Customized Bus Network Planning And Route Design Based On Travel Data

With the rapid development of motorization,urban traffic congestion,environmental pollution and other problems become more and more serious.In order to alleviate these contradictions,the public transport priority development policy has become the consensus of all cities.However,with the diversified and personalized development of travel demand,traditional conventional public transport has not been effective in attracting car users or responding to the increasing demand for diverse travel demand.The development of Internet technology promotes the appointment travel,and when the demand of travelers converges or diverges,there is the possibility of mode transfer between customized bus(CB)and online car-hailing.Based on the big data of travel demands,this dissertation discusses the problem of CB network planning and line design.Through the analysis of travel data,the potential demand of CB is mined,and then the method of CB network planning and line design is put forward,which has important theoretical research and practical significance for encouraging the priority development of public transport,alleviating urban traffic congestion and promoting the sustainable development.Firstly,with the analysis of the demand for online car-hailing and public transportation,it is found that the two are similar in time and space in the morning and evening peak hours for commuters.This dissertation summarizes the shortcomings and characteristics of the existing CB system,and provides expectations and feasibility for the demand transfer between the online car-hailing and CB services.Combined with the travel demand analysis,the different distribution of origin(O)and destination(D)is summarized and the CB application scenarios are divided according to the different distribution and land usage of OD.The CB service modes are designed under different scenarios,including recommended applicable models,operation time segment suggestions,etc.Secondly,based on the travel data,the CB network planning process is established,which mainly includes four stages:the travel demand data processing,the origin and destination region division based on the cohesive hierarchical clustering under the regional demand and distance constraints,the set generation of the route to be selected under the constraints of the line demand and the line length,and the construction of the route selection model.The first three stages of case analysis are based on Beijing network car Hailing data,and the original data processing is based on Arc GIS.Beijing is divided into 334 starting areas and 369 ending areas by combining with hierarchical clustering method.Considering the minimum demand constraints and the minimum line length constraints,267 start areas and 265 end areas are selected.489 lines to be selected are generated by pairing the start area and the end area to form a set of lines to be selected.Thirdly,based on the set of lines to be selected,taking the total generalized cost of customized public transportation system as the objective function,considering the constraints of full load rate,line length,vehicle capacity,number of vehicles to be configured,a customized public transportation network planning model is constructed.Using MATLAB programming to solve the problem,the number of passengers served by customized public transport and network car reservation on each line is calculated,and the service level of customized public transport network planning scheme is evaluated.In addition,this dissertation discusses the influence of customized bus types,fixed operating costs of network car hailing,the weight of each shared cost and different fare rules on the customized bus network planning scheme.489 lines to be selected are analyzed based on the data of car hailing in Beijing network.The calculation results show that(1)the combination of CB vehicles with 49 seats and 18 seats is the most cost-effective regardless of factor values.Using the CB vehicles with 18 seats on one CB line is sensible when only one type of CB vehicle can be used.(2)The CB network has the longest total CB line length and serves the most CB passengers with the highest service acceptance rate and the maximum spatial coverage when decision-makers pay more attention to environmental pollution and congestion issues.It also has the lowest vehicle use.Conversely,the CB network has the shortest total CB line length and serves the least CB passengers with the lowest service acceptance rate and the minimum spatial coverage when decision-makers pay more attention to operating cost,environmental pollution,and congestion issues.This scheme also has the highest vehicle use.(3)The CB network's service acceptance rate and the spatial coverage increase with the fixed operating cost per OCH vehicle per day c _0C.CB vehicle use decreases asc_0Cincreases.(4)In the case of passenger flow density,the CB fare standard has little influence on the evaluation index of the line network,among which the effect of the tiered fare is better.Finally,based on the DBSCAN clustering algorithm and the actual situation of the road network,the initial location of the selected bus stops is obtained.Then,based on the node importance of the station,the station to be selected is divided into the bus stops that should to be serviced and bus stops may be serviced.Then,a customized bus route optimization model is established,which takes the minimum generalized cost composed of time value and operation cost of travelers as the objective function.Two typical customized public transportation lines are selected from the planning scheme of Beijing customized public transportation network for case analysis.The results show that:(1)Line 1 is a line with high demand for travel in the urban area,and the specific rout of the line can be generated through model calculation,including 6 boarding stops and 8 alighting stop;among them,3 stops(No.:1,3,5)in the boarding stations are the stops that should be stopped.(2)Line 2 serves the area composed of high-end industrial service center,technology research and development center and its supporting facilities.The travel demand is also high.The generated alighting station points of the line are relatively concentrated,including 5 boarding stations and 3 alighting stations.Among them,there are also 3 stations(No.:1,4,5)in the boarding station that should be stopped.