Characterization Of Network Performance And Coupling Optimization Of Timetables For Rail Transit

Rail transit has become an essential part of the comprehensive public transportation due to its advantages of high capacity,punctuality,and low carbon emission.In metropolises such as Beijing and Shanghai,the passenger volume of rail transit accounts for more than 50% of the public traffic’s passenger volume,and the operating mileage and total passenger volume rank first in the world.At present,the operating environment of rail transit is becoming more and more complex,with the characteristic of networked operation pattern and the heavy passenger demand.To this end,it is imperative to study the methods for improving the passenger transportation efficiency with the guarantee of operation safety for a rail transit network.This thesis focuses on the routing problem,characteristics of rail transit networks,and timetable optimization from a network perspective.The main work is summarized as follows:1.The routing problem for the rail transit network is studied with the consideration of transfers between lines.Two routing algorithms,passengers’ intuitive routing algorithm and Gaussian Mixture Model clustering based routing algorithm are proposed from the perspective of topology and data-driven methodology,respectively.The passengers’ intuitive routing algorithm is based on an un-weighted L-space model,and the passenger selects the travel path from the shortest path set and the minimum transfer path set.This algorithm could help to accurately characterize the network topology by integrating the transfer behavior with the topological structure.The Gaussian Mixture Model clustering based routing algorithm obtains the effective path set based on an augmented weighted L-space model and effective path rules,and then the probability of each effective path being selected is computed by clustering the automatic fare collection data.This algorithm effectively guarantees the accuracy of path results with lower computational complexity.2.The characteristics of the rail transit network are studied with the consideration of the passenger routing pattern.Considering the characteristics of the passenger travel path of the rail transit network,firstly,the topology efficiency is proposed to characterize the speed of passengers traveling in the network according to the convenience of network passenger travel;further,the node occupying probability is represented by the station utilization rate in travel paths of the entire network to characterize the node centrality;finally,the path disable rate and path cost adjustment are proposed to characterize the network robustness under random failures and target attacks.Through the simulation and comparison experiments with different rail transit network data,the network characteristics of each rail transit network are analyzed and the accuracy of the proposed method is verified.3.The supply-demand equilibrium optimization is studied based on the rail transit network passenger transportation dynamics using a three-layer structure of the railway system.Based on the three-layer structure of rail layer,train layer and passenger layer of rail transit system,a dynamic model of rail transit passenger transportation is established with train capacity limitation.Then time efficiency and maximum average train load rate are proposed to characterize the relationship among passenger entrance rate,headway time,and the network traffic performance,unfolding the evolution of network traffic from free state to congested state and then to the saturated state.Moreover,an optimization model is proposed to balance the transportation efficiency and traffic load by adjusting the train headway time under a given passenger demand.The simulation results verify the relationship among passenger demand,train resources,and traffic performance,and show the optimal headway time under give passenger demand to guarantee the network traffic performance.4.A railway line timetable optimization method for dynamic load balancing is proposed with consideration of the operating cost and service quality of the rail transit system.The rail transit line timetable under the condition of dynamic passenger is optimized to reduce passenger waiting time and total energy consumption,and improve passenger comfort level and train resource utilization.An index named average maximum load deviation is proposed to identify both the utilization of trains and the congestion level for passengers,providing an indicative parameter for achieving dynamic balance of train loads.Finally,a hybrid algorithm combining the particle swarm optimization algorithm and the simulated annealing algorithm is designed to solve the problem.Case study is performed on a metro line with the passenger automatic fare collection data.Results show that the proposed approach could reduce the average waiting time,total energy consumption,and the average maximum load deviation by 5.14%,14.68%,and 69.69%,respectively.5.A weighted coupling optimization of the rail transit network timetable based on the station centrality for transfer coordination is studied.In view of the heterogeneity of the rail transit network,a weighted transfer time index with station centrality as the weight is proposed.Then a weighted coupling optimization model is established to reduce the total transfer waiting time and improve the spatiotemporal connectivity of the entire network.Then,a hybrid algorithm combining the particle swarm optimization algorithm and the simulated annealing algorithm is designed.Finally,the operation data and the structure data of a metro network are used as an example for case study.Results show that the proposed approach could reduce the the transfer waiting time by 12.97%,and improve the coupling number by 6.12%,indicating the effectiveness of the proposed model to improve the transport efficiency and spatiotemporal connectivity.