Research On Optimizing Design Method Of Customized Bus Route Based On Passenger Travel Mode Choice

In recent years,with the development of economy and society,the construction of urban transport infrastructure in China has developed rapidly,but the impact of motorization is still a problem faced by many cities.In view of the contradiction between the demand of urban traffic motorization and resource constraints,the government’s duty is not only to provide transportation infrastructure,but also to guide traffic behavior through traffic policy,and to build a multi-mode public transport service system with resource and energy conservation,so as to realize the priority of public transport and the sustainable development of urban traffic.Therefore,a new and innovative demand-responsive transport service—customized bus(CB)is proposed.It aims to provide one-stop direct arrival,personalized and high-quality public transport services for medium and long distance commuters with similar origin and destinations.In view of the difference and connection with the traditional bus,this paper proposes an optimization method for the line design of the CB according to the operation characteristics of CB.Firstly,this paper begins with the analysis of passenger travel demand to understand the passengers’ travel characteristics and perceive the distribution of passenger origin and destinations.Then,the model of passenger’s travel mode selection is established,and the parameters of the model are calibrated by using the actual stated preference(SP)and revealed preference(RP)data to analyze the main factors affecting passenger’s choice of CB,which can provide reference for the design of CB lines.Finally,based on the results of passenger’s travel mode selection,a bi-level programming model is established to design CB lines,in which the lower level model is a stochastic user equilibrium(SUE)model of hybrid bus network and the upper level is a vehicle routing optimization model with time windows(VRPTW)considering the characteristics of CB lines.Therefore,the research results of this paper can provide suggestions for CB operators to optimize line design and operation management,and contribute to further promote the development of CB.The main research contents include the following aspects:(1)This paper analyzes the travel demand characteristics of CB passengers.By comparing with other commuting modes in the city,the traffic characteristics of CB are analyzed.Based on the SP survey data of potential CB passengers and the RP survey data of actual CB passengers,the characteristics of different types of passengers are analyzed,including traveler characteristics,time-space demand characteristics and traveler preference characteristics.The demand characteristics of CB stations and routes are also introduced.(2)This paper perceives the origin and destination distribution of the travel demand of CB passengers.Considering the line design characteristics and travel conditions of CB,and relying on the clustering idea of DBSCAN algorithm,an improved algorithm of clustering origins and destinations in pairs,P-DN algorithm,is proposed.Based on the IC data of traditional bus,a method of identifying potential CB passengers from traditional bus passengers is proposed;the origins and destinations of potential CB passengers’ travel demand are determined,and feasible distribution schemes of route origins and destinations is proposed.(3)This paper establishes the mode selection model of CB passengers.A three-level nested logit model based on the SP-off-RP survey data is established;information about commuter’s travel mode selection is collected by online and offline SP-off-RP questionnaire survey;parameters of the model are calibrated by Stata 12.1 software and actual survey data,and significant factors affecting commuter’s travel choice of CB are analyzed.Based on the marginal effect,the transfer effect between different travel modes are proposed.(4)A bi-level programming model of CB route design based on passenger travel mode selection is established.Based on the results of the CB passenger travel mode selection model,a bi-level programming model combining the vehicle routing optimization model with time windows and the stochastic user equilibrium model of hybrid bus network is established to obtain the CB line design schemes considering the interaction of minimum passenger travel time and their travel choice.Among them,the lower model is solved by improved allocation algorithm combined with Dial algorithm and method of successive averages(MSA),and the upper model is solved by the branch and bound algorithm.The research results are validated by a case study of CB lines designed by the road network of the Eighth District of Beijing.