Study On The Influence Of Obstacles On Pedestrian Evacuation Efficiency At Exit

With the change of times,high-rise buildings and urban areas continue to expand,crowded public places frequently occur emergency incidents,seriously affecting the people’s property and life safety.The bottleneck effect at the exit of the building is the main cause of these accidents,and the bottleneck structure commonly exists at the exit of the emergency evacuation of public space.How to effectively relieve the congestion at the exit and evacuate the pedestrian quickly and orderly has become an urgent problem to be solved by the national public safety department.On the basis of previous studies,aiming at emergency evacuation scenarios such as fire and earthquake in public space,this paper studies the optimization performance of barriers on crowd evacuation efficiency at building exits.The conclusion of this study can provide relevant management departments with the design and optimization of public space facilities as well as the theoretical basis in the field of emergency evacuation management.Based on previous studies on pedestrian evacuation of obstacles,this paper uses Pathfinder to simulate the simulation model of single exit to study the influence mechanism of different obstacle conditions on pedestrian arch at exit bottleneck.In order to systematically optimize the obstacle parameters,Matlab and Pathfinder simulation software were used,combined with BP neural network,Latin Hypercube sampling and particle swarm optimization algorithm,to establish a BP-PSO neural network model to solve the optimal obstacle parameter combination.Finally,the algorithm is applied to real residential buildings,and the obstacle can optimize the evacuation performance.The main research work of this paper is as follows:(1)Firstly,the influence of obstacles on pedestrian arch at exit bottleneck is studied.By comparing the formation conditions of earth arch effect and pedestrian arch effect,a method of adjusting earth arch effect derived from civil engineering is put forward.The obstacle is compared with anti-slide pile to adjust the position of pedestrian arch.Based on Pathfinder simulation software,four working condition models were established: no obstacle,square obstacle,transverse barrier obstacle and longitudinal fence.Through quantitative analysis of evacuation time of different initial evacuators and qualitative analysis of pedestrian path diagram and instantaneous density diagram,the influence mechanism of obstacles on pedestrian arch at exit was obtained.The results show that obstacles affect the formation of the arch in three ways: the deformation of the pedestrian arch,the movement of the arch center,and the separation of the space around the arch.(2)Secondly,the Latin Hypercube sampling toolbox in Matlab is used to obtain obstacle parameter samples,which are used for PSO-BP neural network prediction model to solve the optimal obstacle parameters.Since there is no clear proportional relationship among the length,exit spacing and width of obstacles,in order to solve the optimal obstacle parameter combination,the particle swarm optimization algorithm(PSO)was introduced on the basis of BP neural network to establish the optimal obstacle parameter combination prediction model.The network generalization ability is evaluated by the output expectation of Pathfinder software,which proves that the model has the ability of fast convergence,high precision and global search excellence.Finally,the optimal obstacle parameter combination corresponding to the minimum evacuation time is obtained from the output data of the neural network model.The results show that the PSO-BP neural network model has the characteristics of high accuracy,good fitting degree,fast convergence rate and powerful global optimization ability,simplifies the calculation of coupling effect between obstacle parameters,and can provide the optimal obstacle design scheme for building designers.(3)At last,an example of evacuation efficiency optimization of barriers on the exit bottleneck of residential buildings is studied.Through field research,residential buildings in a village in the city are selected as the research model.The Latin Hypercube sampling method and PSO-BP neural network model were used to calculate the optimal obstacle parameters,and the evacuation time,instantaneous density map and pedestrian arch were analyzed by quantitative and qualitative methods.It is further verified that the optimal layout of obstacles can improve the evacuation performance at the exit bottleneck.It is verified that the research results of this paper can realize the layout optimization of evacuation space and provide emergency evacuation adjustment strategies for actual public places.