Study On Location Selection And Dispatching Model Of Urban Public Bicycle Rental Points

With the advantages of flexibility,convenience,low price,green and low carbon,the public bicycle system has rapidly integrated into the urban traffic system and become an important part of it.However,in the process of development of public bicycle system,the unbalanced borrowing and returning demand of users in time and space leads to the phenomenon of "no car to borrow,no pile to return" in some leasing points,which directly reduces the satisfaction of users and the overall utilization rate of vehicles,and indirectly affects the operational efficiency of public bicycle companies.Therefore,this paper will take Nanjing public bicycle system as an example,explore the using characteristics of the public bicycle system,analyze the reasons for the unbalanced borrowing of the public bicycle system,and establish a scientific and reasonable location model and scheduling model to provide new ideas and methods for solving the unbalanced borrowing and returning problem.This paper elaborates the definition of urban public bicycle in detail,summarizes the system framework of public bicycle system,analyses the function orientation of public bicycle system,and classifies the leasing points according to the surrounding land properties.Secondly,the evaluation model of public bicycle rental point location scheme is established,and the combination weighting method is formed by combining the subjective weighting method with the objective weighting method based on the game theory,which alleviates the unbalanced borrowing and returning problem of public bicycles from the spatial location.Thirdly,this paper predicts the borrowing and returning demand of public bicycle rental points by using the random forest algorithm with sliding panes.Based on the loan demand forecast of the leasing point,a two-stage scheduling model with Hub-and-Spoke network architecture is established.The first stage is Hub process which forms the central leasing point,and the second stage is Routing process of the optimal demand path.Finally,taking the data of public bicycle rental points in downtown Nanjing as an example,the dispatching model is validated and analyzed.The main achievements of this paper include: introducing the definition,system architecture,function orientation and type of leasing point of public bicycle system through the research of existing papers;establishing a combination Weighting Method of subjective and objective weighting method based on game theory,and establishing a micro-level evaluation system of leasing point location scheme;a random forest prediction algorithm based on pane sliding is established.A two-stage static scheduling model based on Hub-and-Spoke network architecture is established by studying Hub-and-Spoke network,and the ABC algorithm is used to solve the scheduling model.