Research On Optimization Of Train Schedule Based On Short Turning Strategy For Urban Rail Transit Lines

With the continuous expansion of the urban rail transit network,transport capacity of some urban rail transit lines cannot meet the growing passenger demand during peak hours.In order to alleviate excessive pressure for passenger transport and overcrowding during peak hours,operation companies generally take measures to increase rolling stocks to further shorten the headway,but with a largely increase in operating costs.In addition,other measures such as setting obstacles to bypass,closing the entrance and exit are also adopted by operating companies to ease the traffic tension and ensure passengers’ safety.However,passenger flow restriction will increase passengers’ travel times and reduce passenger satisfaction.Therefore,it is significant to find a reasonable train scheduling method that can reduce operating costs and improve passenger satisfaction according to a thorough analysis of the spatial and temporal distribution of passenger flow.Based on the analysis of the spatial and temporal characteristics of the ladder-type passenger flow distribution,this paper proposes an automatic train operation jump algorithm with short turning strategy,which provides a basis for the automatic train scheduling and rescheduling for urban rail transit lines in China.Specifically,main contributions of this thesis are given as follows:(1)This paper expounds the spatial and temporal distribution characteristics of passenger flow on urban rail transit lines,and analyses the influencing factors of the routing schemes selection on the basis of the classification of train routing schemes,which provides theoretical support for the study of train scheduling with short turning strategy.Meanwhile,this paper briefly discusses the classification and propagation principle of train operation delay in order to study the automatic train rescheduling under small disturbance.In addition,solving algorithm for the multi-objective optimization problem and mixed integer linear programming(MILP)algorithm involved in train scheduling are briefly introduced.(2)A macro-model of an urban rail transit line and stations is established on the basis of analyzing the configuration information of urban rail transit lines,stations,and depots as well as traffic organization rules.According to the ladder-type passenger demand distribution characteristic,the study on train scheduling with short turning strategy is carried out considering the train circulation plan.The mixed integer non-linear programming model is transformed into a mixed integer linear programming model by several linearization transformation properties.Branch and bound algorithm is adopted to solve the model,which realizes the integrated optimization of train schedule and train circulation plan with short turning strategy.(3)With the increasing number of train services,the scale of the mixed integer linear programming problem increases sharply.Besides,the mixed integer programming problem is NP-hard problem,which causes the computational complexity increases exponentially.Therefore,this paper proposes static rolling horizon algorithm and dynamic rolling horizon algorithm to solve this problem.The daily train schedule can be obtained by setting a reasonable length of time and designing an effective connection mechanism for train services across time intervals.With the comparison of these two algorithms,the results show that the static rolling horizon algorithm can improve the utilization rate of rolling stocks better.(4)Train rescheduling problem based on running grades under small disturbance condition is studied to achieve the goal of energy-saving and punctual operation.The running time is adjusted by selecting the running grades in each inter-stations,and the dwell time is determined by the load factor of the section and the number of times that the platform screen door closes.In order to enhance the robustness of train rescheduling,an optimal train rescheduling model considering stochastic scenarios with short turning strategy is constructed.Finally,a robust train rescheduling scheme is obtained.