面向故障树结构的不精确知识管理  

A Management Algorithm for an Uncertain Knowledge Oriented Fault Tree Structure

作　　者：余彬海[1] 周祖德[1] 袁楚明[1] 胡文彬[1] 

机构地区：[1]华中理工大学机械科学与工程学院

出　　处：《华中理工大学学报》1995年第2期27-29,共3页Journal of Huazhong University of Science and Technology

基　　金：国家"863"高科技CIMS跟踪资助课题

摘　　要：结合故障树在故障诊断中的应用，运用Ｂａｙｅｓ方法，提出了一个面向故障树诊断问题的证据推理算法。此算法能实时、动态地根据征兆信息（即证据）更新树中各事件发生的概率，且计算量小，可应用于基于故障树的实时故障诊断系统。Although the application of the fault tree(FT)in the fault diagnosis of sophisticatedsystems is successful,so far no satisfactory management algorithm for the uncertain knowl-edge oriented fauIt tree structure has been reported. Some cases of the application of thefault tree in fault diagnosis are reviewed and an FT-based evidential reasoning algorithm withthe Bayes method is proposed,which is capable of updating and managing the probabilities ofevents in FT according to evidences and with less computing time,It is applicable to realtime fauIt diagnosis expert systerns.

关 键 词：故障树 不精确知识 故障诊断 专家系统 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置] TP18[自动化与计算机技术—控制科学与工程]