提取绝缘子泄漏电流特征时母小波的选择

Mother Wavelet Selection for Feature Extraction of Insulator Leakage Current

母小波的选择对提取绝缘子泄漏电流中的突变成分具有重要作用。根据对大量泄漏电流信号的分析,得出小波分解后的细节的能量越集中,所采用的母小波越适合提取泄漏电流的突变成分。据此采用熵作为评价母小波优劣的方法,并给出了泄漏电流熵的计算过程,进而给出了选择最佳母小波的步骤。以高压试验室采集到的泄漏电流为分析对象,以db(Daubechies)小波系为待选母小波集合,根据各小波分解泄漏电流之后的熵值,确定了db2小波为最佳母小波。采用不同母小波对泄漏电流进行了分析,结果显示db2小波优于其他小波。从而得出结论:给出的判断最佳母小波的方法是有效的。

The mother wavelet selection influences the extraction of the abrupt change components in the insulator leakage current greatly.Based on the analysis of many leakage currents,the more concentrated the details energy after wavelet decomposition,the more the mother wavelet was fit to extract the abrupt change components in the leakage current.Entropy was used as the method to evaluate the mother wavelet.The calculation process of the entropy of the leakage current was provided,and then the steps of the optimal mother wavelet selection were given.The leakage current collected in the high voltage laboratory was taken as the analysis object,the db（Daubechies） family was taken as the potential candidate set,based on the entropies calculated after each wavelet decomposition,the db2 wavelet was determined as the optimal mother wavelet.The leakage current was analyzed by the different mother wavelets.The results showed that the db2 wavelet is better than others.The conclusion was that the proposed method to determine optimal mother wavelet is effective.