Multi-Agent Evacuation Simulation Of Buildings Considering Leaders And Panic

How to organize crowd escape reasonably and efficiently in an emergency is an important research content in the field of crowd evacuation.However,there are many factors that affect the evacuation in reality,which makes the study of crowd evacuation challenging.Therefore,factors that are important factors that influence the evacuation of crowd are studied,such as vision,leader,panic and obstacle,and a crowd evacuation model is established.Furthermore,the crowd evacuation behavior and crowd evacuation characteristics in buildings are studied,which can effectively organize crowd evacuation and reduce casualties in emergency situation.A Multi-Agent evacuation model for buildings considering leaders and panic is proposed.The evacuation space in buildings is divided and a real evacuation scene is established;To truly reflect the movement of evacuated individuals,the evacuation scene is further abstracted,and a continuous evacuation space scene model is established;The attributes and behaviors that affect individual evacuation are analyzed and summarized,based on the modeling of evacuation individual attributes,perception,decision-making and behavior,the Agent model of evacuation individual is established;Considering the interaction among different individuals,individuals and environment,the evacuation crowd is modeled,and the Multi-Agent crowd evacuation model is established.An improved Particle Swarm Optimization algorithm for evacuation of buildings is proposed,which is used as an individual route selection strategy in evacuation process.The evacuation scene in the building is divided into several regions,and the global optimal individual selection strategy of the evacuated individual is optimized,considering the influence of the individual’s vision and location on the path selection;For individual panic psychology,the quantitative description of panic and the individual speed update mechanism under panic mood are given to show the influence of panic factors on individual evacuation;The Social Force Model is used to solve the problem of interaction among individuals,between individuals and obstacles,so that the improved Particle Swarm Optimization algorithm can better meet the actual situation of crowd evacuation in buildings.The crowd evacuation simulation experiment in buildings is designed to simulate the crowd evacuation behavior in buildings under emergency conditions.The effects of visual field and herd behavior,exit barriers and leaders on crowd evacuation are studied,and the phenomena of emotional infection and emotional attenuation in evacuation process is simulated.The authenticity of the Multi-Agent evacuation model for building is verified,with the analysis of experimental data,and suggestions are put forward for crowd evacuation in buildings.