Agent-based traffic coordination control method for two adjacent intersections

Agent technique constitutes a focus of distributed artificial intelligence. In this paper, this technique is applied to urban traffic control. Based on the single traffic signal control agent (TSCA) with Q-learning, the method combining game theory and social rule is applied to solve the coordination between two TSCAs. By comparing the method with and without coordination of two intersections in a simulated environment, the result indicates that the coordination method used in this paper is effective.