An Eligibility Traces based Cooperative and Integrated Control Strategy for Traffic Flow Control in Freeways

Abstract

Traffic congestion and gridlocks are considered as main problems of designing an urban motorway network. For this purpose, traffic flow control strategies are presented through recent decades to address this problem. In this paper, an Eligibility Traces based Reinforcement Learning (ETRL) traffic flow control strategy was proposed. This strategy is based on cooperative and integrated control of Ramp Metering (RM) and Variable Speed Limits (VSL). To test the proposed method, first the traffic macroscopic model was calibrated via Genetic Algorithm (GA) optimization to simulate traffic behavior and further, the traffic control strategy is applied to M62 highway stretch in England which is one of the smartest highways, and the results are presented.