Approximation Dynamic Programming for multi-intersections traffic signal intelligent control

Traffic congestion is ubiquitous in the modern city, and optimizing the behavior of traffic lights for efficient traffic flow is a critically important goal. In this paper, an urban road traffic area-wide coordination control algorithm based on Approximation Dynamic Programming is proposed. The agent model of traffic intersections is demonstrated. The algorithm substantially reduces average vehicular delay by using an approximation to the value function of the dynamic programming to update the approximation, and can cooperative control of multiple intersections to achieve a near optimal control policy. The computer simulation results show that the control algorithm can effectively reduce the average delay time and play a very good control effect with multi-intersections, so the coordination method used in this paper is effective.

I. Introduction

Intelligent Transportation Systems is an effective way to solve transportation safety and traffic congestion, while the urban transportation system is an important of the intelligent transportation. As it has multiple information sources, time variability and randomness characteristics, so it is difficult to control them effectively with traditional theories and methods [1]. Therefore some advanced control theories and intelligent methods applied to urban traffic control. For the cooperative traffic signal control, the green wave band method has been studied firstly [2]. But this method asks that the traffic flows keep uniform speed so that a better control solution can be achieved, so it is difficult to be used in the traffic networks.