Research On Operating Characteristics And Dynamic Scheduling Optimization Of Public Bicycle Based On IC Card Data

The urban traffic problems have become increasingly prominent. The development of public transit system has become a consensus. However, the "last kilometer" problem restricts the service level of public transit system. To solve this issue, many cities established the public bicycle system,In actual operations, the imbalance of renting bikes hinders the healthy development of the system. Based on the National Key Basic Research Program (No. 2012CB725400), this thesis targets improving the operational efficiency. The traffic characteristics and site variation are explored based on IC card data. Further in-depth study focuses on site demand forecasting and dynamic scheduling optimization methods.Firstly, format optimization and data cleansing of the IC card data is completed. Rental features are summarized by statistical analysis from time and space dimensions. Public bicycle sites are divided into five classes to discuss the demand variation of different types according to cluster analysis.Secondly, on the basis of the suitability of prediction methods and features of variation, BP neural network time series prediction method is choosed to forecast site requirments. Time correlation is used to predict short-time demand of five types of instances.Furthermore, dynamic scheduling optimizational methods are discussed to of determine the possible solutions. Multi-objective constrained scheduling route optimization model is built based on the goal of low transportation costs and high customer satisfaction. The functional relationship between passenger satisfaction and time delay punishment is also established. According to the characteristics of the model, genetic algorithm is designed to solve the problem.Finally, Ningbo public bicycle system is taken as an example to testify the application of public bicycle dynamic scheduling problem. The results indicate that the model and algorithms used in this thesis applies to the public bicycle dynamic scheduling optimization. Besides, it also balances the contradictions between scheduling cost and user satisfaction and enhances the scheduling efficiency of public bicycle system.