基于样本分位数散布熵的汽车电机轴承故障诊断方法  

Fault Diagnosis Method of Car Motor Bearing Based on Sample Quantile Dispersion Entropy

作　　者：孙文明[1] 闫晟煜[2] SUN Wenming;YAN Shengyu(Department of Automotive Technology,Zhejiang Agricultural Business College,Shaoxing Zhejiang 312000,China;school of automobile,ChangJ an University,Xi’an 710064,China)

机构地区：[1]浙江农业商贸职业学院汽车技术系,浙江绍兴312000 [2]长安大学汽车学院,西安710064

出　　处：《机械设计与研究》2021年第4期110-114,共5页Machine Design And Research

基　　金：国家自然科学基金资助项目(52002282)。

摘　　要：为保证提取出的汽车电机轴承故障特征对轴承的故障具有更好的可分析,在对散布熵(DE)进行研究的基础上,提出了轴承故障信号复杂度度量的新方法——样本分位数散布熵(SQDE)。SQDE是在DE中引入了样本分位数的概念,克服了DE中嵌入维数参数的选择问题。轴承故障诊断实例表明,相比于DE,SQDE对轴承不同类型和不同程度故障的辨识度有了较大提升,支持向量机(SVM)的故障诊断精度也更高,具有一定的优势。In order to better analyze the extracted car motor bearing fault features,a new sample quantile distribution entropy(SQDE)method of measuring the complexity of bearing fault signalsis proposed in this paper based on the study of DE.SQDE introduces the sample quantile to DE and therefore overcomes the issue of selecting embedded dimension parameter in DE.The experiment results of bearing fault diagnosis show that SQDE improves the identification of faults of various types.The fault diagnosis accuracy of the support vector machine(SVM)is also higher when compared with DE,and has certain advantages.

关 键 词：散布熵 样本分位数 故障诊断 轴承 

分 类 号：TH17[机械工程—机械制造及自动化]