建筑物消防系统可靠性分析

为评估建筑物消防系统的可靠性,首先将系统分为探测报警系统和灭火系统等部分,再用事件树法分析事故发生时消防系统不同部分的反应情况.通过贝叶斯理论及历史统计数据得到系统失效率的不确定性概率分布,建立消防系统可靠性随时间变化的数学模型; 用蒙特卡罗方法模拟求得系统可靠性的时间函数并对模型的不确定性参数进行敏感性分析.该方法将统计数据与经验公式、理论方法相结合,并利用蒙特卡罗方法处理模型中的不确定性,不仅能够有效估计消防系统的可靠性,还可对其他类似系统的可靠性进行分析,并通过敏感性分析为进一步减少估计的不确定性提出合理建议.     

车人碰撞事故再现技术研究进展

车人碰撞事故再现已成为国内外研究热点,提高结果可信性为事故再现的核心。根据事故再现所需痕迹将再现方法分为6类,即"行人抛距"、"车辆制动距离"、"行人损伤"、"车辆变形"、"其他"以及"仿真"。通过分析各类方法的优缺点,提出可综合利用这些方法获得客观、可信的事故再现结果。然后探讨提高车人碰撞事故再现结果可信性的新发展方向:开发国产、高精度的事故再现软件,研究事故再现结果的不确定性问题,以及研究痕迹间关系在事故再现中的应用。而其中仿真所得事故再现结果的不确定性问题、车人碰撞事故再现区间不确定优化方法以及事故现场痕迹间关系为值得期待的新研究内容。     

基于Monte-Carlo方法的长输管道气体扩散模拟研究

目前气体扩散模拟研究多采用流体力学的计算方法,分析气体扩散过程中的动力学特性.有限体积、有限元等方法都需要对事故区域整体进行网格划分,计算过程效率无法满足长输管道事故应急跨区域、多气象以及复杂地形的要求.Monte-Carlo方法利用RAMS预测的平均风场,模拟有限气体粒子在风场中的随机行走特性,有效地弥补了计算效率与网格精度冲突所导致的模拟性能下降的缺点.通过HAVEGE方法收集计算的硬件信息熵形成随机源,修正了以往伪随机数问题,增强了Monte-Carlo方法的计算精度.结果表明Monte-Carlo气体扩散模拟研究方法满足了长输管道事故灾害应急决策的需要.     

LPG沸腾液体扩展蒸气爆炸火球热辐射概率风险评估

分析了火球热辐射的特点,针对LPG沸腾液体扩展蒸气爆炸火球造成的热辐射危害存在不确定性,建立了爆炸火球热辐射风险评估模型,提出了一种基于Monte-Carlo模拟的不确定性处理方法,引入实例计算了LPG沸腾液体扩展蒸气爆炸火球伤害范围、事故风险概率曲线方程和累计概率曲线方程,对于定量评估火球热辐射风险具有重要意义.     

固定顶LNG储罐蒸气云爆炸后果的随机抽样推测

考虑LNG储罐泄漏过程的压力变化,通过压力修正建立了固定顶LNG储罐VCE后果分析模型。在总结前人不确定分析研究成果的基础上,将泄漏孔上方初始液位高度与气云TNT当量系数作为不确定性分析参数。利用随机抽样推测法通过模型计算获得了VCE事故人员死亡半径,并通过数据统计分析获得人员死亡半径大小及其概率密度分布,对有效降低事故风险以及提高LNG管理水平具有价值。     

易燃易爆危险物质泄漏扩散仿真及其应用的研究

以某轻烃厂泄漏事故为研究对象，研制并开发了一个事故后果仿真软件（ＰＣＩＭＳ）．针对仿真过程中存在的问题，如模型的复杂性、参数的不确定性等提出了处理方法．通过具体生产装置的泄漏扩散后果分析，演示了ＰＣＩＭＳ的仿真结果     

基于蒙特卡罗法的圆柱形储罐泄漏时间分析

泄漏持续时间是影响事故后果定量风险评价的关键因素之一.已有的泄漏时间计算多是基于参数确定性考虑,计算结果往往与实际数值有较大的偏差.在分析泄漏源模型的基础上,推导了圆柱形储罐泄漏时间理论计算公式,并从影响因素的不确定性分析出发,选用已知状态变量下的分布函数加以描述.同时,结合蒙特卡罗法解决了输入参数的不确定性.通过实例模拟,揭示了连续泄漏事故场景下储罐内物料发生完全泄漏的时间分布规律,并提出了最大有效泄漏持续时间概念.     

基于故障树分析法的液氨泄漏事故分析

采用故障树分析法对液氨储罐泄漏事故进行分析。通过定性分析找到导致顶事件发生的重要因素;在定量分析过程中,处理含有模糊不确定因素的底事件时,采用模糊集理论与专家综合评判相结合的方法。该方法将专家自然语言通过运用模糊数学的相关知识转化为模糊数,再利用左右模糊排序法转换为模糊失效概率。结合液氨储罐泄漏事故树的分析,给出了相应的防范措施。     

交通事故责任认定之问题和对策初探

《道路交通安全法》确定交 通事故认定书的证据属性,从一 定程度上改变了原有交通事故处 理行政裁决之定性,由此导致对 原有交通事故处理责任认定、申 请复议之旧有程序模式的重构。 但是对于公安机关交通管理部门 履行事故处理职责均无明确规 定,法律规定之不确定性给交通 事故处理实践带来许多新的矛 盾。     

氯化氢事故性泄漏扩散的后果模拟分析

通过对1起BOC(叔丁氧羰基)生产中误操作造成的氯化氢泄漏事故的后果模拟分析,说明扩散模型与后果伤害模型对预测事故伤亡后果和人员应急响应策略选择的作用.首先确定事故中氯化氢的泄漏量和泄漏方式,采用高斯烟团模型模拟扩散过程,利用室内外换气公式,得出室内外氯化氢气体质量浓度随时间变化的关系,然后结合概率函数法得出室内外暴露剂量以及概率变量随时间变化的关系.模拟结果表明,室内外的概率变量都小于-7,因此人员死亡概率为0,与事故实际造成的后果基本相符.因此,选择相应的扩散模型与后果伤害模型对毒气泄漏事故进行后果模拟分析,能定量确定泄漏事故对特定范围内造成的危险程度,预测事故伤亡后果.     

A stochastic approach for risk analysis in vapor cloud explosion

A stochastic approach for evaluating the risk of vapor cloud explosions is proposed in this work. The proposed methodology aims to incorporate the effect of uncertainty into the risk analysis to produce a better overall view for the risk. Some stochastic variables are used to estimate the probability of vapor cloud explosions: frequency of the release, the probability of not having an immediate ignition, the probability of delayed ignition and the probability of a vapor cloud explosion given a delayed ignition, as well as different possible meteorological conditions. These stochastic variables are represented with probability distribution curves. Different curves for the frequencies of releases from process equipment types (steel process pipes, flanges, manual valves, actuated valves, etc.), different equipment diameters and different leak sizes are also used in this analysis. Monte Carlo simulation is performed to obtain the risk as a probability distribution using the Analytic Solver Platform. Then the risk distribution curve obtained by Monte Carlo simulation is used to estimate the probability of satisfying the risk tolerance criterion.     

安全系统可靠性分析方法

讨论了安全系统可靠性分析的最新趋势,即:由对系统的定性分析转向定量分析.给出了可靠性分析的一些方法,包括定性和定量方法.重点研究了马尔可夫故障模型法.针对马尔可夫故障模型法计算量大的缺点,提出了相应的改进方法微马尔可夫故障模型法,分析了此法在可靠性分析中的应用.     

生命周期评价中清单的不确定性分析

生命周期评价中存在着广泛的不确定性,尤其是生命周期评价中清单的不确定性因素会严重影响研究结果的可信性.结合生命周期清单中不确定性的特点,指出不确定性分析的作用.综述目前常用的生命周期清单的不确定性分析方法,包括在不确定性评估中常用的质量指标法、蒙特卡罗(Monte Carlo)法、结合法及常用的敏感性分析方法.指出各种方法存在的优缺点以及在实际中的应用,最后提出生命周期清单中不确定性分析的研究方向.     

地铁人员可用疏散时间模拟

人员可用安全疏散时间是衡量公共场所安全性能的重要指标.为了能从定量风险分析的角度考察公共场所的安全疏散性能,引入蒙特卡罗模拟技术.选取天津小白楼地铁系统作为研究对象;建立极限状态方程以刻画地铁人员可用安全疏散时间;最后,通过Matlab提供的M语言编程,实现了蒙特卡罗模拟,得到了与极限状态方程相对应的概率密度函数,累加分布函数和余补累加分布函数.通过余补累加分布函数定量描述了人员在地铁火灾事故中安全逃生的概率.结果表明,蒙特卡罗模拟技术是描述同时具有参数不确定性和模型不确定性的复杂问题的有力工具.     

参数不确定性对隧道火灾人员安全时间的影响研究

为分析参数不确定性对隧道火灾人员安全时间的影响,根据烟气失能指数建立人员安全时间预测模型。并将所建立的模型与Monte Carlo抽样方法相结合,用概率密度函数描述火灾增长系数、最大热释放速率等参数的不确定性,提出考虑参数不确定性的隧道火灾人员安全时间分析方法。计算分析结果表明:应用该方法可得到隧道火灾人员安全时间的上下限、均值、标准差等;就10 MW以下中小型客车火灾而言,火灾增长系数的不确定性对人员安全时间影响不大;随着热释放速率的增大,人员安全时间的均值逐渐减小,标准差逐渐增大,分布区间逐渐变宽;当热释放速率超过120 MW时,人员安全时间的分布区间、均值、标准差基本趋于稳定。在热释放速率给定条件下,随着通风风速的增加,人员安全时间上下限有所提高且分布区间增宽,同时均值和标准差均有所增加。     

基于ANSYS的可靠度分析方法及应用

通过对可靠度分析的概述理解到现代工程分析中实体模型不确定因素众多,进而带来分析的不准确性。基于有限元分析软件ANSYS提供的概率设计系统(PDS)的概率分析功能,使对结构的概率分析非常容易。根据结构的失效模式来确定结构功能函数,由此建立结构极限状态方程,再运用结构可靠度分析中的蒙特卡洛法(MCS法)利用结构的失效频率来估算其失效概率。在本文中提出了利用ANSYS的概率分析功能结合MCS法进行结构的可靠性分析的方法,并通过一个实例具体说明了利用ANSYS的概率分析功能实现结构的可靠性分析的可行性。     

Uncertainty representation and propagation in quantified risk assessment using fuzzy sets

It is generally acknowledged that there are substantial uncertainties present in any analysis of risk. This paper provides a brief overview of the current techniques used for uncertainty analyses, and highlights their inappropriateness for practical use in the complete risk assessment process. The concept of fuzzy sets as a means for quantifying uncertainty is introduced and a case study demonstrates the application of this method to a simple consequence analysis where parameter uncertainty is considered. The results of this fuzzy analysis are compared with those of a more traditional probabilistic approach using a Monte Carlo simulation. This comparison demonstrates that the novel approach of fuzzy sets is a more appropriate technique due to its non-statistical nature and that the amount of computation required is substantially reduced compared to the traditional probabilistic approach. The versatility of fuzzy set theory suggests that this approach could also be used to quantify other types of uncertainty present in the risk assessment process, including model uncertainty and expert opinion.     

Fuzzy logic approach to calculation of thermal hazard distances in process industries

In this paper, a general procedure to deal with uncertainties in each stage of consequence modeling is presented. In the first part of the procedure, the sources of uncertainty are identified and confirmed by sensitivity analysis for the source term, dispersion, physical effects and consequence analysis. While the second part comprises an application of the fuzzy logic system to each step of the consequence modeling. The proposed procedure is verified by the case study for a pool fire liquefied natural gas (LNG) on water. The results in terms of thermal radiation distances are compared with calculations obtained using the Monte Carlo method and with experimental data. The consequence model based on fuzzy logic approach provides less uncertain and more precise results in comparison to the deterministic consequence model.     

Uncertainty techniques in liquefied natural gas (LNG) dispersion calculations

The dynamic development of the LNG sector increases the risk of major accidents. Uncontrolled releases of LNG during the processes of manufacturing, distribution, storage, and regasification can pose a serious threat to people, facilities, and the environment. Therefore, an important goal is to determine hazard zones and the extent of potential consequences associated with a release of LNG. The key issue is to estimate these with the least level of uncertainty. The largest part of uncertainty comes from the modeling of LNG release sources and performing dispersion calculations. It is connected with the application of different mathematical models, the adoption of a number of simplifying assumptions, approximations, empirical relations, constants, and a lack of knowledge.This paper proposes a general procedure for calculating the release rate and duration time of the LNG release, pool spreading, vaporization, as well as dispersion, taking into consideration the uncertainty. The procedure consists of two parts. The first part concerns the sensitivity analysis to identify the most uncertain parameters of the LNG source term and dispersion models. The second part applies to two techniques used to include the uncertainty aspects of fuzzy sets and the Monte Carlo method for calculating hazard zones. In order to provide a basis for comparison between these two approaches, the shape of the membership functions used in the fuzzy methods are the same as the shape of the probability density function used in the Monte Carlo simulation. The case study, concerning an LNG release, illustrates the application of the proposed techniques.     

Safety performance functions with measurement errors in traffic volume

One of the most common and important predictors in safety performance functions (SPFs) is traffic volume which is known to be measured with uncertainty. Such measurement errors (ME) can attenuate the respective predictors’ effect and also increase dispersion. This paper proposes an approach which involves the use of a ME model based on traffic flow time series data. The model is used in conjunction with the negative binomial SPF to circumvent the bias in predicting the aggregate number of accidents during the time period under study. The proposed approach (denoted by MENB), was compared with the traditional negative binomial (NB) technique by way of Monte Carlo simulation. Furthermore, both approaches were applied to two datasets corresponding to 131 and 130 road segments in British Columbia. The full Bayes method was utilized for parameter estimation, performance evaluation and inference through the use of Markov Chain Monte Carlo (MCMC) techniques. The simulation results showed that MENB has outperformed NB when large measurement errors are present. The goodness-of-fit statistics showed that MENB has provided a slightly better fit to the data. However, in the presence of measurement errors, the NB has underestimated the predicted number of accidents for heavy traffic on long road segments and vice versa. The use of MENB is justified when the variance in volume between years is large otherwise both approaches yield comparable results.