摘要
滚动轴承振动信号中的强噪声背景对准确提取轴承的故障特征信息具有很大影响,本文提出利用多小波降噪与CEEMD方法相结合,提取滚动轴承的故障特征信号的方法:首先通过多小波方法自适应地消除滚动轴承故障信号中的噪声,然后利用CEEMD进行分解,将分解后的IMF的最大Shannon信息熵值作为判断标准,最大可能地保持原始信号中的故障信息,提取Shannon信息熵值最大的有效IMF进行频谱分析,利用频谱特性提取滚动轴承故障特征。利用数值算例和滚动轴承数据验证了该方法的可行性,为轴承故障诊断提供参考。
Strong noise background in rolling bearing vibration signal has great influence on accurate extraction of bearing fault feature information.This paper presents a method of extracting rolling bearing fault feature signal by combining multiwavelet denoising with CEMD method.Firstly,the fault signal of rolling bearing is eliminated adaptively by multi-wavelet method.Then the noise is decomposed by CEEMD.The maximum Shannon information entropy of decomposed IMF is taken as the criterion to keep the fault information in the original signal as possible.The effective IMF with the maximum Shannon information entropy is extracted for spectrum analysis,and the fault characteristics of rolling bearings are extracted by spectrum characteristics.The feasibility of this method is verified by numerical examples and rolling bearing data,which can provide reference for bearing fault diagnosis.
作者
胡敏
张娟娟
贾松阳
Hu Min;Zhang Juanjuan;Jia Songyang(Luoyang LYC Bearing Co.,Ltd.,Luoyang 471039,China;State Key Laboratory of Aviation Precision Bearings,Luoyang471039,China)
出处
《哈尔滨轴承》
2018年第3期3-8,12,共7页
Journal of Harbin Bearing
关键词
滚动轴承
故障特征
多小波
CEEMD
功率谱
rolling bearing
fault characteristics
multi-wavelet
CEEMD
power spectrum
