摘要
为解决石化装置贝叶斯网络(BN)拓扑结构节点状态不确定系统难以分析节点重要度的问题,借助广义集对分析(GSPA)理论探讨节点重要度。针对节点的不同逻辑关系,讨论基于GSPA理论与BN方法集成的节点状态概率表达方式。从同一性、差异性、对立性3方面,提出节点同异反概率重要度分析方法,并将其应用到某反应器火灾爆炸事件的BN拓扑结构中。结果表明:各根节点故障状态概率会影响该反应器火灾爆炸事件发生概率,各根节点联系度同一性、差异性、对立性不同,会产生不同的同异反概率重要度分布情形。
GSPA theory was used to solve the problem that the importance degree of node state uncertain system of BN topology structure on petrochemical device is difficult to analyze. Bearing in mind different logical relations between nodes,a way was discussed for expressing node state probabilities based on BN and general set pair analysis. Taking into account identity,difference,and opposition between nodes,a method was worked out for evaluating probability importance degrees of nodes. The method was applied to BN topological structure for fire and explosion event in a certain reactor. Results show that state probabilities of root nodes affect the probability of reactor fire and explosion event,among other things.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2015年第6期81-85,共5页
China Safety Science Journal
基金
辽宁省自然科学基金资助(2013020137)
辽宁省教育厅科学研究一般项目(L2012060)
关键词
石化装置
广义集对分析(GSPA)
贝叶斯网络(BN)
同异反
概率重要度
联系度
petrochemical device
general set pair analysis （GSPA）
Bayesian network （ BN ）
same different and opposite
probability importance degree
connection degree
