Research on Bayesian network algorithm Based on the Hierarchical Module Elimination Method

摘要：

　　Advantages and disadvantages of the fault tree and bayesian network are discussed.Considering that node connection forms of the Bayesian network are confluent connections in most cases,a new algorithm,namely hierarchy module elimination method,is proposed.Firstly,a new module consists of constructed layers; secondly,probabilities of the new module in various state are calculated respectively; At last,probabilities of the leaf node in various states are obtained.The examples show that the probability of leaf node from the hierarchical module elimination is the same as the probability of leaf node from bucket elimination method,so the hierarchical module is suitable for calculating probabilities of device fault in various states.

关键词： fault tree hierarchy module fault probability

作者: LI De-shun Tang Gui-cheng XU Kai-li GONG Bo

作者单位: School of Environmental & Chemical Engineering,Shenyang Ligong University,shenyang110159,China School of Resources & Civil Engineering,Northeastern University,shenyang110819,China;Liaohe petroche School of Resources & Civil Engineering,Northeastern University,shenyang110819,China School of Resources & Civil Engineering,Northeastern University,shenyang110819,China;School of Chemi

会议类型: 国内会议

会议名称: 2012(沈阳)国际安全科学与技术学术研讨会

会议地点: 沈阳

会议语种:英文

页码: 80-84

在线出版日期: 2012-11-11（万方平台首次上网日期，不代表论文的发表时间）