Research On Location Model Of Urban Public Bicycle Scheduling Centers

With the advantages of large capacity, high efficiency, low cost and low pollution, urban transit has become the inevitable choice to ease urban traffic congestion, bus priority is widely acknowledged by state and local governments. Under the guidance of the development of multi-modal urban transit, public bicycle system (PBS) is attracting increasing interest. On the one hand, PBS is ideal for urban short-distance travel; on the other hand, PBS can be connected to the public transportation with large capacity, to solve the’last mile’problem of residents travel. However, the unbalanced borrowing and returning requirements at peak hours, makes it easier to have problems of’no bicycle can be borrowed or returned’, and severely reduced PBS’s service level. Thus, in the critical period of PBS’s rapid development, research on public bicycle scheduling centers’location problem is of great theoretical and practical significance, which is beneficial to enhance the service ability of the public bicycle scheduling system, and improve PBS’s service level.Firstly, PBS’s model of static scheduling is built. Each dynamic event type has been analyzed to determine the corresponding trigger mechanism of scheduling plan adjusting dynamically. Focus on the newly increased scheduling demand, mathematical model of dynamic scheduling is built.Secondly, departing from the regional scheduling mode of PBS, basic location model of public bicycle scheduling centers has been built, based on the analysis of traditional siting problem and vehicle allocation routing problem. This model considers scheduling centers’ construction costs, scheduling vehicles’ daily use-costs and transportation costs. Fuzzy processing the service time windows of public bicycle rental points, penalty function of fuzzy time windows is proposed.Thirdly, methods to determinate service time windows and scheduling demand are proposed, after the sitting of public bicycle rental points’parking rates safety threshold. Uniting the basic location model and the penalty function, location model of public bicycle scheduling centers is built based on location-routing problem, with the goal of minimum costs. On the basis of simple genetic algorithm, a hybrid adaptive genetic algorithm is designed, importing self-adaptive strategies and simulated annealing algorithm.Finally, based on the operational data of Jiangning District, study period and key objects are determined. After setting model parameters, traditional method and present model are used to solve the case with the help of MATLAB. Analysis results show that the present model has obvious advantages.