Analysis And Assessment Of Metro Station Operation Vulnerability Due To Passenger Behaviors

China is undergoing a boom period of urban metro construction and operation. Along with numerous metro lines being put into service comes frequent occurrence of metro accidents, which impacts people’s safe travel as well as cities’normal operation. An accident case study indicated that a significant number of metro operation accidents was directly or indirectly caused by passenger’s behaviors and almost all of these accidents occurred at metro stations; however, to date the effects of passenger’s behaviors on metro operation safety has not drawn enough attentions. Therefore, analyzing how to enhance the safety and reliability of metro operation in terms of reducing passenger’s risky behaviors and improving station operation is of great theoretical value and practical significance. Based on the facts that the operation of metro station is vulnerable to the disturbance of passengers’behaviors, the present dissertation, focusing on the effects of passenger’s behaviors, explored the vulnerability of metro station operation from the relationship between behaviors and accidents and from the sensitivity of station operation facing behavioral disturbance. The final aim of this study is to identify and assess the vulnerable scenarios in metro station operation so as to provide metro operation management with improving targets and suggestions.(1) On the basis of a detailed literature review, the overseas and domestic research status of metro operation safety, vulnerability theory and assessment approaches, metro passenger behaviors and their behavioral psychology was summarized and made comments to indicate the shortage of the current research; and then focusing on these shortages and the problems in the practice, the aims, contents and research framework of this dissertation were specified.(2) The conceptions and principals of metro station operation vulnerability (MSOV) were defined. On the basis of establishing the theoretical system of metro system operation vulnerability as a whole, the level and scope of vulnerability study in this context were clarified. Moreover, accidents due to passenger aberrant behavior (PAB) were analyzed to figure out the formation mechanism of PAB accident and its influence factors.(3) An analytical model of PAB accident was established to identify the vulnerable scenarios that lead behaviors to accidents. According to the mechanism of PAB accident, the important influence factors in accident occurrence and development as well as the causality between these factors were confirmed. Based on this, a fuzzy petri net (FPN) model describing the process of PAB accident was established and the inference rules of the FPN model were provided. Matlab was used to fulfill model computation. The vulnerable scenarios with a larger likelihood to cause accidents were identified and ordered according to their vulnerability degrees.(4) PABs were given a full identification and assessment. Through literature review, expert interview and field observation, a metro passenger behavior questionnaire (MPBQ) was developed to investigate the occurrence frequency of PABs, with which was the factor structure of MPBQ was explored. With the identical PAB items, a metro staff estimation questionnaire (MSEQ) was developed to assess the risk level and latent affect objects of each PAB. ANOVA method was adopted to analyze the difference among different groups about the frequency of different PAB categories. A Logistic model was built to explore the relationships between different PAB categories and accident occurrence.(5) The generation mechanisms of two important passenger risky behaviors (PRBs) were analyzed. Based on the extension of adding 4 additional variables (moral norms, past behavior, risk perception and self-identity) into the Theory of Planned behavior (TPB), a theoretical hypothesis model for PRB generation mechanism was put forward. After a typical scenes of the two behaviors were built, a passenger risky behavior questionnaire was developed to measure the psychological cognition variables of PRBs. The independent-samples t test was applied to identify the key beliefs that differed most between behavioral intenders and non-intenders. Using structural equation model (SEM), the theoretical hypothesis model was tested and modified under two behavioral scenes, based on which the psychological cognition of the two PRBs were confirmed.(6) A simulation model of PRB evolution within metro station operation was established to identify the vulnerable scenarios with a surge increase in risky behavior level. A systematic analysis of passenger risky behavior evolution based metro station operation vulnerability (PRB-MSOV) was made from the perspective of system modeling. Then, the causal diagram of the PRB-MSOV system was drawn, through which variables and feedback schemes in the system dynamics (SD) model was defined. The relationships of all the variables within the 6 sub-modules constituting the system and between these sub-modules were described and thus the overall flow graph of PRB-MSOV system was formed. An example simulation based on the established PRB-MSOV SD model was performed by using AnyLogic software and the data from Wukesong station of Beijing Metro. Finally, a simulation platform for sensitivity analysis was built based on the SD model, which allows multiple simulation and comparison. The sensitivity of some important metro station operation elements to the PRB effects was analyzed to identify the vulnerable scenarios. These elements included train time intervals, the maximum passenger flow of peak time, the types of screen doors and agent numbers on platforms.