摘要
结合北京市某高层住宅小区火灾构建故障树,合理建立各事件之间的逻辑关系,并将故障树转化为贝叶斯网络模型,再对其中的起火部分进行优化,通过贝叶斯网络的推理计算出顶上事件发生的概率和中间事件、基本事件的后验概率。计算得到的火灾发生概率1.46×10^-4起/户,与国内住宅火灾平均发生概率较为接近。
Considering a high-rise residential community in Beijing, the fault tree was built, and the logical relationship between the events were arranged reasonably. The Fault tree was transformed into a Bayesian network model, and then the fire outbreak part was optimized. The probability of the top event as well as posterior probability of intermediate events and basic events was calculated by the Bayesian network. The fire probability calculated is1.46×10^-4,which is closed to the average probability of residential fire in China.
作者
杨世全
黄晓家
谢水波
吴懂礼
程超
赵耀
YANG Shi-quan;HUANG Xiao-jia;XIE Shui-bo;WU Dong-li;CHENG Chao;ZHAO Yao(School of Civil Engineering,University of South China,Hunan Hengyang 421001,China;China Zhongyuan Internation al Engineering Co.,Ltd.,Beijing 100089,China)
出处
《消防科学与技术》
CAS
北大核心
2019年第9期1322-1325,共4页
Fire Science and Technology
关键词
住宅火灾
故障树
贝叶斯网络
residential fire
fault tree
Bayesian network
