Calibration Of VISSIM Simulation Parameters Based On Aerial Video

The research and development of Intelligent Transportation System(ITS)plays a good role in improving the current serious traffic congestion,frequent accidents,increasing environmental pollution and other major issues.Its two key research areas are traffic parameter extraction technology research and micro traffic simulation system research.In the research field of traffic parameter extraction technology,Unmanned Aerial Vehicle(UAV)has the advantages of low cost,small occupation volume,large coverage,less occlusion interference and so on,which is loved and supported by more and more intelligent traffic researchers.In the research field of micro traffic simulation system,parameter correction of traffic simulation model can improve the simulation modeling accuracy and make the established traffic model highly representative.Aiming at the video of traffic status at the intersection captured by UAV,and using the COM interface technology of VISSIM,the methods of vehicle detection,traffic flow parameter extraction and VISSIM parameter correction for UAV video are studied.Based on pycharm development platform,opencv computer vision library is called by Python to complete the extraction of traffic flow parameters of video data.Aiming at the shortcomings of the existing virtual detection line detection method,a combined virtual detection line method suitable for intersection is proposed to count the number of vehicles turning in different directions in aerial video of intersection According to the actual cases,it is proved that the statistical method can ensure the accuracy of traffic flow statistics.Then the statistical results are transferred to VISSIM micro traffic simulation model by computer programming,and the model parameters are corrected by genetic algorithm.Finally,taking the intersection of Xueyuan Road and Changning Road in Tangshan City as an example,a micro traffic simulation model is established,and then parameters are corrected.The research results show that the delay error is reduced by 9.47%,which proves that the method in this paper is feasible.Figure 18;Table 24;Reference 72...