机械故障诊断中征兆的粒度原理及应用  

The principle of symptomatic granularity and its application in the mechanical fault diagnosis

作　　者：江志农[1] 马晋[1] 蔡镇 

机构地区：[1]北京化工大学机电工程学院,北京100029 [2]中石油辽阳石化公司,辽宁辽阳111003

出　　处：《振动工程学报》2008年第6期630-634,共5页Journal of Vibration Engineering

基　　金：国家高技术研究发展计划(863计划)专题课题(2007AA04Z433);国家自然科学基金重点资助项目(50635010)

摘　　要：提出了一种基于知识粒度原理的故障诊断中征兆参数重要度评价方法。解决了目前故障诊断征兆参数约简和压缩过程中得到的是最小征兆集而不一定是最优最小征兆集的情况,通过对两个实例的计算,很好地求出了最优最小征兆集。A method based on the principle of the knowledge granularity to estimate the significance of symptomatic parameters is presented in this paper. The current problem, that the smallest symptom sets obtained in the process of reducing and compressing the symptomatic parameters of fault diagnosis is not always the smallest and optimal one, has been solved. By calculating on two instances, the smallest and optimal symptom set has been successfully obtained.

关 键 词：粒度熵 故障诊断 征兆参数 重要度 

分 类 号：TH165.3[机械工程—机械制造及自动化] TP39[自动化与计算机技术—计算机应用技术]