经验模态分解在滚动轴承故障诊断中的应用与扩展  

Application and Extension of Empirical Mode Decomposition in Fault Diagnosis for Rolling Bearings

作　　者：张旭[1,2,3] 李超强[1,2,3] 杨柳[1,2,3] 王玉良[1,2,3] 钞仲凯 

机构地区：[1]洛阳轴研科技股份有限公司,河南洛阳471039 [2]河南省高性能轴承技术重点实验室,河南洛阳471039 [3]滚动轴承产业技术创新战略联盟,河南洛阳471039

出　　处：《轴承》2016年第6期59-62,共4页Bearing

摘　　要：针对滚动轴承故障诊断过程中的故障信号特征提取环节,介绍了经验模态分解的基本原理、特性,以及其存在的端点效应和模态混叠等不足,并以此为基础分析了局部均值分解、总体平均经验模态分解和经验小波变换。结果表明:通过改进经验模态分解的曲线拟合方法,加入高斯白噪声及结合小波分析等方法,可以提高EMD特征提取的可靠性及计算速度。Aiming at feature extraction for fault signals during fault diagnosis process for rolling bearings,the basic principles and characteristics of empirical mode decomposition are introduced,and the deficiencies about end effect and mode mixing are discussed. Based on this,the local mean decomposition,ensemble empirical mode decomposition and empirical wavelet transform are analyzed. The results show that the reliability and calculation speed of feature extraction for EMD are improved by improving curve fitting method for empirical mode decomposition and adding white Gaussian noise and combining wavelet analysis.

关 键 词：滚动轴承 故障诊断 经验模态分解 小波分析 

分 类 号：TH133.33[机械工程—机械制造及自动化] TN911.7[电子电信—通信与信息系统]