Research On Last-train Timetables Coordination And Optimization Method In Urban Rail Transit

Due to the increasing number of rail lines, cities gradually form an urban rail network. The original single-line mode of operation is difficult to adapt the needs of network operating. Transfer coordination between different last trains is an important part of network operations. Adjust the last train departure time to achieve the coordination between the last trains becomes a significant research subject.This paper is studying the transfer coordination between different last trains in the urban rail network operations. Built a model of last train timetable coordination, its object is the maximum number of transfer coordination passengers, considering the specific transfer direction constraints. Meanwhile, built a model of network last train timetable optimization, by extending the dwelling time of last train to increase the accessibility of last train and achieve two-way transfer. For the model, we can use the genetic algorithm to get the answer.Firstly, analyzed the characteristics of the transfer passenger flow and made a comparison about the passenger behavior between the last train passenger and the regular time. For the last train passengers, we proposed that the last train transfer traveling time should consider the needs of most transfer passengers. According to the different circuit layout of the transfer station, we analyzed the coordination conditions of all kinds of transfer station and proved some conclusions that the two coordination relations between the two last trains cannot be achieved without prolonging the dwelling time.Secondly, built a model of last train timetable coordination and solved it by genetic algorithm. The goal of the model is the transfer demand of the most transfer coordination last train in the network, considering the specific transfer direction constraints. Meanwhile, we improved the computing speed of the algorithm by changing the constraints from the specific transfer direction to the departure time domain. On the basis of the coordination model, in order to build the model of network last train timetable optimization, we proposed to extend the dwelling time to achieve two-way transfer and increase the accessibility of network, and analyzed the extending value and applicable conditions of the dwelling time.Finally, based on the network of Beijing Subway Line1, Line2and Line5, we calculated three cases:the transfer direction without constraint, the constraint in one direction, and the prolonged dwelling time. The network with the original last train timetable only can coordinate19transfer direction and610passengers, while20transfer direction and906passengers with no transfer direction constraint, and20transfer direction and858passengers with a particular transfer direction constraint. In the S2transfer station, the last train dwelling time of LI and L5were four minutes and three minutes. With this situation, the coordination flow was964passengers and L1and L5achieved a two-way transfer. The last train timetable coordination model and optimization model designed in this paper, and the genetic algorithm, is effective for the coordination optimization of the last train in the transfer station, and has a certain utilities.