基于粗糙集与信息熵的不完备测试信息条件下故障诊断  

Fault diagnosis under condition of incomplete test information based on rough set and information entropy

作　　者：陈利安[1] 肖明清[1] 赵鑫[1] 

机构地区：[1]空军工程大学自动测试系统实验室,西安710038

出　　处：《振动与冲击》2012年第22期24-28,共5页Journal of Vibration and Shock

基　　金：总装备部"十二五"国防预研重点资助项目

摘　　要：针对测试信息不完备条件下故障诊断决策问题,引入粗糙集与信息熵方法。利用决策属性支持度求相对核,以此作为启发信息;通过建立属性知识与信息熵的联系,提出基于信息熵的属性约简方法。结合决策属性支持度和信息熵设计约简算法流程,减少属性集搜索空间,求得最优属性约简集。实例表明,该算法适用于协调不完备决策信息系统与不协调不完备决策信息系统,能解决不完备测试信息条件下故障诊断决策问题。Aiming at a decision-making problem of fault diagnosis under imcomplete test information, an approach with rough set and information entropy was introduced. The decision-making attribute support function was used to find a relative core used as heuristic information, through establishing connections between attribute knowledge and information entropy, an attribute reduction method based on information entropy was put forward. The attribute reduction algorithm was designed based on the decision-making attribute support function and information entropy, with this algorithm the best attribute reduction set was obtained and the attribute set search space was reduced. The example results showed that this algorithm is Suitable to a complete coordinative and an incomplete coordinative decision-making information systems; it can solve a decision-making problem of fault diagnosis under incomplete test information.

关 键 词：粗糙集 信息熵 不完备测试信息决策系统 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]