Abstract: The emergency evacuation of metro station passenger flow is studied in the perspective of combination modeling in the event of emergencies. Firstly, the influencing factors of passenger flow outburst and its evacuation is analyzed. A cellular automata theory-based combined model on emergency evacuation of subway passenger volume is constructed, including optimal route choice model, pedestrian movement model, obstacle avoidance model, lane change and transcendence models, etc. Following an empirical research on the Qinglongsi Subway Station in Xi'an, the AnyLogic simulation platform is adopted to simulate and analyze the emergency evacuation process of passenger flow. The result indicates that the channel located between the platform floor and the underground hall will be crowded if the exit and path information cannot be obtained timely and completely as the pedestrian flow intensity is relatively large under the emergency evacuation process. However, the crowds could be evacuated effectively with the real-time release of evacuation information and under the guidance of the order of the flow and its walking path.

Key words: metro, emergency events, passenger flow evacuation, AnyLogic simulation