Study On Connection Optimization Of Last Train Timetable On Urban Rail Transit Network

A reasonable last train timetable not only directly reflects the operation of urban rail transit service quality, but also fully embodies the’people-oriented’service concept. Urban rail transit system is developing rapidly from lines to network, and it puts forward higher requirements on last trains’connection optimization. Therefore, how to arrange the arrival time and departure time effectively has become a key issue in the last train operation plan. To fully satisfy passengers’as well as operation companies’ different demands, and improve the practicability of the optimized scheme, this paper studies on connection optimization of last train timetable in order to match the characteristics of urban rail transit network. Specific researches are as follows:Firstly, on the analysis of last train operation as well as passenger flow’s time and space distribution characteristics, this paper summarizes the existing last train timetable’ connection condition, and puts forward some problems on the evaluation method of main transfer directions as well as timetable scheduling.Secondly, this paper analyzes transfer directions’ conflict relationships, and proposes a method to measure the relative importance of transfer directions, which combines ’node betweenness’ and ’transfer flow’ simultaneously. Then a multi-objective optimization model in deterministic circumstance, with the objectives of maximizing passengers’ total transfer dissatisfaction, maximizing passenger flow transfer successfully, and minimizing the latest operation time of all line directions is built. The NSGA-Ⅱ algorithm is designed to solve the problem at the same time.Thirdly, based on the deterministic last train timetable optimization model, this paper introduces random variable ’transfer walking time’ into the model which contains stochastic chance constrain. The solving methods, such as certainty equivalent transformation and hybrid intelligent algorithm, are designed to solve the random environment optimization model.Finally, takeing Beijing subway network and actual passenger flow data as an example, this paper solves the two models given respectively. The results show that, compared with present situation, the indicators of deterministic environment optimization are improving. Among them, passengers’ total transfer dissatisfaction has reduced 14%, passenger flow transfer successfully have increased 6%, and the latest operation time of all line directions have optimized 57%.Besides, random environment optimization also gets good results, more in line with the actual situation. It proves that the two models and corresponding algorithms are feasible and effective.