Traffic State Identification And Key Points Mining Of Urban Road Network Based On Floating Car Trajectory Data

Traffic congestion has become one of the main factors affecting the quality of residents’ life and the bottleneck of economic development of the city.Traditional means such as expansion of traffic capacity is unable to meet the growing traffic demand anymore.How to make full use of the obtained traffic data to accurately grasp the traffic congestion situation of the overall road network has become a new way to solve the problem of urban traffic congestion.Based on the floating car data in Chengdu,this paper studies the identification of key nodes of the road network and identify the state of urban traffic congestion,which provides theoretical reference for urban traffic congestion control and management decision-making.The specific research contents are as follows:(1)First of all,based on the floating car data,the data preprocessing process is adopted to realize the time conversion,coordinate system correction and traffic feature recognition of the original track data.And the feature mining technology is applied to extract the traffic speed features from the position information of the original track sequence.Besides,the data set of November 1,2016 is taken as an example to implement the all-day track travel in the research scope,and the characteristics analysis of the differences of the characteristics of traffic travel in the study area in time and space distribution are fully excavated.(2)Then,by introducing the H3 spatial index grid system,a spatiotemporal index model of traffic characteristics within the research scope is constructed,which provides a reasonable spatiotemporal segmentation method for traffic feature set analysis.The DBSCAN(density based spatial clustering of applications with noise)clustering algorithm is applied to realize the spatiotemporal clustering of traffic state in the scope of global road network,and the spatiotemporal analysis of typical traffic congestion modes is implemented based on the clustering results,which provides a full conclusion reference for mastering the spatiotemporal distribution and congestion characteristics of traffic state in the scope of global road network.(3)Furthermore,the key node identification model of urban road network is constructed.Firstly,according to the previous research,the definition of key nodes is proposed.Then,based on the directed weighted complex network,the urban road network model is constructed.Furthermore,based on the static characteristics of complex network,the evaluate index system of key nodes of urban road network is constructed from the aspects of road network topology and traffic attribute.And the four indexes of node betweenness,node efficiency,node congestion degree and node traffic weight are selected.Finally,the algorithm of key nodes identification is proposed.(4)Finally,taking the local area of the second ring road in Chengdu as an example,the identification analysis of key nodes is carried out.Firstly,the general situation of the research area and the road network matching algorithm based on Arc GIS are introduced.Then,the data of November 1(Monday)and November 6(Saturday)are selected to carry out the dynamic identification analysis of key nodes in peak hour and off-peak periods.By analyzing the distribution of key nodes in peak hour and off-peak period on weekends and weeks,the distribution characteristics of key nodes are summarized.And the management suggestions for key nodes are put forward.