Research On Crowd Congestion Evacuation Method Based On Knowledge Graph And Gravity Field

With the material and cultural progress of today’s society,large-scale crowd gathering activities are becoming more and more frequent.When an emergency occurs,disorderly evacuation not only leads to inefficient evacuation,but also makes it very easy for congestion,stampede,and other malicious accidents to occur.Crowd congestion is an important reason for reducing evacuation efficiency,and avoiding congested areas and planning efficient evacuation routes for the evacuating crowd are effective means to improve crowd evacuation efficiency.The traditional response is to conduct crowd evacuation drills,but this drill method will waste a lot of cost and time,and it is difficult to plan the best escape route through human resources.To overcome the shortcomings of traditional evacuation drills and provide scientific guidance for the crowd evacuation process,academia and government departments are paying more and more attention to crowd evacuation simulation technology.This technology plays an important role in preventing public safety accidents and safeguarding people’s lives and properties by exploring the law of pedestrian movement in the form of simulated crowd evacuation process.Therefore,it is of great significance to study the scientific evacuation problem when crowd congestion occurs through computer simulation technology.Crowd congestion is an important factor affecting evacuation efficiency,and reasonable regulation of crowd congestion in the evacuation process is a means to improve crowd evacuation efficiency.In this thesis,we focus on how to improve crowd evacuation efficiency by avoiding large-scale crowd congestion.Since the traditional crowd evacuation simulation method is based on hypothetical scenarios and rules,it makes the crowd evacuation simulation lack of realism.To solve this problem and reduce the scale of crowd congestion during evacuation,this thesis proposes a crowd congestion evacuation method based on knowledge graph and gravity field.Firstly,this thesis extracts the information of real scenes in the video through sensors to build a model of evacuation scenes,to improve the realism of evacuation simulation.Secondly,an area state knowledge graph(ASKG)is constructed to represent the congestion information in the evacuation scenes through real evacuation data,which improves the ability of the model to characterize the crowd information.Then,the gravity field is generated based on the information in the knowledge graph,and the crowd is evacuated under the guidance of the gravity field.Finally,the constructed area state knowledge graph is used to predict the next moment of congestion and regulate the evacuation route of the crowd in real time to reduce the occurrence of crowd congestion.The experimental results show that using sensors to extract real scene data can improve the realism of crowd evacuation simulation.At the same time,the introduction of knowledge graph can improve the characterization ability and prediction effect of the model,and the model in this thesis avoids large-scale crowd congestion and improves the efficiency of crowd evacuation.The main work and innovation points of this thesis are as follows.(1)To address the drawback that the existing crowd evacuation simulation models are based on hypothetical evacuation scenarios,which leads to the lack of realism in the simulation process.In this thesis,we propose a knowledge graph construction method based on real data to represent various heterogeneous data in evacuation scenarios in the knowledge graph,to improve the model’s ability to characterize real data.Firstly,this thesis adopts YOLO V3 method to detect pedestrians in the video and extract the pedestrian positions and motion trajectories in the real crowd evacuation video.Secondly,the evacuation information is combined with the knowledge graph to establish the pattern layer of the area state knowledge graph.Then,the area state knowledge graph is constructed according to the pattern layer,which effectively preserves various information in the process of crowd evacuation.Finally,it is experimentally verified that this thesis introduces the knowledge graph into the crowd evacuation field,which deeply improves the characterization ability and evacuation efficiency of the model.(2)To address the problem of lack of efficiency of existing crowd evacuation methods.In this thesis,we propose a gravity field path navigation method based on knowledge graph.The method introduces the concept of gravity field,generates gravity field according to the crowd and environment characteristics,predicts the congestion at the next moment by using the constructed knowledge graph of the area state,avoids the crowd congestion in advance and combines the gravity field navigation model with real-time path planning.The experimental results show that the proposed method can guide the crowd to choose the optimal evacuation route,avoid a large number of congestions and improve the efficiency of crowd evacuation.(3)Based on the above research,this thesis adopts a crowd evacuation navigation simulation system based on knowledge graph and gravity field to simulate and guide the evacuation of pedestrians.The system consists of simulation control,camera control and rendering output.The system is applied to carry out evacuation simulation experiments in the lobby of the teaching building of Shandong Normal University and the real model of Quancheng Square.The simulation experiments can study the crowd movement process more intuitively,increase the realism of the simulation effect,and have guiding significance for the emergency crowd evacuation in safety accidents.