, exit retention rate, select ambiguity, cellular automata