The basic theory of reinforcement and the idea of distributed Q-Learning are introduced in this paper. Based on the analysis of distributed Q-Learning in the urban traffic coordination control, reward function and weight function are presented. The e