Study On Integrated Train Timetabling And Rolling Stock Scheduling Model And Algorithm Based On Time-Dependent Demand For Urban Rail Transit

The traffic jams and congestion of urban transportation is a problem which each metropolis has to confront. Urban rail transit system has become one of the key approaches to mitigate urban traffic problems in China, due to its safe, efficient, large capacity, and sustainable characteristics. Train timetable and rolling stock schedule are the basis and core for urban rail transit system.This dissertation analyzes the relationship among passenger flow, train timetable and rolling stock schedule in urban rail transit system, and proposes an innovative methodology using a simulated-annealing-based heuristic algorithm to solve an integrated train timetabling and rolling stock scheduling bi-level programming model for urban rail transit line. Based on it, this dissertation considers typical lines (Beijing rail transit Line 2, Line 5 and Line 10) in different scenarios (workday, weekend and mass passenger flow) as real-world applications. The main work and conclusions of this dissertation are as follows:(1) This dissertation analyzes passenger flow, train timetable and rolling stock schedule in urban rail transit system, and discusses their relationship. Based on it, this dissertation develops an integrated train timetabling and rolling stock scheduling optimization methodology.(2) Urban rail transit lines include loop lines and linear lines; number and location of depots are various. In order to formulate an integrated train timetabling and rolling stock scheduling optimization model, this dissertation puts forward a general train flow model for urban rail transit lines, which can be applied to all types of urban rail transit lines.(3) This dissertation formulates an integrated train timetabling and rolling stock scheduling bi-level programming model:the upper model is to optimize train timetable, minimize passengers’waiting time and operation cost simultaneously; the lower model is to schedule rolling stock, minimize number of infeasible train paths. After that, it proposes a simulated-annealing-based heuristic algorithm to solve it, wherein the evaluation function consists of passengers’waiting time, operation cost and penalty value of infeasible train paths.(4) This dissertation considers typical lines (Beijing rail transit Line 2, Line 5 and Line 10) in different scenarios (workday, weekend, mass passengers flow) as real-world applications and uses two commercial softwares to compute for comparison. The results and comparisons show that the proposed model and algorithm has a good versatility and application value.