Research On Location And Planning Of Bike Sharing System Based On Markov Chain Model

In order to promote green travel,reduce the proportion of vehicles,and generalize the use of bicycles,the bike sharing system built and put into use in different cities,while the domestic construction and operation has just started,both in theory and experience have some deficiencies.The imbalance phenomenon of borrowing and returning is likely occurred in peak period of bike sharing system,and generally use the factitious repositioning method to adjust,which cost a lots of government financial resources and ignoring the unreasonable bike station location and facilities scale also can lead to imbalance.This paper will focus on the location and planning of bike sharing system,analyze the main factors that affect the user choice behavior,weigh the pros and cons of conditions including the distance between demand points and rental stations,or each adjacent stations,the flow of bicycles and the lack in bike and lock piles.Increase the user riding choice behavior considerations in the general location and planning research to obtain the smallest cost location program,and to make a more reasonable design of station location and facilities scale distribution.It will make the demands of borrowing and returning more closer to balance,decrease the difference of borrowing and returning,and finally reduce the workload of repositioning.The main research contents include:(1)The study of bike sharing syetem user selection characteristics.It analyzes the characteristics of different gender,age,travel destination,and travel modes,also the influences of user selection by the characteristics of seasonal,time,land attributes,distance,which provides reference for model to determine the parameters;(2)Analysis of riding factors and demand factors based on Markov chain method.Briefly introduces the principles and conditions of the Markov chain modle,and the applicability in the bike sharing system.We choose the two factors of riding time and facilities scale to using this model.And then,we build the quantity transfer matrix and transfer probability matrix.Finally we obtain the steady state probability vector by the Markov chain method which presents the general selection probability of state transfer based on riding time and demand of borrowing and returning by users when the system reaches the steady state.It provided data for determining cost of riding and construction;(3)The location and palnning model of bike sharing system.Build a cost model based on four factors: walking time,riding time,construction scale and penalty of lack.We considering the relationship between the stations and making it as a whole,then choosing the different aggregates of stations.Through the quantification and unit transform,we compared the different location planning program at the price level.(4)Model application example.Based on the bike sharing system in Beijing,this paper builds an example,using the model to determine the final location planning program,and presents the detailed calculation process with the final progran as an example.In the calculation of the results,we analyzed the chocice method under different standards,and compared different types of progranm.