Research On Multi-Objective Customized Bus Network Planning

In recent years,the market share of public transport continues to slide,and the loss of passenger transport enterprises is increasingly aggravated.How to get out of the predicament of passenger transport market has become an important issue at present.As a new demand responsive public transport travel service mode between traditional public transport and online car booking,customized public transport has the characteristics of"door to door","point to point","follow the passenger" and"mobility and flexibility".It has gradually become a new starting point,a new way and a new trend of public transport business transformation and upgrading.Under the condition of market economy,the passenger transport market should pay more attention to operation,pay attention to subdivision,and promote customer growth by means of fine operation and service.However,due to the late start of customized bus in China,the customized bus operation service system is not perfect,lack of innovation,part of the operation mode is still based on the traditional bus mode,customized bus network and operation organization planning is not reasonable,customized bus network has not formed a scale.Firstly,from the relevant theories of customized bus,this thesis deeply analyzes customized bus theory,standard vehicle routing problem theory,conventional bus network planning theory,and makes a new definition of customized bus from the perspective of operation.Preprocessing the collected data,combining with the customized bus station planning theory,determine the location of the station,and update the customized passenger demand data.The customized bus station planning is described,and the improved DBSCAN algorithm design is proposed for the clustering of passenger travel demand,which improves the discrete degree within the cluster,the inter-cluster importance,the automatic updating selection of Eps and MinPts parameters,and the determination of cluster center location.The refined classification steps of clusters are given,the running time between stations is calculated,and the set of primary selected stations obtained by the improved DBSCAN clustering algorithm is used to prepare for customized bus route network planning.Secondly,from the perspective of operation service system,this thesis analyzes the four main aspects of customized bus:operation service,operation scheduling,operation management and operation guarantee,and demonstrates the operation service elements,so as to build a more complete operation service system to meet the needs of passengers at different levels.In addition,the "point,line,surface,line network system and other modes" of customized bus network operation mode is deeply studied to make the customized bus operation organization and line network structure reasonably connected.Through comparison,this thesis takes the station as the multi-point and multi-point,the station is the combination mode of fixed station and non-fixed station,the line is the flexible line,and the line network adopts the operation mode of the combination line network.The key research is carried out by comparing the selected stations,lines and network elements,paving the way for the model construction of chapter four.Thirdly,aiming at the two aspects of passenger travel efficiency and enterprise operation efficiency,considering the site as"multi-point at most",line operation mode for multi-mode mixed operation,and consider the multi-model,different needs mixed load,starting and ending schedule of static travel demand.In this thesis,a multi-objective line network planning model with multi-constraints such as vehicle capacity,number of stations and line length is constructed.In order to achieve the optimal effect of two aspects of benefits at the same time,an improved genetic algorithm—NSGA-Ⅱ algorithm is designed to solve the model,and the Pareto solution of the network planning scheme and the multi-objective planning model is obtained.Finally,through the example of city B,the specific location of the station is obtained,the validity and accuracy of the improved algorithm are verified,and the candidate station of the customized bus is obtained.Combined with the actual operation of the customized bus in city B,the obtained data information of the alternative stations is substituted into the model,and NSGA-II is used to solve the problem,and the customized bus network planning scheme is obtained.Through the analysis of the results,suggestions for customized bus network planning are given,so as to obtain more reasonable and more practical customized bus operation planning routes.