Abstract: According to the operation problem of public bicycle systems, this paper focuses on proposing a public bicycle partition scheduling method. Based on improved ant colony algorithm, a public bicycle partition scheduling model with lowest cost and biggest traffic volume was established. The bicycle scheduling between sites was simulated and the impact of different factors was analyzed. The data of Danyang public bicycle system was taken as an example to verify the improved ant colony algorithm model. The result proves that the algorithm can solve the optimal scheduling loop with a high accuracy and high solution efficiency.