基于GSPABN的拓扑结构系统状态动态分析  

Dynamic Analysis of Topological Structure State Based on GSPA-BN

作　　者：汤规成[1,2] 许开立[1] 李德顺[1,3,4] 刘家喜[1] 韩恩慧 

机构地区：[1]东北大学资源与土木工程学院,沈阳110819 [2]中国石油辽河石化公司,辽宁盘锦124002 [3]沈阳鼓风机集团股份有限公司安技环保部,沈阳110425 [4]沈阳理工大学环境与化学工程学院,沈阳110159

出　　处：《工业安全与环保》2015年第3期58-60,共3页Industrial Safety and Environmental Protection

基　　金：辽宁省自然科学基金(2013020137);辽宁省教育厅科学研究一般项目(L2012060)

摘　　要：介绍了贝叶斯网络(BN)应用现状及其优势,引入GSPA理论,提出了基于GSPA-BN拓扑结构节点状态分析模型,讨论了节点状态联系度及其重要度表示方法,提出了BN拓扑结构节点态势概念。结果表明,以构建的锅炉缺水烧干事件的BN拓扑结构为例,得到了各根节点重要度排序;同时得到了叶节点、中间节点状态联系度及其态势等级,且中间状态的变化会对其节点态势及其等级产生一定的影响,甚至发生节点态势逆转现象,为动态了解BN拓扑结构节点态势发展奠定了理论分析依据。The application status and advantages of Bayesian network are introduced,the node state analysis model of bayesian network topology based on GSPA is proposed,the state of the node connectivity and the expression method of its important degree are discussed and the node trend concept of the Bayesian network topology is also put forward. Application results show that the important degree rankings of all the root nodes are obtained in the Bayesian network topological structure of the boiler water boil dry event; the state connection of the leaf node and the intermediate node and its trend level are obtained and the change of the intermediate state will have some certain impact on the corresponding node situation and the node level,even the situation is reversed,which provides the theoretical analysis basis for the dynamic development situation of node state of Bayesian network topology.

关 键 词：广义集对分析 贝叶斯网络 节点重要度 联系度 节点态势 

分 类 号：X913[环境科学与工程—安全科学]