基于贝叶斯网络的化工园区地震次生灾害情景分析

摘要

为了研究化工园区内发生地震灾害后的事故演化过程,利用事故链模型对地震次生灾害演化过程进行分析,并将其转化为贝叶斯网络,确定各节点的变量与状态取值范围;通过相关文献及专家经验判断分析,获得贝叶斯网络中各节点的条件概率;依据贝叶斯网络的推理策略,比较不同地震烈度下,各化工事故发生的后验概率值,并探讨应急救援的及时性对地震次生灾害发生概率的影响,从关键要素遏制化工事故的发生,采取针对性的应对措施,尽可能的降低化工事故,有效的降低人员伤亡及财产损失.%To study the accident evolution process after the earthquake disaster in the chemical industry park , the evolution process of secondary disasters of earthquake was analyzed by using the accident chain model , then it was transformed into the Bayesian network , and the valuing ranges of variables and state of each node were determined .The conditional probabilities of each node in the Bayesian network were obtained through the analysis on relevant literatures and expert empirical judg -ment.The posterior probability values of chemical accidents under different earthquake intensities were compared according to the reasoning strategy of Bayesian network , then the timeliness of emergency rescue and its influence on the occurrence probability of secondary disasters of earthquake were discussed .The occurrence of chemical accidents should be contained from the key factors , and the targeted countermeasures should be taken to reduce the chemical accidents as far as possible , so as to reduce the casualties and property loss effectively .