LMD-ICA联合降噪方法在滚动轴承故障诊断中的应用题稿.docx

LMD-ICA联合降噪方法在滚动轴承故障诊断中的应用研究 卞家磊 朱春梅 蒋章雷 吕俊燕 北京信息科技大学 现代测控技术教育部重点实验室，北京，100192 摘要：针对经典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（LMD）和独立分量分析（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。 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 中图分类号：TH212；TH213.3 Application of LMD-ICA to Fault Diagnosis of Rolling Bearings BIAN Jia-lei，ZHU Chun-mei，JIANG Zhang-lei，LV Jun-yan Beijing Key Laboratory on Measurement and Control of Mechanical Electrical System, Beijing Information Science and Technology University, Beijing, 100192, China; Abstract：The classical ICA can only be applied to Overdetermined blind source separation problem which means the source of the observed number should be not less than of signal source , According to this feature , a new method of local mean decomposition (LMD) combined with independent component analysis (ICA) is proposed . With the approach , collected single-channel vibration signal is first operated with LMD, then rearrange each component to build a virtual channel noise based on cross-correlation criterion , input the virtual channel noise with collected signal into ICA , using FastICA algorithm based on negative entropy to realize the separation between source signal and noise signal so as to achieve the noise reduction purpose. At last , use spectrum analysis method to compare the two signals before and after noise reduction, the noise of the signal is filtered out in a certain degree, and the spectrum is less burr, and the fault characteristic frequency is more obvious, which is advantageous for the fault feature extraction,experimental analysis prove that the new denoising method proposed in this paper is valid. Key words ：ICA；LMD；Noise reduction；Rolling bearings；Fault diagnosis 0 引言 振动信号分析是实现设备运行状态监测与故障诊断的重要方法，信号降噪又是振动信号分析中重要的步骤。传统的振动信号降噪方法通常使用滤波器设置不同通带，如低通滤波，高通滤波，带通滤波等，这种方法本质上是利用滤波