摘要
针对微弱故障信号易被强噪声淹没的难题,提出了一种基于小波包降噪与改进LMD相结合的提取微弱信号特征向量的方法。首先选择恰当的小波基进行小波包分解,再根据计算出的最优小波包树进行信号重构,实现对原始信号的降噪处理。然后对重构的信号进行LMD分解,再计算PF分量的互相关系数和峭度值,减少虚假分量同时增强故障信号幅值。最后对真实的PF分量进行包络谱分析,提取弱信号的故障特征。实例研究结果表明:该方法能够有效地提取出淹没在强噪声中的故障弱信号的特征向量。
Aiming at the problem of weak fault signal with strong noise,a weak fault feature extraction method is proposed based on wavelet packet de-nosing and improved LMD.Firstly,the original fault signal is used to wavelet packet decomposition with appropriate wavelet base and reconstructed according to the calculated optimal wavelet packet tree to realize de-noising for the original signal. Then,the reconstructed signal is decomposed by using LMD method,and calculating the correlation coefficient and kurtosis of PF components in order to reduce the false component and enhance the amplitude of fault signal. Finally,envelope spectrum analysis of real PF components is carried out,and the fault feature of weak signal is extracted. Example results show that the proposed method can effectively extract the feature of fault weak signal under strong noise.
作者
陈长征
魏巍
CHEN Chang-zheng;WEI Wei(School of Mechanical Engineering,Shenyang University of Technology,Liaoning Shenyang110870,China)
出处
《机械设计与制造》
北大核心
2020年第1期165-168,172,共5页
Machinery Design & Manufacture
基金
国家自然科学基金面上项目（51675350）
关键词
弱信号提取
小波包降噪
LMD
包络谱
Weak Signal Extraction
Wavelet Packet De-Noising
LMD
Envelope Spectrum
