基于故障树和贝叶斯网络的仓储类建筑火灾风险评估

以降低仓储类建筑火灾风险为目的，通过分析影响仓储类建筑火灾发生原因，分析顶层事件与基层事件之间自上而下的因果关系，建立故障树，并将其映射到贝叶斯网络中，通过计算得出事件发生的先验和后验概率，并将两者有效的联系起来。给出各事件发生的重要度指标有结构指标、概率指标和关键指标，分别进行定量分析指出仓储类建筑最有可能存在的风险。文章以2010—2020年十年期间仓储类建筑火灾发生案例为样本对本文提出的模型进行验证。检验结果：文章提出贝叶斯网络模型能够有效评估仓储类建筑风险等级并降低火灾事故发生。

Fire risk assessment of storage type buildings based on fault trees and bayesian networks

With the aim of reducing the risk of fire in storage buildings, the top-down causal relationship between top-level events and grassroots events is analyzed by analyzing the causes affecting the occurrence of fire in storage buildings, establishing a fault tree and mapping it to a Bayesian network,and calculating the prior and posterior probabilities of event occurrence and effectively linking the two. The importance indicators for the occurrence of each event are given as structural indicators, probability indicators and critical indicators, and the quantitative analysis is carried out to point out the most likely risks of storage type buildings respectively. The article validates the model proposed in this paper with a sample of cases of fire occurrence in storage type buildings during the decade of 2010-2020. The results of the test: the Bayesian network model proposed in the paper can effectively assess the risk level of storage buildings and reduce the occurrence of fire accidents.