Research On Topological Reconstruction Of Road Network In Residential Area Based On GPS Data Of Slow Traffic

Electronic map is an important foundation of its.At present,the production of electronic map is mainly made of artificial mapping and satellite remote sensing image after processing.This kind of traditional method needs a lot of human and material resources,and the production cycle is long,and the update is slow.With the rapid development of 5g communication and the Internet of everything,the massive GPS data generated by vehicle GPS system and mobile GPS terminals,such as mobile phones,contains the spatial information of urban roads.The use of GPS data mining and extraction of urban road network information has become a hot issue in recent years.At present,scholars at home and abroad mainly rely on floating car or taxi data to extract road network by GPS.However,due to the requirements of vehicle on road conditions,it is difficult to obtain part of path and branch information by using floating car GPS data alone,which makes the extracted road network information incomplete.Therefore,this paper proposes a road topology generation method based on slow traffic GPS data.The main work is as follows:Firstly,the characteristics of GPS track data are analyzed,and the processing methods of stagnation point data,offset data and missing data are put forward,and the data preprocessing of Shanghai shared single vehicle GPS data is carried out.Then extract the network topology node.At present,most of the methods to judge turning point are by setting an angle threshold,which is easily affected by accidental data.In this paper,a concept of trend vector is introduced to describe the motion trend of the trajectory.The turning point of the trajectory is determined by the angle between the front and back trend vectors,After all the trend vectors are unitized,the minimum spanning tree algorithm is used to cluster,and then the import and export direction vectors are obtained,The topological points are determined by turning analysis.Then,the geometric alignment of the road is extracted.In this paper,the sub trajectory clustering framework is used to segment the whole track,cluster the sub trajectory and extract the feature points.Among them,the ow(opening window)algorithm is used to divide the track into sub tracks.Based on the DBSCAN algorithm,this paper defines the measurement method between sub tracks,proposes a clustering method for sub tracks,then scans the cluster to extract feature points,and finally uses B-spline curve to fit feature points to obtain road alignment.The topological structure of road network is constructed by combining the topological points with the road centerline.Based on the generated road network topology,Dijkstra binary heap is used to optimize the shortest path and the optimal path planning,in which the travel time extracted from the GPS track is the link impedance of the optimal path.Finally,the accuracy of the road network is verified by using buffers with different radii.The results show that the generated road can not only ensure the topological structure but also have high coverage.And the generated road network completed the application of path guidance.