Passenger Demand-Driven Optimization Of Customized Bus Lines Based On Bus IC Data

In recent years,with the rapid development of urbanization,the ownership of car countinues increasing,and there is a contradiction between supply and demand when considering the limited road resources and serious road congestion.To solve this problem,the governments establish public transportations(PT)in high priority.However,there have been existing the problems of low punctuality rate,poor comfortable performance,and lots of transfers in the public transportation,especially the bus transportation system.Therefore,thses problems make bus transportation less attractive to passengers and slow to develop.At the same time,passengers' demand for travel is getting largerer and largerer.Customized buses cannot meet these needs.Therefore,how to clearly and intuitively understand passenger travel characteristics and the distribution of temporal and spatical needs is the essitioaul issue for the development of suitable bus service line designe.Therefore,the bus transportation system can enhance the comfortable and efficient services for passengers,which is also the fouce of research in this paper.First of all,the background and current situation of the bus route optimization have been analyzed deeply.The detailed introduction of the characteristics,operating procedures,and current status of the customized bus transportation have been made.On this basis,the bus network design,optimization principles,objectives etc are elaborated for the customized bus route optimization,which provides a theoretical basis for the subsequent analysis of passenger flow characteristics,station selection and route optimization.Secondly,based on IC card data and the bus travel chain,the passenger travel information recognition method combing with bus station information and route information is proposed.The proposed method mainly includes original and destional station identification,the matching between station name and station number,and OD station identification,station passenger flow statistics and other aspects.According to the obtained travel information,the characteristics of passenger travel are analyzed in case of temporal and spatical aspects.From the temporal perspective,the characteristics of daily hour passenger flow and peak hour passenger flow during the moring and evening are analyzed.From the spatical aspect,the spatical distribution of the departure stations in whole day and peak hour are analyzed.Moreover,the passenger flow characteristics for stations,spatical characteristics of passenger flow,and the characteristics of distance between OD passenger flow are also analyzed.Based on the analysis of passenger flow characteristics,excavating passenger travel laws and discovering the potential passenger flow of customized bus routes,the customized bus route optimization method is provided in this paper.Then,according to the passenger flow distribution area and passenger outflow demand OD,the bus sharing station of the customized bus is planned,and the improved K-means clustering algorithm based on the fusion hierarchical clustering method is proposed.The proposed clustering method combines with the station passenger flow and the maximum carrying capacity of the station constraints which are foundations for the selection of customized bus stops.Finally,based on passenger flow demand and station selection results,considering passengers,bus companies,and social environment aspects,a customized bus route optimization model is formulated,which minized the travel time of passengers,the bus operation cost,and the environmental pollution cost.Furthermore,an improved genetic algorithm is proposed to solve this model.The experiment based on the area of "OrchardInternational Trade" in Beijing is carried out.The results show that the feasibility of the model was verified and some strategic suggestions for the optimization of customized bus routes are provided.There are 65 figures,32 tables and 83 references in the body.