CAJ | 학술논문

文章提出了一种基于粗糙集理论的改进DMI(改进可区分矩阵)属性约简算法和已建立的故障树相结合的方法来解决压制观瞄装置故障诊断专家系统因经验数据难以识别而产生的知识获取的瓶颈问题.采用改进DMI属性约简算法导出诊断规则,对应故障树的顶事件;建立压制观瞄装置故障树,得到最小割集,并由其转化为规则,实现知识获取.实际应用结果表明,该方法既保证了数据的约简效率,又能够准确地找出故障原因,避免了分析错乱的问题.
문장제출료일충기우조조집이론적개진DMI(개진가구분구진)속성약간산법화이건립적고장수상결합적방법래해결압제관묘장치고장진단전가계통인경험수거난이식별이산생적지식획취적병경문제.채용개진DMI속성약간산법도출진단규칙,대응고장수적정사건;건립압제관묘장치고장수,득도최소할집,병유기전화위규칙,실현지식획취.실제응용결과표명,해방법기보증료수거적약간효솔,우능구준학지조출고장원인,피면료분석착란적문제.
Affected by the control system of rules and empirical data are difficult to identify,the acquisition of knowledge has be-come a big problem that need to be solved in the designing of fault diagnosis expert system.This paper presented an improved DMI algorithm of attribute reduction which was based on the Rough Set Theory,and used it to combine with the Fault Tree to solve the bottleneck problem in the knowledge acquisition,which caused by the difficulty of identifying the empirical data in the fault diag-nosis of the Suppress Sighting Device .The improved DMI algorithm of attribute reduction was used to export diagnosis rules so as to correspond top event of the fault tree;The Fault Tree was established to analysis and transform the minimal cut sets to rules,so as to realize the knowledge acquisition.The practical algorithm of case indicated that this method not only keeps the efficiency of the reduction,but also find the reason of failure exactly, so as to avoid the problem of deranging while we are analyzing.