Development of a Behavior-Based Passenger Flow Assignment Model for Urban Rail Transit in Section Interruption Circumstance

Urban Rail Transit (2015) 1(1):35–46 DOI 10.1007/s40864-015-0002-0 http://www.urt.cn/ ORIGINAL RESEARCH PAPERS Development of a Behavior-Based Passenger Flow Assignment Model for Urban Rail Transit in Section Interruption Circumstance Jing Teng1 • Wang-Rui Liu1 Received: 3 December 2014 / Revised: 15 February 2015 / Accepted: 19 February 2015 / Published online: 9 May 2015  The Author(s) 2015. This article is published with open access at Springerlink.com Abstract At present, the urban rail transit (URT) system has achieved network operation in many major cities of China. But, little attention has been given to the vulnerability of the URT system. The purpose of this study is to assign the passenger flow under the condition of section interruption in URT system. Two surveys (a passenger behavior survey and a stated preference survey) were conducted and a multinomial logit model was developed. The results show that although the first choice of passengers in emergency situation is to stay in URT system by a circuitous way, more than half of the respondents express interest in the temporary shuttle bus. For the temporary shuttle bus, the sensitivity analysis show that the relative speed is more important than crowding degree for passengers. The significant variables mostly fall in the personal attributes such as income, gender, age, etc. The impacts of trip feature factors are similar to the previous research in normal situations. These results provide basic support for passenger flow assignment at the shuttle bus level and reducing the risk of crowding at some special stations. Moreover, it is also good for reducing passenger delay and recovering the trip. Keywords Urban rail transit  Section interruption  Passenger flow assignment  SP survey  Multinomial logit model & Wang-Rui Liu 1 Key Laboratory of Road and Traffic Engineering, Ministry Education, Tongji University, 4800 Cao’An Road, JiaDing District, Shanghai, China Editor: Marin Marinov 1 Introduction URT, with its large capacity and high reliability, is gradually developing to be the favorite traffic mode in metropolis. However, for the concentrated passenger flow, limited space, closed running, and high capacity characteristics, once an emergent event happens in URT system and operation is interrupted, will come out in a huge amount of delay and spread rapidly in network. According to the process of emergent measures in URT system, the organization of passenger flow can be divided into two steps—evacuation and recovery. The former is at the beginning of emergent events, trying to evacuate passengers to a safe place as fast as possible; the latter is active after the evacuation step and all potential risks will be cleared. During the recovery phase, the common practice is to maintain the integrated service of OD trips, using other traffic modes, usually bus transit, to replace the interrupting section, which is the research question in this paper. The basic need in emergency situation is to maintain the accessibility of rail lines and the integrated service of OD trips; that is to say, some indispensable measures need to be proposed on transport organization level. Passenger flow assignment under emergent condition provides essential support for emergency decision, such as adjusting train routings and departure intervals, scheduling plans of shuttle buses (i.e., the temporary shuttle buses), etc. Quite a few studies [5–8] have focused on passenger flow assignment in URT network. User-equilibrium model and discrete choice model are the most commonly used methods in passenger flow assignment. The factors in the generalized cost function consist of rail-ride travel time and passengers’ stay time at a transfer or passing station. An amplification factor is usually introduced to describe transfer time. Though some different disposal ways were 123 36 adopted for these factors in previous papers, they were not considered from passengers’ standpoint. For example, invehicle crowding degree is hardly considered in previous papers for it is hard to be measured in practice. In addition, comparing with the variable of the number of stations, most papers prefer travel time. Although travel time seems to be more accurate and objective, the variable of the number of stations has its own advantages, such as, the number of stations between the interrupting section and the origin station, the number of stations between the interrupting section and the destination station. These variables are probably more intuitive and convenient for passengers to choose the best route in emergency situations. In recent years, some researches [9–11] pay attention to passenger flow assignment under emergent condition in URT network. Delay on passengers and the affected area are considered. The number of affected passengers is also calculated for every affected station. However, these researches are still in the theoretical stage, and many practical factors are out of consideration, which will be hard to provide enough support for the emergent decision makers. This paper selects factors, in the utility function, from both the passenger behavior survey and previous researches. These factors are brought in the state preference (SP) survey, including the crowding degree, relative speed, the number of stations in the shortest route, etc. In the SP survey, a shuttle bus passageway replacing the interrupting section is assumed to establish an integrated URT network. Then, a discrete choice model is built and analyzed. The objective of this paper is to assign passenger flow under emergent condition and to provide support for the emergent decision makers. The paper is organized as follows: first, previous researches on passenger flow assignment in normal and emergent situations in URT system are reviewed, followed by a detailed explanation of the methodology and data preparation, including passenger behavior survey, SP survey, and route choice model. Then, with the data collected from a passenger behavior survey and a SP survey in Shanghai URT system, the route choice model is calibrated and analyzed. 2 Literature Review Several existing reports [1, 2] presented the importance of using the bus transit as the connecting mode when railway or metro emergent events happen. Between the URT and bus systems, the synergy of timetables was the key to realize the intermodal transportation in emergency. Shanghai subway system stipulated [3] the shuttle buses need to be deployed when the delay of the system is more than 30 min. Beijing subway system also stipulated [4] traffic 123 Urban Rail Transit (2015) 1(1):35–46 control need to be implemented and the shuttle bus need to be deployed in emergency. These researches show that the bus is a crucial traffic mode in emergent event of URT system, and the passenger flow assignment is the basis of all emergent decisions. Previous researches on passenger flow assignment in URT system can be divided into t (...truncated)