Agent-Based Approach for Crowded Pedestrian Evacuation Simulation
Xianye Ben,Xifa Huang,Zhaoyi Zhuang,Rui Yan,Sen Xu
DOI: https://doi.org/10.1049/iet-its.2011.0236
IF: 2.7
2013-01-01
IET Intelligent Transport Systems
Abstract:Pedestrian evacuation has long been a vital safety concern for any large indoor facility, for example, gymnasium or stadium. To efficiently simulate crowded pedestrian evacuation, an agent-based modelling approach in the cellular automata (CA) environment is proposed in this study. Different from a stand-alone CA method which roughly describes the external environment, the proposed agent-based modelling approach can describe individual behaviour more accurately. In order to verify the crowded pedestrian evacuation simulation and reflect the behaviour of pedestrian crowds' evacuation, this study simulates four evacuation scenarios as follows: (i) an ordered activity area with no obstacles, (ii) an unordered activity area with no obstacles, (iii) an ordered activity area with obstacles and (iv) an unordered activity area with obstacles. The effects of the parameters on the evacuation simulation process and the effects of maximal endurance capability on the number of casualties are also analysed. Furthermore, the order of evacuation of pedestrians with different competitive capabilities is estimated. The simulation results show that the proposed modelling framework, principles and methods are effective, and the model has a strong capability to describe, represent and explain the reality of evacuation.