基于EMD阈值方法的轴承故障振动信号去噪

针对随机噪声信号影响对有用信息的获取，提出了EMD分解阈值去噪方法，将小波阈值去噪原理应用于经验模态分解（Empirical Mode Decomposition，EMD）中。首先对实际含噪信号进行EMD分解，根据分解后得到的内蕴模态函数（Intrinsic Mode Function，简称IMF分量），采用自适应阈值去噪，进行信号重构，得到消噪后的信号，获取有用信息。将该方法应用于实际工程故障振动信号中分析研究表明，该方法可以获得较高的信噪比，能够对实际信号进行有效的故障特征频率提取，降噪后比降噪前的诊断效果更明显。

De-noising of rolling bearing fault vibration signal based on empirical mode decomposition threshold

Caused by random noise signals in the analysis of actual signal interference, and influence to obtain useful information seriously, in this paper, the wavelet threshold denoising principle is applied to the Empirical Mode Decomposition（EMD）, the EMD threshold denoising method is proposed. Firstly to actual noise signal EMD, according to the decomposition after each Intrinsic Mode Function（IMF） adopts adaptive threshold denoising, then reconstructs signal and gets after de-noising of the signal, then carries on the analysis to obtain useful information. Finally the method and the traditional denoising method is applied to practical engineering fault vibration signal analysis in the comparative study shows that the method of denoising performance is superior, can be obtained the higher signal-to-noise power ratio and the actual signal extraction of fault feature frequency effectively, after de-noising diagnosis effect is obvious.