An optimal guidance strategy for fire evacuations: A hybrid modeling approach

Abstract

Evacuation plans are subject to many restrictions in the event of a fire, and unreasonable evacuation guidance methods can affect efficiency and even cause danger. In the current study, we investigate the optimal guidance strategy for fire evacuation to obtain the optimal guider layout and the evacuation paths. The environmental constraints are generated in conjunction with FDS fire simulation to transform the evacuation space into a pseudo-static obstacle grid. A multi-objective model of the guider layout is constructed, and the guider number is used as an optimization variable. The pseudo-static obstacle grid is combined with the exit fluency strategy to plan paths. The evacuation time and cost of the crowd are adopted as two objective functions. Finally, the optimal guider layout is generated by iterative optimization of NSGA-II algorithm with the introduction of a chromosome fragment deletion operator. The simulation results and comparative analysis indicate that the model can improve fire evacuation safety, enhance the utilization and balance of each exit, and reduce evacuation time and resource redundancy.