A hierarchical public bicycle dispatching policy for dynamic demand

The dispatching policy of public bicycles during peak hours largely determines the sharing rate of bicycles and service level of public bicycle system. Firstly, a dynamic demand prediction model of public bicycle system is proposed considering reservation data and historical data. Secondly, a hierarchical dispatching policy and network partitioning method are discussed. Thirdly, an optimization model with time window to decide the number of bicycles that need to be dispatched is put forward considering the cost of dispatching and waiting time of passengers, and the method to decide the path of dispatching is also discussed. Finally, a case study in Guangzhou shows that the proposed policy could provide lower waiting time for passengers and better bicycle sharing rate of bicycles.