基于复小波变换相位信息的谐波检测算法  

A NOVEL METHOD FOR HARMONICS MEASUREMENT USING PHASE INFORMATION OF COMPLEX WAVELET TRANSFORM

作　　者：赵成勇[1] 何明锋[1] 

机构地区：[1]华北电力大学电气工程学院,河北省保定市071003

出　　处：《中国电机工程学报》2005年第1期38-42,共5页Proceedings of the CSEE

基　　金：原国家电力公司 2002 年度重点学科建设项目。

摘　　要：小波变换因其良好的时频局部化特性,可用来进行谐波分析。但由于不同尺度的小波函数在频带相互混叠,使得现有利用小波系数幅值及在此基础上改进的各种算法都无法实现谐波的准确检测。该文提出利用复小波变换的相位信息来分析谐波的方法,利用改进递归复小波变换在不同尺度时,信号复小波变换系数的相位变化周期来确定信号的谐波频率,进而确定谐波的幅值;与 FFT 和幅值检测法相比,该方法能消除 FFT 算法的频谱泄露,提高了谐波检测的精度,且适用于含有非整数次谐波的信号。通过在 LabVIEW中的实例验证,说明该方法能有效地消除 FFT 算法的频谱泄露和小波函数频带混叠造成的不良影响,有效地提高了谐波检测的精度。Because of excellent time-frequency characteristic of wavelet transform, wavelet transform can be used to analysis harmonics. But the existing algorithm and its improved version based on the amplitude of wavelet transform coefficients can not realize accurate harmonic detection because of frequency band aliasing phenomenon in different scale. Phase information of complex wavelet transforms is proposed to detect harmonics. The harmonics frequency and its amplitude can be gotten by phase period of signal’s coefficient of improved recursive complex wavelet transformation in difference scales. Comparing with the FFT and the amplitude measure arithmetic, the novel method can avoid the spectrum leakage and fence effect in ideal extent and improve the precision of harmonics measurement. The method is valid for non-integer harmonics measurement. Simulations made in LabVIEW have proved that the method can realize accurate detection of harmonics because it can eliminate bad effects of frequency leakage in FFT and aliasing of frequency band in wavelet transform.

关 键 词：复小波 谐波检测 频带 幅值 换相 次谐波 谐波分析 相位信息 混叠 FFT算法 

分 类 号：TM835[电气工程—高电压与绝缘技术] TM711