Abstract: Recently, more and more practical applications show that in excessively noisy electromagnetic environment, using wavelet transform (WT) only may bring waveform distortion. The Fast Fourier Transform (FFT) can suppress sinusoidal interference without concerning the signaltonoise ratio (SNR). Unfortunately, the FFT is almost helpless in suppressing white noise. Thus, combining the advantages of the two transforms, a new method to suppress both white noise and sinusoidal interference is presented. At first, the automatic threshold FFT is used to suppress sinusoidal interference partially and to raise the SNR. Then WT is used to suppress both the residual sinusoidal interference and white noise. Simulation results and practical applications demonstrate that this method can extract partial discharge (PD) signals effectively in excessively noisy electromagnetic environment.

Key words: fast Fourier transform, partial discharge, wavelet transform, signaltonoise ratio

摘要： 近年来，越来越多的实际应用显示，在强噪声环境下，只用小波变换去噪有可能带来波形畸变。相比之下，快速傅里叶变换可以用于去除窄带干扰，但它不能去除白噪声。因此，本文在综合二者优缺点的基础上，提出基于FFT自动阈值滤波法和小波变换的联合去噪技术。即先用FFT自动阈值滤波法去除部分窄带干扰，提高信噪比，继而使用小波变换将白噪声和残余窄带干扰一并滤除。仿真结果和实际应用表明，此法能有效地从较强噪声环境中提取出局部放电信号。

关键词: 快速傅里叶变换, 局部放电, 小波变换, 信噪比