LMD-ICA联合降噪方法在滚动轴承故障诊断中的应用

发布时间：2018-02-28 20:52

本文关键词： 独立分量分析局部均值分解降噪滚动轴承故障诊断　出处：《中国机械工程》2016年07期 　论文类型：期刊论文

【摘要】：针对经典独立分量分析(ICA)只能应用于观测源数不少于信号源数的超定盲源分离问题,提出局部均值分解和ICA相结合的欠定盲源分离新方法。该方法将采集的单通道振动信号进行局部均值分解,基于互相关准则对分解的分量进行重组,构建虚拟噪声通道;将虚拟噪声通道与振动信号作为盲源分离的信号输入,采用基于负熵的FastICA算法实现信号源和噪声的分离,从而达到降噪目的。将该方法应用于滚动轴承故障信号,频谱分析结果表明,该方法处理后的信号中噪声得到一定程度滤除,频谱中毛刺更少,故障特征频率更加明显,有利于故障特征的提取,实验分析证明了该方法的有效性。
[Abstract]:The classical independent component analysis (ICA) can only be applied to the problem of over-determined blind source separation where the number of observed sources is not less than the number of signal sources. A new method of undetermined blind source separation, which combines local mean decomposition and ICA, is proposed, which decomposes the single channel vibration signal into local mean decomposition, recombines the decomposed components based on cross-correlation criterion, and constructs virtual noise channels. The virtual noise channel and vibration signal are used as the input of blind source, and the FastICA algorithm based on negative entropy is used to separate the signal source and noise, so as to reduce the noise. The method is applied to the fault signal of rolling bearing. The results of spectrum analysis show that the noise in the signal processed by this method is filtered to a certain extent, the burr in the spectrum is less and the frequency of fault feature is more obvious, which is beneficial to the extraction of fault features. The experimental results show that the method is effective.
【作者单位】：北京信息科技大学现代测控技术教育部重点实验室;
【基金】：国家自然科学基金资助项目(51275052) 北京市自然科学基金资助重点项目(31311002) 北京市教委科研计划资助重点项目(KZ201311232036)
【分类号】：TH133.33

【相似文献】