模糊DBN的室内燃气泄漏动态风险评估方法研究  

Research on dynamic risk assessment method of indoor gas leakage based on fuzzy DBN

作　　者：吕良海 梁艺苑 张淏彬 白永强 LYU Lianghai;LIANG Yiyuan;ZHANG Haobin;BAI Yongqiang(Institute of Urban Safety and Environmental Science,Beijing Academy of Science and Technology,Beijing 100054,China)

机构地区：[1]北京市科学技术研究院城市安全与环境科学研究所,北京100054

出　　处：《安全与环境学报》2024年第4期1337-1345,共9页Journal of Safety and Environment

基　　金：北京市科技计划课题(Z221100005222024);北京市科学技术研究院创新工程课题(20230101)。

摘　　要：为有效分析并评估室内燃气泄漏风险,运用蝴蝶结模型对室内燃气事故危险源进行识别;并利用模糊集理论改进动态贝叶斯模型,弥补因数据缺失带来的误差,实现风险评估从静态到动态的转变,从而构建一种基于蝴蝶结(Bow-Tie,BT)模型模糊动态贝叶斯网络(Dynamic Bayesian Network,DBN)的室内燃气事故动态风险评估方法,并结合实际案例验证模型有效性和可行性。结果表明:依据该模型得到的关键风险因子能够为居民燃气安全风险防控提供参考;同时,该方法能够分析原因事件失效后各事故后果发生概率在各时间片的变化,模拟结果与实际相吻合。To analyze the main causes of indoor gas leakage and the dynamic changes in the probability of different accident consequences that may occur after leakage,this paper establishes a Bow-Tie(BT)model for indoor gas leakage and uses the mapping algorithm of Bayesian networks to convert it into Dynamic Bayesian Network(DBN),making up for the shortcomings of previous static risk assessment.This method uses expert evaluation and trapezoidal fuzzy numbers to calculate the prior probability of basic events,and obtains the prior probability of barrier nodes through reference data;Using the reasoning and inverse inference functions of DBN,the failure probability of indoor gas leakage and its consequences as well as the posterior probability of each basic event are obtained;Use a prior probability and a posterior probability to calculate the sensitivity of each node,obtain the key risk factors leading to gas leakage through comparison,and propose targeted preventive measures based on the analysis results.Besides,the hidden danger is set as an evidence node to compare the dynamic changes of gas leakage risk when there is no evidence node and when the hidden danger exists.Finally,the model is applied to practical cases,which are divided into four nodes according to the time of the accident occurrence.The accident causes are input into the model in chronological order as evidence nodes,and the dynamic changes in the probability of gas leakage and the consequences of each accident are analyzed.The results show that the key risk factors obtained from the model can provide a reference for the prevention and control of residential gas safety risks;Finding hidden dangers through dynamic analysis will gradually increase the risk of indoor gas leakage over time;According to the actual case,the results obtained are consistent with the actual situation,verifying the feasibility of the model.

关 键 词：安全工程 动态风险评估 蝴蝶结模型 模糊集理论 动态贝叶斯网络 燃气泄漏 

分 类 号：X932[环境科学与工程—安全科学]