Stadium Evacuation based on Multi-agent System

In order to design a reasonable pedestrian evacuation and exclude the security risks in large stadiums, this thesis proposes the research of security model of evacuation in stadiums combined with multi-agent and cellular automata. The research is based on cellular automata model and the process makes extended analysis to the cell’s behavior of autologous, and then it makes simulation experiments of the process of simulating the evacuation in large stadiums. Simulation results finally show that cell Agent combines the advantages of multi-agent and cellular automata, which fully considers the individual internal factors. Compared to traditional cell cellular automata, it is more close evacuation situation of the realistic major sports stadium, and it shortens the evacuation time and improves the safety.