Development Of Trajectory Flow Statistics And Visualization Software For Traffic Network

With the increasing scale of urban construction,the growth of urban population,the traffic travel information has also increased,and the large number of self-driving cars has exerted tremendous pressure on the urban traffic network.Therefore,the key to solve the pressure of urban traffic network is to observe the characteristic of these collected traffic data and analyze the law of these data hiding.According to traffic conditions and urban development status,this paper designs and implements a traffic road network that counts trajectory flow and develop visualization software system.The collected traffic flow data is processed to realize the trajectory flow statistics,and the results of trajectory flow statistics are displayed by visualization technology,the travel rules and traffic conditions of the vehicle are summarized and analyzed,and the instantaneous trajectory flow period division algorithm is used to solve the difficulty of dividing the peak-valley period.It is rich in meaning to help understand the current situation of urban traffic and improve road network traffic.The main work and achievements of this thesis are as follows:1.Firstly,the Dijkstra algorithm is introduced to lay the foundation for the subsequent trajectory flow statistics algorithm.The mathematical description of the road network is based on the graph theory.The trajectory flow statistics consists of the definition of trajectory and trajectory flow and the flow statistics algorithm.The trajectory flow visualization consists of four parts: geographic coordinate projection,scale method,visual layout algorithm and trajectory flow direct line drawing method,the instantaneous trajectory flow time division algorithm is based on Gaussian regression algorithm to solve the difficulty of peak-to-valley time division,and can accurately reflect the trend of traffic change and improve the effectiveness of scheduling.This chapter lays a theoretical foundation for the trajectory flow statistics and the realization of relevant modules in the visualization software system through the design of trajectory flow statistics and visualization algorithms.2.Based on the proposed theoretical research,through the analysis of the software system,including system requirement analysis,system feasibility analysis and system basic function analysis,the trajectory flow statistics and visualization software system design.The system includes modules for trajectory flow statistics,trajectory flow visualization,instantaneous trajectory flow period division,data management,and system management.The trajectory flow statistics module processes the actual traffic flow data,the trajectory flow visualization module visualizes the results of the trajectory flow statistics,the instantaneous trajectory flow period dividing module performs Gaussian process regression processing on the instantaneous trajectory flow data to obtain a smooth regular instantaneous trajectory flow curve and classification of peak and valley periods,the data management module provides actual traffic flow data,and the system management mainly stores the records of the system operation and the situation of daily user use.3.Using the implemented system,the road network composed of nine intersections in Jiaojiang District of Taizhou is used for practical application,and the trajectory flow statistics and visual analysis are carried out.Based on the actual traffic flow data,the statistics of all-day,early-peak,late-peak,flat-peak trajectory flow and instantaneous trajectory flow are sequentially performed,and then the trajectory flow is visualized.According to the results of trajectory flow statistics and the visualization of trajectory flow,the trajectory flow in the road network is analyzed in detail,and the rules of self-driving travel behavior are summarized,and some suggestions for vehicle travel strategies and residents' travel are put forward.Finally,the results of the instantaneous trajectory flow period division are described.