基于粗糙集理论的故障诊断规则获取方法研究  

A Method of the Rule Generation in Fault Diagnosis Based on Rough Set

作　　者：周庆敏[1] 李永生[2] 殷晨波[2] 陆金桂[1] 

机构地区：[1]南京工业大学信息科学与技术学院,南京210009 [2]南京工业大学机械与动力工程学院,南京210009

出　　处：《计算机工程与应用》2003年第26期64-65,118,共3页Computer Engineering and Applications

基　　金：江苏省自然科学基金资助项目(编号:BK2001095)

摘　　要：提出了基于粗糙集理论的故障诊断规则获取方法,该方法通过决策系统的可辩识矩阵和分辨函数计算核和所有的约简集,以约简集为初始节点,建立层状节点网络模型,对于各节点,提取符合可信度要求的诊断规则写入规则集中。通过应用表明,方法是确实可行的。A method of rule generation based on Rough set is presented.The method creates the core and a series of the reduction sets from discernibility matrix and discernibility function of the known decision system.With the reduction sets as first nodes,the nodes network model of different reductive levels is developed.Its own rule sets with suitable confidence is reasoned at each node.By an example the correctness of the method is verified.

关 键 词：粗糙集 故障诊断 可辩识矩阵 决策系统 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]