基于单子带重构改进小波变换的电力系统谐波检测方法

Harmonic detection based on wavelet transform improved by single -band reconstruction

Mallat快速离散小波变换注重总体抗混叠和信号重建,忽略分解分量中的频率混叠抑制,在电力系统谐波检测中产生严重的频率混叠现象。虽然单子带重构小波变换提高了频率分辨率,但是分析结果仍存在一定程度的频率混叠。单子带重构改进小波变换算法,对信号进行分解与重构时,利用快速傅里叶变换(FFT)和快速傅里叶逆变换(IFFT)对各子带信号进行处理。该方法可以克服传统小波变换算法存在频率混叠现象的固有缺陷,仿真计算结果证明了该方法的有效性,为电力系统谐波、间谐波和时变谐波分量的精确检测提供了一种有效手段。

Mallat fast discrete wavelet transform emphasizes the total anti- aliasing and signal reconstruction,but omits the frequency aliasing restraint of components,which produces serious frequency aliasing in power system harmonic detection. Although the single- band reconstruction wavelet transform improves frequency resolution ratio,the experimental results indicate that it restrains frequency aliasing to a certain extent. When wavelet transform improved by single- band reconstruction decomposes and reconstructs signals,it processes band signals with FFT and IFFT to avoid the inherent frequency aliasing of traditional discrete wavelet transform,which is validated by simulative results. It could be applied in harmonic,inter- harmonic and time- varying harmonic detection.