The Scheduling Optimization Study Of Public Bicycle

With the community economy developing, the number of vehicles in China is growing, urban traffic congestion is getting more and more significant. Developing public transportation to solve traffic problem has become the most effective mean. However, the coverage of subway station, BRT station and conventional bus stop is not enough. The "last mile" has become a bottleneck of improving the services of public transport. In that context, the public bicycle system breeds. With the advantages of flexible, convenient, cheap and eco-friendly, public bicycle has become an important travel option for local people. But with the influence of some factors, such as layout, characteristics of travel behavior, it is quite hard for people to lend and return bicycle during the rush hour. Therefore it is imperative to establish efficient public bicycle scheduling system.In order to achieve these goals, this thesis has proposed a thorough and comprehensive research in public bicycle scheduling optimization problem. Specifically, this thesis mainly studied the following aspects:(1) On the basic of reading a large number of domestic and international literatures, this paper firstly summarized the applications and theoretical study of current public bicycle system. The function, scheduling and process of public bicycle system were analyzed, and key problems of public bicycle scheduling were researched emphatically.(2) Based on the data of public bike system in Zhongshan from IC card, this thesis analyzed characteristics of usage of public bicycle, such as lease number, lease time, length of lease. Also using these data, characteristics of demand and the related influence factors were studied.(3) This paper set up the single-depot PBS scheduling optimization model. And on the basis of this model, the dynamic-demand Scheduling optimization model was proposed. Two models meet optimization objectives of shortest scheduling path and highest passenger satisfaction. It is helpful in the field of dispatching bicycle in practical work.(4) This paper made a full consideration to the characteristics of public bicycle scheduling optimization and used genetic algorithm to solve the model in Matlab. The model and algorithm were verified by public bike system in Huoju development zone in Zhongshan.