Simulation of evacuation based on Multi-Agent and cellular automaton

On the basis of analysis of traditional discrete cellular automata and Agent-based models, a simulation model has been improved and optimized. The model combines method of the traditional discrete cellular model´s investigation on the surrounding environment with the idea of describing individual behavior by the Agent model. Fully considering the escaping individuals´ difference and taking into account the respective sense, the adjacent persons´ behavior, the person density and the environment simulates, reflect the process of evacuation accurately. By the crowd evacuation model simulation of a supermarket, the outcome of the simulation is explicitly analyzed and studied, which proved the rationality and applicability of the model.