Simulation of crowd evacuation in fire based on agent behavior

Zhuomo Lu,Wenhu Qin
DOI: https://doi.org/10.3969/j.issn.1001-0505.2011.06.031
2011-01-01
Abstract:For guiding the crowd evacuation in fire, a new crowd evacuation model based on agent's behavior which can improve the realism of crowd evacuation is proposed. Through analyzing the influences of many factors on the evacuation process, including agent's different reaction time when encountering the fire alarm, different moving velocities in the process of evacuation, application of group movement and the selection of exits, a model combining the characteristics of agent behaviors and cellular automata is established. The model includes the agent's response of fire alarm model, the agent's evacuation speed model, the group movement model and exit selection model. The simulation results show that different agent behavior in fire remarkably influence total time of crowd evacuation in building. Crowd evacuation simulation integrating agent's reaction time and group movement can improve the realism greatly.
What problem does this paper attempt to address?