Study On The Passenger Flow Assignment Model And Algorithm Of Urban Rail Transit Network

With the rapid construction of urban rail transit (URT) systems in China, the rail lines are gradually developing into the shape of networks. The expansion of rail network and improvement of the structure have obvious impacts on the assignment of passenger flow and network operation. Under the condition of network operation and seamless transfer, passengers have more choices of paths, making the passenger flow assignment more complicated. The study of passenger flow assignment could not only help to improve the operation of URT, but also helps to forecast the passenger flow and study the problem of transit fare clearing.To describe the path choice behavior, a network topological method is proposed. First, based on the previous studies, the paper gives a new definition of hyperpaths by studying the differences of hyperpaths between URT and urban traffic. Second, based on the breadth-first search method, the paper proposes a new search method. Third, a case study is conducted to show the effectiveness of proposed definition and path search method. The paper also examines the parameters of the suggested new definition. A group of parameters pairs which are considered more reasonable are given.One major concern of passenger flow assignment is the path choice problem. This paper discusses the factors influencing the path choice behaviors. By comparing the differences of path choice behavior between URT and urban traffic, the major qualitative and quantitative factors influencing the path choice are proposed. According to the discrete choice methods, a utility function which takes passenger attribution and path attribution into account is proposed. A passenger flow assignment model based on Logit model is formed, and the calibration and verification of the model together with the static and dynamic assignment algorithm are discussed. The thesis further generalizes the basic passenger flow indexes of the network. And the estimations of the indexes of stations, lines, and network are discussed.Finally, a example is given to illustrate the application of the proposed model and solution algorithm. The results show the effectiveness and the feasibility of the proposed model. The model is suitable for analyzing URT passenger flow assignment.