Research On Data-driven Balance Control Of Queue Length At Urban Road Intersections

With the increasing number of social motor vehicles year by year,the problem of urban road congestion is becoming increasingly serious.Intersection control is one of the important ways to alleviate traffic pressure.At present,the intersection control of urban roads is mainly based on different forms of single-point timing and accurate traffic flow model.Under this control condition,if the traffic flow increases sharply,there will be a large difference in vehicle queue length at each phase.Therefore,it is necessary to consider the balanced control of queue length at multi-intersection of urban roads to improve the efficiency of traffic network.Queue length equalization control is a method to adjust the queue length by adjusting the signal timing method,so as to solve the problem of queue length equalization at multiple intersections.In addition,most of the control methods are based on accurate road traffic flow model,which is difficult to adapt to the characteristics of multi-intersection road network system,such as time-varying,strong nonlinear and periodic.Therefore,this paper proposes the research of queue length data-driven equilibrium control for multi-intersection of urban roads.The main work of this paper is as follows:Firstly,considering the factors of the length between the points and the degree of vehicles in the multi-intersection system,a multi-intersection data flow model is established.The communication between MATLAB and VISSIM is realized based on COM interface,and the co-simulation platform of MATLAB-VISSIM is established.Secondly,based on the queue length of each intersection generated by VISSIM,the model free adaptive control(MFAC)is proposed with the queue length balance as the control objective and the green time length of each phase as the control input.The queue length balance control scheme for multiple intersections of urban road is proposed and the algorithm is verified by co-simulation platform.Thirdly,for the intersection traffic system which has the characteristics of repeatability,strong nonlinear and uncertainty,there are some problems such as modeling difficulties,learning gain settings and so on.Based On the introduction of Model Free Adaptive Iterative Learning Control(MFAILC)method,the queue length equalization control scheme at multiple intersections is studied.The feasibility of the proposed scheme is verified by the MATLAB-VISSIM co-simulation platform.Fourthly,based on the above control method,considering the problem of data packet loss caused by communication in multi-intersection control,the queue length equalization control of multi-intersection MFAILC under data packet loss is proposed.Firstly,a controller is constructed using only the input and output data of the system under packet dropout.Secondly,considering the impact of data loss on the system control performance,a data loss compensation algorithm is further designed.Finally,MATLAB-VISSIM simulation platform is used to verify the effectiveness of the algorithm.