基于FCM和条件熵的风机属性约简  

Attribute Reduction Method of Fan Based on FCM and Conditional Entropy

作　　者：玄兆燕[1] 封红梅[1] 

机构地区：[1]河北联合大学机械工程学院,河北唐山063000

出　　处：《机械研究与应用》2013年第6期43-45,52,共4页Mechanical Research & Application

摘　　要：矿井风机是矿下作业的重要机械之一,针对矿井风机振动故障诊断中故障征兆的使用问题,基于FCM聚类分析和条件熵相结合的方法提出了一种新的属性约简方法,首先按照特征频率实现对风机故障的聚类,接着结合最大隶属原则和条件熵完成对属性的约简并最终为专家系统提供有效的诊断规则库,从而能更好的实现故障的诊断。Mine ventilator is one of the important mechanical device in the underground work, A method based on FCM and conditional entropy is proposed to realize attribute reduction.It can provide an effective diagnosis expert system rule base, ac-cording to characteristic frequency the clustering of the fan is realized, and then combining with the maximum membership principle and conditional entropy , the attribution reduction is complished , and the valid diagnosis rule base is provided , which can better realize fault diagnosis.

关 键 词：特征频率 FCM 条件熵 属性约简 最大隶属规则 

分 类 号：TH123[机械工程—机械设计及理论]