The coordination between traffic signal control agents based on Q-learning

Agent technique constitutes a new focus of distributed artificial intelligence (DAI) and its application has covered many areas. This technique is applied to urban traffic control area. The traffic signal control agent improves its control ability with the Q-learning method. A new method combining game theory and society rules was proposed to solve the problem of coordination between two TSCAs (Traffic Signal Control Agents). To test the efficiency of the coordination mechanism, a prototype traffic simulator was programmed in MS C++. In such a simulative environment, the methods with and without coordination of two crosses were researched. The result indicates that the new coordination method proposed in this paper is effective.