基于最大幅值变分模态分解和均方根熵的滚动轴承故障诊断  

Rolling Bearing Fault Diagnosis Based on Maximum Amplitude Variational Mode Decomposition and Root Mean Square Entropy

作　　者：孟宗[1] 岳建辉 邢婷婷[1,2] 李晶 殷娜 MENG Zong;YUE Jian-hui;XING Ting-ting;LI Jing;YIN Na(Institute of Electrical Engineering,Yanshan University,Qinhuangdao,Hebei 066004,China;Tangshan Polytechnic College,Tangshan,Hebei 063000,China)

机构地区：[1]燕山大学电气工程学院,河北秦皇岛066004 [2]唐山工业职业技术学院,河北唐山063000

出　　处：《计量学报》2020年第4期455-460,共6页Acta Metrologica Sinica

基　　金：国家自然科学基金(51575472);河北省自然科学基金(E2019203448)。

摘　　要：针对变分模态分解中模态个数的设定会对分解结果产生重要影响的问题,提出一种求取最优分解层数的方法,该方法以瞬时频率的幅值特性为依据,通过分析变分模态分解过程中,各分量最大幅值之间的关系来确定最佳分解参数;均方根熵可以反映不同振动信号的能量值,以信号均方根熵为故障特征参量,通过优化支持向量机建立故障分类模型,实现故障模式分类。将基于最大幅值变分模态分解和均方根熵的故障诊断方法应用于滚动轴承实测信号中,实验结果表明基于最大幅值变分模态分解和均方根熵的方法能够有效识别滚动轴承运行状态,识别准确率高达98.75%。The setting of modal number in the variational mode decomposition is very important for the decomposition results,based on this,a method to obtain the optimal decomposition layer number is proposed.The method is based on amplitude characteristics of the instantaneous frequencies and determines the optimal decomposition parameter by analyzing the relationship between the maximum amplitude of each component in the variational mode decomposition process.The root mean square entropy can reflect the energy of different vibration signals and is used as the characteristic parameter of the fault.And a fault classification model is established by optimized support vector machine to realize fault patterns classification.The fault diagnosis method based on maximum amplitude variational mode decomposition and root mean square entropy is applied to the measured signal of rolling bearings.The results show that the method based on maximum amplitude variational mode decomposition and root mean square entropy can identify rolling bearing running state efficiently and realize rolling bearing fault diagnosis.The recognition accuracy of this method is 98.75%.

关 键 词：计量学 滚动轴承 故障诊断 变分模态分解 均方根熵 支持向量机 

分 类 号：TB936[一般工业技术—计量学] TB973[机械工程—测试计量技术及仪器]