Research And Implementation Of Spatial-temporal And Safetyaware Route Planning Method For Urban Travel

With the development of urbanization,the internal environment of cities has become more and more complicated,and the issue of urban travel safety has become the focus of common concern for city managers in all countries.From the perspective of safe travel,there are many unstable factors in cities,such as criminal incidents or traffic accidents.People should try to avoid the influence of these factors during travel to ensure travel safety.Therefore,the study of urban route planning methods for safe travel scenarios has important practical significance.The goal of path planning is to plan a path for users to meet their needs.The path planning to achieve safe travel can be divided into two steps.First,it is necessary to determine whether each area in the city is safe,and then plan the travel route according to the time and space safety of the area.Existing work has carried out research on these two aspects,but there are still shortcomings: in the characterization of regional risk,the existing methods mostly focus on the temporal and spatial correlation of public safety events between regions,ignoring the sparseness of public safety events in the region.In addition,the uneven distribution characteristics make it difficult to accurately determine high-risk areas.In terms of safe path planning algorithms,most existing methods focus on planning a path from the starting point to the endpoint,without considering the constraints of intermediate nodes,which cannot satisfy the user’s travel route.How to plan a safe path while meeting the individual needs of users is a more complicated problem,and requires better path planning algorithms.In response to the above problems,this thesis builds an urban regional risk prediction model based on graph neural network technology,and designs a corresponding path planning method for the personalized travel needs of users in safe travel scenarios,which specifically includes three parts:(1)Aiming at the problem of ignoring the temporal and spatial distribution characteristics of public safety events in current research work,a regional public safety event temporal and spatial distribution modeling method was designed to more comprehensively describe the temporal and spatial characteristics of security event data in the region,and a time-sensitive city was designed Regional risk prediction model to improve the level of regional risk prediction.(2)Because of the lack of path planning algorithms to meet the personalized travel needs of users in current research,a safety-sensitive city personalized path planning algorithm is designed to meet the personalized travel needs of users while ensuring path safety and optimizing paths Travel costs.(3)Based on theoretical results,designed and developed a safe travel-oriented urban path planning system.The system adopts a front-end and back-end separation mode.The back-end is responsible for managing data,predicting regional risks,planning paths and the front-end is responsible for providing user interaction interfaces and path visualization functions.To sum up,this article faces the problem of safe travel in urban scenarios,designs urban regional risk prediction models and urban route planning algorithms,and builds an urban route planning prototype system for safe travel scenarios.Experiments show that the proposed method can effectively predict the risk status of urban areas and plan safe travel routes for users.The proposed theoretical mechanism and prototype system proposed are helpful to construct urban safe travel services,which can be further applied to many fields such as urban perception and intelligent transportation.