Transfer-Based Modeling Flow Assignment With Empirical Analysis For Urban Rail Transit Network

Along with the construction of urban railway network, passenger flow displays network characteristic. The time-space distribution and path choice behavior greatly changes. Under the network operation and seamless transfer, passengers need not to pay for transfer, which results in the uncertainty and difficulties in the statistics of passenger volume both on path and in transfer station. And this gives rise to great difficulties for urban railway operation and management. Therefore, theoretic research on urban railway path choice and passenger assignment does great help to capture the passenger network pattern and enhance the management level of urban railway operation.The core of this dissertation is transit assignment of urban railway network. Major factors impacting the transit are taken into account and the urban railway flow distribution pattern is analyzed. On base of these, some mathematical models are presented for urban railway traffic network flow assignment problem. The application of these models and algorithms are illustrated with Beijing urban railway traffic network. The contents of this paper are summarized as follows:(1) In practice, the flow assignment pattern through urban rail transport network is resulted with the aggregation of all passengers' combined choices. Therefore, understanding and mastering the passenger's choice psychology is the foundation of modeling the passenger's route choice and flow assignment through whole network. In order to fully grasp the characteristics of the passenger route choice behavior, meanwhile, in order to provide reliable data for future research support and related parameter estimation, this paper designed a questionnaire of passengers' choice behavior and gives fully data statistical analysis of Beijing urban railway. The result proves that age, gender, income level, occupation and trip purpose have great influence on path choice. Travel time is the most significant factor while time of transfer and crowdedness follows. Passengers with different attribute have different response to these factors so that the presented model gives fully consideration to this besides the time in cars, transfer time and numbers needed of transfer.(2) The major factors (including travel time and transfer) that influence passenger flow pattern in URTS are taken into account. The general cost function for generalized urban rail transit network is presented and corresponding route choice behavior of passengers is analyzed. On base of these, an improved logit-based model is then presented for URTS flow assignment problem which is tested by Beijing urban railway data in2008. The relative error of this transfer-considered model is31.22%, which is16.44%lower than the no-transfer model.And sensitivity study of general cost model shows the result is very high sensitive which means that the more number of transfer needed, the lower chance the path be selected. This is partially because that the enlargement of urban railway network and increase of transfer station make number of transfer a critical factor for passenger's path choice. (3) According to survey data, the passengers with different socio-economic attributes will have different standards and preferences of traffic choices. This paper considers the different choice behaviors between different passengers fully and establishes the generalized travel cost functions for different types of passengers. Then the urban railway flow assignment model is built considering the passenger classification. And improved Dial algorithm based on random network assignment is proposed to solve the model rapidly. And the result indicates that the assignment model based on passenger classification can demonstrate the influence of passenger attributes on assignment result and is more precisely on passenger's preference on path choice.(4) This paper gives full consideration of the crowdedness factor in the passenger's route choice and proposes crowdedness-based generalized travel cost function and Probit based assignment model which is solved by MSWA algorithm. And a case study is done based on this model. With one pair of OD, the effect of parameter on result precision is analyzed which shows that (α,β)=(1,2) gives the best result and reliability and lowest relative error. With multipaire of OD, the effect of parameter on result precision is analyzed and the analysis proves that MSWA algorithm gives lower speed and larger relative error with more OD pair.(5) Based on the proposed model and Fratar model based OD, the design and development of network passenger flow analysis software for Beijing URTS are introduced in this paper, which can calculate statistical index of railway flow automatically.