Optimization Of Last Trains' Arrival And Departure Time At Transfer Stations

Recently,the speed of urbanization in China improves significantly,the population and transportation demands in city increase progressively,results from the rapid development of economy.Consequently,traffic problems have been increasingly serious.Subway system,which is the backbone of urban transit system,plays an important role in large and mediumsized cities.In subway system,some passengers cannot reach their destinations by metro prior to the end of operation.,because of the last train coordination problem.This problem may cut off their trip during last train period.Few of existing researches on timetabling problem took full advantages of actual passenger flow demand.And few of studies considered the time accessibility of paths between OD pairs.Hence,this paper establishes optimization model for coordinating last trains,in order to maximize the number of passengers who can reach their final destinations and minimize time caused by dwelling at transfer stations.This can not only promote the service level of subway system,but also provide travel and transfer guidance for passengers prior to the end of operation.Firstly,this paper introduced Automatic Fare Collection(AFC)system and its data structure,and then proposed a method to dig and match the target OD pairs of passengers during last train period,which might provide a new idea of acquiring target OD pairs from limited data resources.Secondly,considering that all paths weights change over time,this paper chose the shortest route having both spatial and temporal accessibility by the usage of deep-first-search algorithm.Then,a method of calculating critical time of last swiping card before passenger boarding the train was put forward.With AFC system data,a multi-object optimizing model of coordinating last trains using arriving and departing time at transfer stations as decision variables was then proposed.Afterwards,a generic algorithm was used to solve this model.Finally,a case study of Nanjing subway system was applied to verify the effect and the feasibility of the model.The results show that this model can improve the temporal accessibility of subway network and help more people reach their final destinations successfully with only a small amount of increase of dwell time at transfer stations.