Timetable Optimization With Transfer Coordination In Urban Rail Transit Systems

With the development of urban rail transit network,the numbers of transit lines and stations increase,and passenger transfer via network becomes commonplace.Given that the conditionality between lines and stations is quite complicated in network,it is tough for operators to coordinate the network relation between lines based on their experience and management expertise.Thus,there is urgent demand of the scientific and intelligent decision support to guarantee the network performance.This dissertation is devoted to studying the problem of train timetabling in urban rail transit systems from the perspective of passenger transfer,which covers the transfer efficiency,network accessibility and transfer robustness.The transfer connection mechanisms and relationships are modelled from different aspects to optimize the network plan.The detailed research contents include:(1)Transfer efficiency optimization.In the case of timetabling with even headway that is most commonly used in actual operation,the periodicity related to the transfer connection relationship and the time delta between the departure time of connection train and the arrival time of feeder train are researched,and then the synchronization timetabling model are constructed.In this way,the scale of variables is significantly reduced.Meanwhile,in order to improve the synchronization performance,an alterative timetabling model are established with the notion of elastic headway by introducing the departure phase and elastic deviation.Moreover,by analyzing the specific structure of the model,four classes of valid inequalities are designed to improve the computational efficiency when implementing the branch and cut algorithm,and a genetic algorithm is developed when tackling the large-scale network case.Finally,the case studies are conducted and results show that valid inequalities can improve the performance of branch and cut algorithm,but it still cannot cope with the large-scale network,while the genetic algorithm can yield the optimal solution fast.Besides,timetabling with elastic headway can effectively improve transfer efficiency.(2)Network accessibility optimization.The differences of the connection relationship between the start-and end-of-service periods are discussed,and the advantages and necessity of timetabling from the perspective of the start-and end-of-service periods are illustrated.By designed three kinds of connection index,the transfer connection relationship is formulated by making use of the train connection time.Then,a coordination model is constructed by minimizing the passenger connection time to simultaneously optimize the transfer efficiency and network accessibility,which can be solved by branch and bound algorithm.Finally,some experiments are carried out based on the Beijing rail transit network and results show that the method can yield coordinated network plans with significant improvement in both transfer efficiency and network accessibility.(3)Transfer robustness optimization.The function and substitution effects of several different types of buffers are discussed and it is worth noting that the running time supplements and transfer buffer times yield different effects on the travel plans of transferring and non-transferring passengers.Based on this,an expected extra travel cost(EETC)function is formulated to appropriately balance efficiency and robustness,which is then implemented in the construction of a robust transfer optimization model with the objective of minimizing the total EETC.Next,to improve the computational efficiency,an approximate linearization approach for the EETC function is proposed.Experimental results show that the proposed method can yield practically applicable solutions with significant reductions in both EETC and probability of missing a planned transfer.