Research On Visual Analysis Method Of Vehicle Travel Pattern Based On Historical Driving Track Set

With the increasing number of vehicles in the city,it is increasingly urgent for the traffic management department to carry out fine management of vehicle traffic.Mastering the travel behavior characteristics of vehicles is an important prerequisite for implementing fine management.The regularity of vehicle activity is one of the important behavior characteristics of vehicle,which has an important reference value for improving the level of vehicle traffic management.Based on the urban intelligent traffic checkpoints equipment,a large number of historical driving track data can be obtained.Therefore,it is of great significance to study how to mine the potential vehicle travel patterns in the track data to solve urban traffic congestion and ensure traffic safety and flow.The degree of confusion in the historical driving track of vehicles can describe the regularity of vehicle activity.Therefore,this dissertation studies how to mine the characteristics of vehicle driving pattern from the confusion degree of vehicle track according to the checkpoint track data.The main research contents are as follows:(1)Research on the visual analysis method of vehicle behavior based on trajectory entropy.Firstly,this dissertation proposes to map checkpoints and trajectories into words and sentences respectively,and calculate trajectory similarity by sentence semantic similarity.Secondly,based on the track similarity,the trajectory entropy is proposed,which measures the regularity of all trajectories of each vehicle.Finally,the behavior characteristics of vehicles are analyzed based on trajectory entropy.In order to facilitate users’ s further analysis,this dissertation provides a visual analysis system containing multiple linked views,which allows users to observe and compare the trajectories and trajectory entropy of different vehicles.Combined with clustering analysis and correlation interactions,it can help users to discover meaningful vehicle behaviors patterns.(2)According to the characteristics of multi-temporal granularity vehicles,this dissertation proposes an interactive clustering visualization analysis method based on vehicle patterns of subspace exploration.Firstly,the time dimension of vehicle trajectory was divided into fine granularity,and the vehicle trajectory set was segmented and quantitatively grouped.Trajectory entropy is used to measure the chaotic degree of vehicle trajectory in different time periods,so as to characterize the multi-dimensional characteristics of vehicles.Secondly,in the visual analysis system,the scatter diagram matrix and subspace search process are used to help users explore meaningful patterns in multi-dimensional data.