| In recent years,the process of network operation of urban rail transit system in China has continued to advance.As the last stage of train operation,the last-shift period determines the transfer efficiency and accessibility of passengers at night.Therefore,with the increasing complexity of the network structure and the increasing passenger demand,how to optimize the train schedules during the last-shift period has become an important issue for the operation of current urban rail transit systems.Based on the current research literature,this paper comprehensively considers the last train and multiple trains in the evening operation period of each line,and cooperatively optimizes the train schedules in the last-shift period considering the transfer connection and OD accessibility in the network respectively and proposes the corresponding theoretical models and solution algorithms.The specific contents of this paper are as follows:(1)The characteristics associated with the last-shift period are analyzed.Firstly,the definition of the last-shift period in this paper is introduced and compared with the general last train timetable optimization,emphasizing the importance of studying the optimization of train schedules during the last-shift period.In the context of the current development of urban rail transit,the differences between the last-shift period and other periods in terms of passenger flow characteristics,train operation characteristics and transfer connection are analyzed,and the influencing factors for the optimization of train schedules are outlined.(2)From the perspective of transfer connection,the train schedules during the last-shift period are optimized by taking transfer stations in the network as the research object.Based on the transfer passenger demand in the network,a dual-objective non-linear optimization model is developed considering the transfer performance and transfer efficiency of passengers,with the objectives of minimizing the number of failed transfer passengers and minimizing the average waiting time for successful transfer passengers,respectively.In addition,a hybrid genetic-simulated annealing algorithm is designed to solve the model to improve the quality and efficiency for large-scale network cases.(3)The OD demand-oriented train schedules are optimized during the last-shift period.As successful transfer connection is not equivalent to OD accessibility,a further study is carried out to optimize train schedules during the last-shift hours based on the path choice behavior of passengers,which takes the OD accessibility and travel convenience into account,and guided by the OD demand of the whole network.A space-time network is introduced to carve out the trajectory of trains and the path of passengers,and a 0-1 programming model is established with the objective of minimizing the total travel time of passengers.In order to improve the solving efficiency under large-scale networks,the model is decomposed into an upper-level timetable optimization and a lower-level path selection problem,which are solved by the adaptive large neighborhood search algorithm and the improved Dijkstra algorithm respectively.(4)The models and algorithms proposed in this paper are validated based on the Wuhan metro network.The results show that based on the transfer passenger demand during the last-shift period,the number of failed transfer passengers is reduced by31.38% and the average waiting time for successful transfer passengers is reduced by13.38% after optimization.Based on the OD demand,the number of inaccessible passengers in the network is reduced by 19.96% and the total travel time is reduced by12.89%,and the model has better robustness under different passenger flow demands.In addition,the optimization results of both models are significantly better than the last train timetable optimization model,thus validating the effectiveness of the last-shift period model.This paper contains 31 figures,33 tables and 70 references. |
PDF Full Text Request