Research On Optimal Timing Management Of Intelligent Traffic Lights Based On Non-parametric Analysis

The optimal matching of signal lights is directly related to the quality of intelligent signal control system,not only involves the problem of traffic congestion,but also involves the life safety of the masses,which is a major livelihood problem.In recent years,intelligent traffic related research has become the main issue of the "two sessions",which has aroused wide concern of the Government,enterprises and the public,and the related research of intelligent signal lights,as a sub-subject of intelligent traffic research,has also been highly valued by all parties in the society.As an effective means to solve traffic congestion,how to optimize and match the intelligent signal lamp effectively,to find a more reasonable and accurate prediction method has become the focus of research.In view of this research,this paper combs the traditional timing method of signal lamp optimization and obtains the common disadvantages of these methods: the traditional method model is based on some hypotheses,the result is only the ideal result,and there is a big gap with the real situation.This paper starts with these hypotheses,looks for the timing model that cannot be made any hypothesis,summarizes the characteristics of non-parametric analysis through the in-depth study of non-parametric analysis,and puts forward the application of this method to the study of signal optimization timing,and constructs a kernel regression prediction algorithm based on European distance matching search.Based on the empirical analysis of the real traffic volume at the cross intersection of Dongxihu District Avenue and Siming Road in Wuhan City,Hubei province,and then evaluating and analyzing the optimized timing scheme,it is concluded that the average delay time of the vehicle passing through the intersection under the optimized timing scheme is 40.85% lower than that under the original scheme.The average queuing length is 48% lower than the original scheme and the road capacity is increased by 11.84%.Finally,the traffic volume of December 18,2018,07:12—08:24 is selected,and the average delay time,vehicle queuing length and road capacity of the other 5 different methods mentioned in this paper are calculated,and through the comparison of the data,It is further explained that the optimization effect of signal timing under non-parametric analysis method is much better than that of these traditional methods.Through the comparison of the data,it is further explained that the optimization effect of signal timing under the non-parametric analysis method is much better than that of these traditional methods.Firstly,this paper expounds the research background of signal lamp optimization,the research status at home and abroad,and the related concepts of signal timing.Secondly,the development history of signal lamp and the world mainstream signal optimization technology and the construction of signal lights at home and abroad are introduced.Then the classical signal timing method is introduced and summarized,the shortcomings of each method and the common problems are pointed out,and the paper puts forward the viewpoint of applying the non-parametric analysis method to the optimal timing study of signal lamp and gives the advantages of the application of non-parametric analysis method.Then,the concrete steps of non-parametric analysis are given,and the actual traffic data of Dongxihu District Avenue and SiMing Road cross intersection in Wuhan City are analyzed empirically,and the validity and feasibility of this method are proved.Then the situation analysis of the optimization and timing development of traffic signal lights in China is carried out,according to this analysis to provide guidance and suggestions for its future development.At the end of this paper,the summary and the next prospect of this study are summarized.