【作者】 陈磊；

【导师】 穆建成；

【作者基本信息】 北京交通大学 ， 交通信息工程及控制， 2012， 硕士

【摘要】 为了满足城市轨道交通安全、节能、高效、灵活运输的特点,近年来全自动化地铁在全球轨道交通领域需求逐渐增加,巴黎、新加坡、香港等城市都已引入全自动驾驶地铁(Fully Automatic Operation, FAO)。全自动驾驶并不意味着没有人的参与,而是代表着在系统中,人将处在更高的决策层级和运营层级上。全自动化运营决定了人因行为对系统有更大的影响,人员的失误后果将直接危害系统的安全,因此在FAO中研究人员失误非常重要和迫切的。据统计当今世界上所有人机系统失效中,约有70%-90%直接或间接源于人的因素。在针对FAO进行人误分析时,最主要的问题是缺乏适用于地铁运营领域的人因数据。本论文的目的在于将人因可靠性分析引入FAO系统,并解决缺乏人误数据的现状。本文提出FAO系统的人误分析原则,针对FAO系统ATS操作设计人误分析系统,通过模拟机实验的方式获取人因数据,并应用适于FAO的方法进行分析。论文首先介绍了FAO系统的基本构架,并对人因可靠性分析技术、人的失误、人的绩效行为以国内外研究现状做了简要概述。其次,概述了全自动驾驶ATS人误分析系统总体设计方案,介绍其各组成模块功能需求；引入FAO系统人误分析、障碍设置及后果评价的原则；提出针对FAO系统OCC人员的模拟机实验设计和执行方案。然后,针对事故场景和正常运营场景的特点,从数据来源和人误分析两个主要方面出发,介绍通过模拟机实验获取数据等方法,详细设计系统组成的模拟ATS、事故显示、事故调查、人误分析和人因数据库等功能模块,引入HFACS人误分类体系、基于XML的人误记录、层次分析法、基于S-O-R人误率测算等关键概念,并对关键部分开发实现,应用SQL Server对人因数据库具体设计和实现,为进一步进行人误分析提供数据来源。最后,对具体案例分析,应用层次分析法分析子任务模板,应用THERP+HCR方法测算其人误率,并针对高人误的环节设置障碍,提出改进措施,降低案例中的人误率,对实际的工程应用给予了建议性指导。同时对论文研究做了总结,展望了未来研究的方向和方法。更多还原

【Abstract】 In order to meet the characteristics of urban rail transport such as safety, energy conservation, efficient and flexible, the world has increased the demand for fully automatic operation (FAO) in recent years, and Paris, Singapore, Hong Kong and other cities are driving the introduction of FAO. Automatic driving does not mean no human intervention, but in the system people will be at a higher operational level and decision-making level. In FAO people have greater impact on the system and the consequences of the mistakes people made will directly endanger the safety of the system, so research on human error of FAO is very important and urgent.According to statistics in systems fails all over the world, about70%to90%fails directly or indirectly derived from the human factor. However, to make human reliability analysis (HRA) for FAO, the main problem is the lack of HE data in the area of subway. The purpose of this paper is to introduce HRA to FAO system and solve the status of lack of human error data. This paper first makes principles of HRA in FAO system analysis principles, focuses on ATS system to design the human error analysis system for FAO, and introduces the way through simulator experiments to collect human error data.Firstly, the thesis introduces the system architecture of FAO. The concepts of human reliability analysis, human error, shaping performance and the research over internal and external are briefly referred.Secondly, overviews the system design plan of human error analysis system for ATS system in FAO and describes the functional requirements of its constituent modules, meanwhile introduces the principles of HRA, barriers set up and evaluation of the consequences.Then aim to OCC staff of FAO system design the simulator experiment and introduces the experiment plan.Then, starting from the two main aspects of data sources and human error analysis, introduce the way through simulator experiments and other methods to obtain data for the operation of accident scene and normal scene, also design the system consisting modules in detail such as simulated ATS, accidents interface, accident investigation tool, human error analysis tool and human factor databases. The chapter introduces the key theory such as HFACS human error classification system, XML-based human error record, HTA and S-O-R based human error probability estimates. Develop the key part of the module of the system and use SQL Server to design and implementation the database to provide data sources for the further human error analysis.Finally, analysis of specific cases and use the hierarchical task analysis(HTA) to analysis the sub-goal templates and estimate the human error probability,at the same time give advice to set barriers for the high HEP tasks to reduce the HEP of the case. This can give the practical application of proposed engineering guidance. At the same time make a summary of the thesis, and look forward to the future research directions and methods.。更多还原