基于改进LMD阈值降噪的滚动轴承故障诊断研究  

Research on Fault Diagnosis of Rolling Bearings Based on Improved LMD Threshold Denoising

作　　者：高峰 胡攀辉[1] 李梦仁 刘海亮 曹红星 李昱良 GAO Feng;HU Panhui;LI Mengren;LIU Hailiang;CAO Hongxing;LI Yuliang(Shanghai Spaceflight Precision Machinery Institute,Shanghai 201600,China;The Sixth Military Representative Office of the Naval Equipment Department in Shanghai,Shanghai 201108,China)

机构地区：[1]上海航天精密机械研究所,上海201600 [2]海军装备部驻上海地区第六军事代表室,上海201108

出　　处：《工程与试验》2024年第3期6-11,30,共7页Engineering and Test

摘　　要：针对滚动轴承振动信号易受噪声干扰从而影响故障诊断精度的问题,提出了一种将改进局部均值分解(LMD)区间阈值降噪算法、多尺度排列熵(MPE)和支持向量机(SVM)相结合的滚动轴承故障诊断方法。采用改进LMD区间阈值降噪算法对信号进行预处理,考虑到去噪信号仍有较强的非线性特性,采用MPE算法构建特征向量集,并将其输入SVM进行故障识别。实测轴承信号分析结果表明,本文所提出的故障诊断方法的故障识别准确率为98.3%,优于其他故障诊断方法。Aiming at the problem that the vibration signals of rolling bearing are susceptible to noise interference,which affects the accuracy of fault diagnosis,a rolling bearing fault diagnosis method combining improved local mean decomposition(LMD)interval threshold denoising algorithm,multi-scale permutation entropy(MPE)and support vector machine(SVM)is proposed.The improved LMD interval threshold denoising algorithm is used for signal preprocessing.Considering that the denoised signal still has strong nonlinear characteristics,the MPE algorithm is used to construct the feature vector set,and the signals are input into SVM for fault identification.The analysis results of the measured bearing signals show that the fault identification accuracy of the proposed fault diagnosis method is 98.3%,which is better than other fault diagnosis methods.

关 键 词：局部均值分解 阈值降噪 滚动轴承 故障诊断 

分 类 号：TH133.3[机械工程—机械制造及自动化]