基于傅里叶和小波变换的电网谐波检测方法研究摘要本文通过对构建的谐波信号模型用傅里叶变换、小波变换、小波阈值去噪以及基于傅里叶和小波变换相结合的谐波检测方法进行仿真，仿真结果说明：傅里叶变换的双峰谱线修正算法利用被检测频率点左右两侧的最大和次最大的离散频谱谱线幅值可求出各次稳态谐波的幅值和频率。小波变换通过用低通滤波器和高通滤波器可将信号分解成高频部分和低频部分，经过多次分解，从而得到各次谐波的时域和频域大致信息。小波和傅里叶变换相结合的谐波检测算法首先利用小波变换得到重构后的稳态分量和衰减分量，接着对稳态分量用双峰谱线修正算法计算得到各次谐波的幅值和频率。小波去噪和傅里叶变换相结合的谐波检测算法首先利用小波阈值去噪方法去除信号中的噪声，接着对去噪后的信号用双峰谱线修正算法计算得到各次谐波的幅值和频率。实验结果表明傅里叶变换能够反映出信号的频域信息，而小波变换通过其时频特性，可以得到信号的时频特性图，对信号的突变点检测和去除噪声具有较好的效果。结合这两种算法的优势，首先利用小波变换提取信号中的突变点和去除噪声，接着利用FFT能够准确地检测信号中的各次稳态谐波的幅值和频率。关键词：谐波检测小波去噪小波变换傅里叶变换vvTHECHARMBASED丨ONICDETECTIONONFOURIERANDTRANSFORMINPOWERGRIDWAVELETABSTRACTThroughtheestablishmentoftheharmonicsignalmodel,TheHarmonicdetectionmethodbasedoncombiningFouriertransformandWavelettransform^Fouriertransform、Wavelettransform^Waveletthresholddenoisingisusingtoanalysisthemodel.Thesimulationresultsaredescribedbellow.TheamplitudesandfrequencyofharmonicscanbeobtainfromthetwoneighboringspectrallinesintheFouriertransformbimodalspectralcorrectionalgorithmwhichuseweightedaverageoftwolinestocorrecttheamplitude.InWavelettransform,Thesignalisdividedintohighfrequencyandlowjfrequencybylow-passfilterandhigh-passfilter.Throughmultipledecomposition,Allthethetimedomainandfrequencydomainapproximateinformationofharmonicscanbeobtain.InHarmonicdetectionmethodbasedoncombiningFouriertransformandWavelettransform,thereconstructedsteady-statecomponentanddampingcomponentcanbeobtainbyWavelettransform,thenbyusingFouriertransformbimodalspectralcorrectionalgorithmtoanalyzethesignalofwhichhassteady-statecomponenttoextracttheharmonicamplitudeandfrequency.InHarmonicdetectionmethodbasedoncombiningWaveletdenoisingandFouriertransform,thenoisesofsignalscanbereducedbyWaveletthresholddenoising,thenbyusingFouriertransformbimodalspectralcorrectionalgorithmtoanalyzethesignalofwhichhasbeenreducednoisestoextracttheharmonicamplitudeandfrequency.TheresultsshowthattheFouriertransformcanreflectthesignalinthePAGE\*MERGEFORMATVIPAGE\*MERGEFORMAT#frequencydomaininformation