Multiperiod-based Timetable Synchronization Optimization In Urban Rail Transit Networks

With the acceleration of urbanization process,a dramatic increase in transportation,with concomitant crisis in energy,issues in environmental,and the shortage in land resource,the traffic problems already seeped to society's each aspect.Facing rising travel costs and an aging population in need of more services,subways offer high security,time-saving travel,wide accessibility,energy efficiency and reliably superior to other modes of transportation in cities service.Thus,travelers adopt subways as their primary mode of transportation in modern society.Currently,subway system has entered a new era of networking operation.Moreover,the density of the network and transfer stations,distribution of train flow would gradually increase along with the new lines access.The denser a network is,the more convenient it becomes to the travelers.However,having more lines and stations to a network increases the complexity of timetable optimization for the system.Hence,it is increasingly difficult to manage the operation of the urban rail transit network systems.This dissertation focusses on the importance degree of transfer stations and passenger flows in urban rail transit networks and tackles the multiperiod-based train timetable optimization problem to enhance the performance of transfer synchronization between different rail lines.Train timetables of connecting lines are adjusted in such a way that train arrivals at transfer stations can be well synchronized.The remainder is organized as follows.(1)This study firstly introduces fundamental timetable theories of urban rail transit networks,i.e.,core elements of train timetable,optimization principles,evaluation indexes,the characteristics of passenger flows,transfer conncetion types,transfer characteristics.Subsequently,the real passengers' flow data of Beijing subway are utilzed to find characteristics of passengers in different time dimensions and spatial dimensions.Moreover,this dissertation establishes a method based on the importance of lines and transfer stations with the features of first train.Subsquently,the train timetable synchronization model is proved to be NP-Hardness.This will further expand the collaborative model as multiperiod-based timetable model for optimizing synchronization events in urban rail transit networks.(2)This study analyzes the operation characteristics and principle for first trains.A sub-network connection method is developed,and a mathematical programming solver is utilized to solve the suggested model.A simple test network and a real network of Beijing urban railway network in 2014 are modeled to verify the effectiveness of our suggested model.The optimized results reduce the selected stations' total connection time by 11172 seconds or 43%.Results demonstrate that the proposed model is effective in improving the transfer performance in that they reduce the connection time significantly.(3)This study particularly focuses on the timetable optimization problem in the transitional period(from peak to off-peak hours or vice versa)during which train headway changes and passenger travel demand varies significantly.A mixed integer nonlinear programming model and an efficient hybrid optimization algorithm based on the Particle Swarm Optimization and Simulated Annealing(PSO-SA)are designed to obtain near-optimal solutions in an efficient way.Finally,a numerical example and a real-world case study based on the Beijing subway and travel demand is conducted to validate the proposed timetabling model.Computational results demonstrate the effectiveness of adjusting train timetables and the applicability of the developed approach to real-world subway networks.(4)With a concern on the last train service synchronization issue,this study developed a mathematical optimization model to determine the optimal departure time for last train services with the objective of maximizing synchronization events for system operators and at the same time minimizing transfer connection time for passengers.Firstly,a multi-objective optimization model was developed to generate favorable train departure time,running time and dwell time.Secondly,an improved version of NSGA(non-dominated sorting in genetic algorithms)was used to find near-optimal solutions efficiently for real-world urban rail transit systems.Computational experiments are conducted based on the Beijing subway network,and the results demonstrated that the proposed model effectively improves the number of successful synchronization events and shortens the longest waiting time with the optimized train service schedule.