Modeling Of Emergency Pedestrian Escape Considering Obstacle Avoidance Behavior And Information Guidance

In recent years,large-scale activities are held more and more in public places with the improvement of economic level,and the flow of people is especially dense.When earthquake,fire and other emergencies occur,what measures should be taken to make pedestrians escape quickly is an urgent technical problem to be solved.Therefore,the scientific formulation of relevant escape strategies is necessary for the smooth evacuation of pedestrians.Comprehensively exploring obstacles characteristics,this paper focuses on the modeling of pedestrian obstacle avoidance behavior and induced behavior in the process of emergency escape.The specific research contents are as follows:The improved social force model based on ant colony algorithm is used to study the movement behavior of pedestrians with/without global escape information in obstacle environment.Obstacle avoidance behavior,as an important part of pedestrian escape,can directly affect the overall escape efficiency of the crowd.According to whether pedestrians are familiar with the building structure,they are divided into pedestrians with only local escape information and pedestrians with global escape information.Considering obstacles,pedestrian preference,herding effect and other influencing factors,the desired direction of pedestrians with only local escape information is improved on the basis of social force model.Ant colony algorithm is used to modify the desired direction to navigate pedestrians with global escape information.Based on the improved social force model,the effects of obstacle avoidance behavior,herding behavior and field of vision on evacuation dynamic characteristics and escape ability are obtained.The results show that pedestrians with global escape information can avoid obstacles smoothly and can choose a shorter route to escape,which improves the escape efficiency.Pedestrians with only local escape information have curved trajectory and low escape ability,while herding behavior is conducive to the rapid escape of such pedestrians to a certain extent.An improved induced pedestrian escape motion model based on clustering algorithm is established to solve the optimization of the initial position and number of leaders,and the selection of multiple leaders in the process of emergency escape.As an important part to ensure pedestrian safety in buildings,the leader plays an key role in improving pedestrian escape efficiency.According to the cost function constructed by the number of induced pedestrians,non induced pedestrians and leaders,the number of leaders is determined.The location of leaders are optimized combined with K-means algorithm.The choices of multiple leaders for individuals are decided due to distance and number.Based on the improved induced pedestrian escape motion model,the relationship between the number of leaders,model parameters and escape ability is obtained.Then the cooperative improvement strategy of escape efficiency and leader optimal position is solved.Results indicate that more leaders can reduce escape time,but there is a maximum threshold for the number of leaders considering the human resource cost and induced coverage.In the process of the selection of multiple leaders,the weights of the distance and number need to be reasonably set according to the actual escape demand.