极值理论在高速铁路运营安全风险分析中的应用研究
发布时间：2018-08-24 14:44
【摘要】：随着我国高速铁路技术的飞速发展，安全这一主题在高速铁路运营管理中越来越突出。2012年起，我国铁路全面引入基于风险的安全管理机制，对适用于我国高速铁路运营现状的风险分析方法的研究也成为热门课题。 目前，铁路运输从人-机-环境-管理等方面采用了先进可靠的新技术确保运营安全，尤其是高速铁路集成了诸多领域的现代高新技术，因此事故发生概率小，但由于列车速度的大大提高，一旦发生就会造成极大损失。这种特性即为低频率—高损失特性，，体现在数据上就是损失数据不服从正态分布，而是具有明显的厚尾特征。 本论文引进极值理论中的POT模型，能够针对铁路重大以上事故（尤其是高速铁路事故）这种低频率—高损失的特性进行风险估计与分析，在选定一个合理的较大阈值μ的情况下，针对数据样本中超过阈值的数据，拟合得出损失分布函数，并在给定的置信水平之下计算得出风险价值VaR P，从而对事故损失做出估计。 本论文在调研分析风险管理理论在国内外的发展现状，以及风险管理基本概念、模型与方法的基础上，研究了风险管理技术应用于中国高速铁路运营安全的基本方法，给出适用于我国高速铁路运营安全的风险分析的一般流程和模型。并结合研究课题“中国高速铁路运营安全风险分析和控制”，对目前高速铁路运营安全风险进行了简要分析。 作为重点研究内容，本论文在研究风险分析数学理论的基础上，首次提出了高速铁路事故低频率—高损失特性的理念，并提出了高速铁路基于低频率—高损失特性的高速铁路运营安全风险分析模型，即POT-GPD方法模型。 基于POT-GPD模型，以运营数据为基础，分别对某运营单位客车脱轨事故造成的经济损失和人员死亡的历史数据、某运营单位货车脱轨事故造成的经济损失数据进行数值模拟，分别选取不同的阈值，计算得到3个数据样本各自的损失分布函数，并分别给定置信水平P=95%和P=99%，计算得到风险价值VaR P值。 最后，本论文通过计算得到的风险价值，结合对样本数据的分析，对阈值μ的选取的合理性进行检验。针对不同的阈值μ，通过样本数据分别进行拟合，分析阈值选取的合理性，并对阈值μ的选取做出建议。
[Abstract]:With the rapid development of high-speed railway technology in China, the theme of safety is becoming more and more prominent in high-speed railway operation and management. Since 2012, China Railway has fully introduced risk-based safety management mechanism. The research on the risk analysis method which is suitable for the operation of high-speed railway in China has also become a hot topic. At present, railway transportation adopts advanced and reliable new technology in man-machine-environment management to ensure operation safety, especially high-speed railway integrates modern high and new technology in many fields, so the probability of accidents is small. However, due to the great increase of train speed, once it happens, it will cause great losses. This property is characterized by low frequency and high loss, which is reflected in the fact that the loss data is not obedient to normal distribution, but has obvious characteristics of thick tail. In this paper, the POT model of extreme value theory is introduced, which can estimate and analyze the risk of the low frequency-high loss characteristic of the major railway accidents (especially high-speed railway accidents). In the case of a reasonable larger threshold 渭, the loss distribution function is obtained by fitting the data exceeding the threshold value in the data sample, and the risk value VaR P is calculated under a given confidence level, and the accident loss is estimated. Based on the investigation and analysis of the development of risk management theory at home and abroad, as well as the basic concepts, models and methods of risk management, this paper studies the basic methods of applying risk management technology to the operation safety of high-speed railway in China. The general process and model of risk analysis for high-speed railway operation safety in China are presented. Combined with the research topic "risk analysis and control of high speed railway operation safety in China", this paper briefly analyzes the operational safety risk of high speed railway at present. On the basis of studying the mathematical theory of risk analysis, the concept of low frequency-high loss characteristics of high-speed railway accidents is put forward for the first time. Based on the characteristics of low frequency and high loss, a risk analysis model of high speed railway operation safety, called POT-GPD method, is proposed. Based on POT-GPD model and operation data, the economic loss and death caused by train derailment in a certain operating unit and the economic loss data caused by derailment of freight car in a certain operating unit are simulated numerically respectively. The loss distribution functions of the three data samples are calculated by selecting different thresholds, and the VaR P value of the risk value is calculated by giving the confidence level PQ 95% and PU 99% respectively. Finally, this paper tests the rationality of the selection of threshold 渭 by calculating the risk value and analyzing the sample data. According to different threshold 渭, the rationality of threshold selection is analyzed by fitting the sample data, and suggestions are made on the selection of threshold 渭.
【学位授予单位】：中国铁道科学研究院
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：U298
本文编号：2201142
[Abstract]:With the rapid development of high-speed railway technology in China, the theme of safety is becoming more and more prominent in high-speed railway operation and management. Since 2012, China Railway has fully introduced risk-based safety management mechanism. The research on the risk analysis method which is suitable for the operation of high-speed railway in China has also become a hot topic. At present, railway transportation adopts advanced and reliable new technology in man-machine-environment management to ensure operation safety, especially high-speed railway integrates modern high and new technology in many fields, so the probability of accidents is small. However, due to the great increase of train speed, once it happens, it will cause great losses. This property is characterized by low frequency and high loss, which is reflected in the fact that the loss data is not obedient to normal distribution, but has obvious characteristics of thick tail. In this paper, the POT model of extreme value theory is introduced, which can estimate and analyze the risk of the low frequency-high loss characteristic of the major railway accidents (especially high-speed railway accidents). In the case of a reasonable larger threshold 渭, the loss distribution function is obtained by fitting the data exceeding the threshold value in the data sample, and the risk value VaR P is calculated under a given confidence level, and the accident loss is estimated. Based on the investigation and analysis of the development of risk management theory at home and abroad, as well as the basic concepts, models and methods of risk management, this paper studies the basic methods of applying risk management technology to the operation safety of high-speed railway in China. The general process and model of risk analysis for high-speed railway operation safety in China are presented. Combined with the research topic "risk analysis and control of high speed railway operation safety in China", this paper briefly analyzes the operational safety risk of high speed railway at present. On the basis of studying the mathematical theory of risk analysis, the concept of low frequency-high loss characteristics of high-speed railway accidents is put forward for the first time. Based on the characteristics of low frequency and high loss, a risk analysis model of high speed railway operation safety, called POT-GPD method, is proposed. Based on POT-GPD model and operation data, the economic loss and death caused by train derailment in a certain operating unit and the economic loss data caused by derailment of freight car in a certain operating unit are simulated numerically respectively. The loss distribution functions of the three data samples are calculated by selecting different thresholds, and the VaR P value of the risk value is calculated by giving the confidence level PQ 95% and PU 99% respectively. Finally, this paper tests the rationality of the selection of threshold 渭 by calculating the risk value and analyzing the sample data. According to different threshold 渭, the rationality of threshold selection is analyzed by fitting the sample data, and suggestions are made on the selection of threshold 渭.
【学位授予单位】：中国铁道科学研究院
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：U298
【参考文献】
相关期刊论文 前10条
1 刘秀娟;杨双利;;铁路道口事故经济损失评估的探讨[J];北方交通大学学报;1991年02期
2 卓志;王伟哲;;巨灾风险厚尾分布:POT模型及其应用[J];保险研究;2011年08期
3 周孝华;张燕;;一种新的风险价值(VaR)计算方法及其应用研究[J];管理学报;2008年06期
4 李宝宝;王言峰;;基于CVaR-POT模型的我国银行业操作风险度量研究[J];华东经济管理;2011年07期
5 田新时,毛洪云;基于POT模型的风险价值估计[J];华中科技大学学报(社会科学版);2003年05期
6 李纲,杨辉耀,郭海燕;基于极值理论的风险价值度量[J];决策借鉴;2002年05期
7 刘睿;詹原瑞;刘家鹏;;基于贝叶斯MCMC的POT模型——低频高损的操作风险度量[J];管理科学;2007年03期
8 李金昌;黄劲松;;风险理论发展的比较分析[J];经济学家;2006年02期
9 武剑;;极值理论下的风险度量模型[J];科技信息;2011年36期
10 高扬;雒旭峰;;基于Bowtie模型的机场安全风险分析[J];中国安全生产科学技术;2009年05期
本文编号：2201142
本文链接：https://www.wllwen.com/kejilunwen/anquangongcheng/2201142.html
