基于小波自相关函数的电磁干扰有源滤波方法研究

Noise Suppression Method for Improvement of Power Quality Using Wavelet Transforms

针对电力电子电路在换流过程所产生高频开关噪声的各项特征,提出了基于小波自相关函数的电磁干扰有源滤波新方法.该方法运用小波多分辨分析实现了电网侧开关噪声的快速提取,再根据复小波变换获得噪声的有效幅度谱和相位谱信息,并结合自相关函数辨识出干扰信号的主频,从而建立了一种多层次处理和多层次描述的噪声辨识方法.研究结果表明,该方案能够克服传统干扰抵消抑制技术中频谱泄露的问题,提高检测精度,检测信号的最大绝对误差不超过2%,噪声的衰减平均达20 dB以上,因而有源抑制为高频开关噪声提供了新的思路.

To reduce the high frequency switching noise generated by power electronics equipment, a novel noise suppression method using wavelet transforms and auto-correlation function is proposed. The switching noise is extracted by wavelet multiresolusion analysis, the enough features of signal in frequency and time domains are obtained by using arguments and magnitudes of complex wavelet transform as well as main frequency evaluation of auto-correlation function, thus the multi-level process and multi-level description method for disturbance identification is estab lished. The simulation results show that the proposed method can not only overcome the defect of classical noise cancellation method that may has the spectrum leakage and fence effect, but also improve the precision of noise measurement, and the maximum absolute error is no more than 2%, meanwhile, the average noise attenuation is larger than 20 dB. This new method provides another concept for the active cancellation of high frequency switching noise.