Research On Road Signal Control Strategy Based On Traffic Flow Identification

Since the 21st century,China's urbanization process has developed rapidly,and traffic intersections have also shown continuous congestion during peak periods.It is imperative to ease traffic congestion.In the case of relatively fixed road infrastructure construction,a reasonable traffic signal control strategy can effectively alleviate congestion at traffic intersections,thereby reducing traffic delays.Aiming at the characteristics of traffic flow control strategies in the existing traffic signal control strategies,this paper uses neural network to classify and identify the traffic flow categories in actual intersections.Based on this,the paper proposes to distinguish the traffic flow patterns.Adapting to the road signal control strategy,the control strategy and the traffic flow category are accurately matched,which reduces the error caused by the artificial judgment of the traffic state subjectivity.The main contents are as follows:(1)Explain the basic mode,parameter model and key performance evaluation indicators of signal light timing control.(2)Aiming at the problem of dynamic information acquisition of traffic flow class,a window road model based on vehicle traffic characteristic information perception is established,and the information interaction process between vehicle network and road communication equipment in window road mode is described.Based on the model,key information such as traffic flow,lane and vehicle priority are acquired in the subsequent intelligent control process.(3)Based on historical statistical data,the traffic entropy method is used to pre-process the traffic data.The SOM neural network is used to cluster the traffic flow state,and the intersection of the intersection time segment and the road segment environment characteristics is realized to realize the scientific division ofthe traffic flow pattern.Based on the classification of traffic flow categories,the PNN neural network is introduced to train the traffic flow pattern of the intersection.Finally,the traffic flow data at different intersections of the intersection is input to verify the accuracy of PNN network identification.(4)The threshold polling service algorithm,Webster control algorithm and emergency traffic control strategy are analyzed.The traditional Webster signal timing algorithm is improved.A Webster timing method based on traffic flow pattern optirmization is proposed.Based on the analysis of the advantages of each control strategy and the characteristics of traffic flow,this paper proposes an adaptive control strategy based on traffic flow identification,and formulates a proposed control strategy table under time and traffic flow mode.(5)Taking the intersection of Juxian Street and Jingming South Road in Kunming as the test object,through the traffic data collection within 24 hours,based on the VISSIM platform,the actual road simulation model is built according to the actual situation of the intersection,and the signal timing scheme is selected as the research object.The delay of driving,the number of parking,and the queue captain as the excellent indicators for evaluating the timing scheme,completed the performance test and analysis of the proposed control strategy.