Research On Cooperative Control Method And Simulation Of Queuing Length Consensus At Multi-intersections

With the development and expansion of major cities in China,the structure of urban traffic network is also expanding.The fast time-varying,multi parameter and strong nonlinear characteristics of urban traffic system also make the establishment of accurate traffic flow mechanics model more and more complex.The control method based on data-driven is to design the traffic control strategy using only the input and output parameters of the traffic system.Therefore,it can solve the problem of limitations caused by many modeling constraints,which are only applicable to specific scenes.At the same time,with the change of network topology and the distributed characteristics of the system,multi-agent system is more and more closely connected with intelligent transportation system.It has achieved good results in solving a series of problems such as strong nonlinearity and non-stationary of urban transportation system.The purpose of this paper is to combine multi-agent system with data-driven control,to work together on urban road traffic system,and to achieve adaptive cooperative control of traffic signals at single and multi-intersections.The main contents of this paper are as follows:1.Aiming at the problem of unbalanced queuing length of each phase in the fixed cycle control of four phase single intersection,the coordinated control model of multiagent intersection is constructed by the communication of intersection signal controller.The global coordination error and tracking target are defined,and a distributed datadriven adaptive cooperative control method is proposed.The validity of the method is verified by MATLAB simulation and comparison experiment.2.Aiming at the problem of mismatch between the time of dynamic change of variable approach lane attributes and the change of traffic flow,the conditions of setting variable approach lanes are analyzed.The saturation of traffic flow in each direction and the queuing coordination error are defined to realize the adaptive coordination control of intersections with variable approach lanes.The rationality of the proposed method is verified by comparing with the single intersection queuing balanced control method.3.To solve the problem of global queuing length imbalance in multi-intersections variable cycle control,considering the correlation between the arrival traffic flow in different directions and the traffic flow at adjacent intersections,a queuing length model for multi-intersections is proposed.According to the global coordination error of queuing length in each direction of each intersection,a data-driven adaptive cooperative control strategy is designed for multi-intersections.Finally,the applicability of the improved multi-agent adaptive coordinated control method to multi-intersections is verified by simulation.