부산대학교 도서관

Frequent accidents are taken in public places in recent years, causing heavy casualties and economic losses. It is necessary for us to research effective evacuation. Provide evacuation route guidance strategy for the evacuated individuals is the key for emergency evacuation. This paper proposes a evacuation model based on Artificial Fish Awarm Algorithm (AFSA). Define the evacuee as an intelligent artificial fish and utilize the preying behavior, swarming behavior and following behavior of artificial fish swarm to simulate the mental activity, path selection, behavior preference of individuals. The model embody the characteristics of evacuation rules and uncertainties in the process and we can get the optimum path planning. In this paper, we take Zhuankou Stadium in Wuhan Hubei as the experiment environment, to conduct a research of the simulation of evacuation process under emergency. The results of simulations show that this method can and balance the congestion and improve the efficiency and fidelity of evacuation, compared with existing algorithm.