According to the operation data of the public bicycle system, we propose an accessibility index evaluation model for the complex network. We analyze the topology characteristics of the complex network, evaluate the accessibility indexes for traffic functional division, and then select the key nodes in the network, which can provide theoretical basis for optimizing the public bicycle system. The overall topology analysis shows that the user demand model of public bicycle is scale-free with the characteristics of a small world, and a small part of the hub node in its network distribution can be further optimized. According to the short distance travel demand and long distance public transport network, key nodes can be filtered out by calculating reachability indicators. Furthermore, the actual flow and potential indexes of key nodes can be compared and analyzed, and more accurate optimization suggestions for key nodes are obtained. Actual data analysis shows that the potential evaluation model based on the accessibility indexes can provide a theoretical basis for the public bicycle to play a better  role in public transport.