Timetable Coordination And Optimization For Transfer Stations In Urban Rail Transit

Urban rail transit, as the backbone of urban public transit, plays an important part in relieving traffic jams and shaping low-carbon, energy conservation and environmental protection trip mode. With the rapid development of urban rail transit, transit network is formed from isolated lines; however, the operation of each line is independent and lack of coordination, which makes great inconvenious to passenger transfers. Thus, transfer coordination between different lines under network operation becomes a new research subject.This paper studies the transfer coordination between different lines in urban rail transit under network operation. Using the probility theory and mathematical statistics method, the paper has proposed models for optimizing across-platform transfers and passage transfers respectively. Genetic Algorithm is designed to solve the optimization model. Besides, the network transfer optimization theories and methods have been introduced and transfer walking time in passage analysed to provide tools and data for modeling on transfer coordination.Firstly, the paper analyses transfer theories in rail transit networks, including the planning principles and operation ideas of transfer classification, transfer station layout, transfer coordination from abroad. Meanwhile, sampling survey of transfer walking time in subway passages has been made and Weibull function is used to describe the walking time distribution from the perspective of Cumulative Distribution. Parameter values in Weibull function are derived through maximum likelihood estimation.Secondly, transfer coordination models on across-platform transfers and passage transfers have been proposed to optimize schedules of transfer station. The arrivals of train are considered to be period square pulse, and the train operation is analysed using wave theory. Transfer synchronization is achieved by adjusting phase difference of square pulse to make trains arrive at transfer stations simultaneously under across platform transfers. To the passage transfers, transfer waiting time is strictly defined and used as objective function. The model of minimizing transfer waiting time is formulated and Genetic Algorithm is adopted to solve the problem.Finally, a case study of scheduling optimization in Yonghegong Lama Temple transfer station is carried out using methods proposed by the paper.The results of the paper suggest that Weibull function could well describe transfer walking time distribution in subway passages; timetable coordination in non-peak hours is worthier of optimization than peak hours; square pulse could vividly present the periodical arrivals of trains, the paper provides a new approach for rail scheduling by considering train movement as wave propagation and using mathematics function to express square pulse; case study shows that transfer waiting time per pedestrian has been reduced by 70 seconds (29.7%) and total transfer waiting time 54145 seconds, equivalent to 15 hours during study period after schedule optimization; the models and Genetic Algorithm proposed by the paper is effective to timetable optimization and could be applied to transfer coordination in actual urban rail transit networks.