基于经验模态分解的局部放电去噪方法

在简单介绍经验模态分解(EMD)的基础之上，将经验模态分解用于局部放电的信号分析。根据含噪声信号分解后固有模态函数(IMF)的统计特征，提出了一种基于向量阈值的新去噪算法，相比于常规的小波去噪算法，该算法具有形式简单、应用方便灵活、不受傅里叶变换及小波函数选择的限制等特点。实际处理结果及与小波的对比表明，新算法可以有效地抑制白噪声，取得和小波变换几乎一致的效果。

Denoising of Partial Discharge Based on Empirical Mode Decomposition

Based on a brief introduction to empirical mode decomposition (EMD), the theory is applied in analyzing the partial discharge signals. According to the statistical characteristics of intrinsic mode function (IMF) , a new denoising algorithm based on the vector threshold is proposed. Compared with the conventional wavelet based denoising algorithms, the new algorithm is simpler, more flexible, and not limited by the prerequisites of Fourier transform. The results obtained from processing and comparison with the wavelet demonstrate that the new algorithm is capable of effectively suppressing white noise and agrees fairly well with the wavelet based denoising one.