基于失效致因模型的动态风险分析方法设计及应用

发布时间：2018-03-13 21:46

本文选题：动态风险分析　切入点：失效致因模型　出处：《北京交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：在安全苛求系统中,风险分析是系统安全性的评价指标。风险一般与事故发生频率以及造成的损失有关。事故的发生与设备故障,人因错误及组织管理都有关系。本文提出了包含技术、设备,人因及组织管理的动态风险分析方法,并构建了动态风险分析模型以及动态风险分析方法。本篇论文的主要工作包括：一、对安全苛求系统种常见的风险分析方法进行综述,简单介绍了风险的定义以及风险分析在各个领域以及国内外的研究现状,并总结了三代风险分析理论中各自的关注点。二、结合系统动力学模型、Petri网模型以及主体代理模型的特点,从组织管理、人因到技术、设备层面,提出动态风险分析模型框架,构建动态风险分析方法,覆盖了风险分析中建模、验证及仿真分析。同时引入蒙特卡洛实验,完成动态随机的模拟仿真。通过模拟仿真、参数优化实验、参数敏感度分析实验以及参数对比实验,满足用户需求。三、在AnyLogic平台上,借助扩展库开发工具,开发与模型相适应的Extend Petri Nets库和Human Factors库,实现动态风险分析模型建模平台的统一。结合蒙特卡洛实验方法,对真实不确定性系统的模拟进行了仿真实验。四、以铁路系统这个安全苛求系统为例,对管理子系统、人力资源子系统、维护和维修子系统以及运营场景四方面进行建模分析,实现了动态风险分析理论从建模到仿真分析的整个过程,验证了动态分析方法的有效性。
[Abstract]:In safety demanding system, risk analysis is the evaluation index of system safety. The risk is generally related to the frequency of accidents and the losses caused by accidents, and the occurrence of accidents and equipment failures, This paper presents a dynamic risk analysis method including technology, equipment, human factor and organization management, and constructs a dynamic risk analysis model and a dynamic risk analysis method. The main work of this thesis includes:. The main contents are as follows: 1. The common risk analysis methods of safety demanding system are summarized. The definition of risk and the research status of risk analysis in various fields and at home and abroad are briefly introduced, and their respective concerns in the third generation risk analysis theory are summarized. Secondly, according to the characteristics of system dynamics model, Petri net model and agent model, from organization management, human factor to technology and equipment, a dynamic risk analysis model framework is put forward, and a dynamic risk analysis method is constructed. It covers modeling, verification and simulation analysis in risk analysis. At the same time, Monte Carlo experiment is introduced to complete dynamic random simulation. Through simulation simulation, parameter optimization experiment, parameter sensitivity analysis experiment and parameter contrast experiment, Meet the needs of users. Thirdly, on the AnyLogic platform, the Extend Petri Nets library and the Human Factors library, which are suitable for the model, are developed with the help of the extended library development tools to realize the unification of the modeling platform of the dynamic risk analysis model. The simulation experiment of the real uncertain system is carried out. Fourthly, taking the railway system as an example, the management subsystem, the human resource subsystem, the maintenance and maintenance subsystem and the operation scenario are modeled and analyzed. The whole process of dynamic risk analysis theory from modeling to simulation analysis is realized, and the validity of dynamic analysis method is verified.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U298;TP311.52

【参考文献】