【摘要】 为了提取故障暂态信号的相位谱特征以进行配电网单相接地故障选线,引入了零序电流解析信号的小波包分解。信号的解析表达形式可以更好地描述信号的幅值、相位和瞬时频率随时间变化的规律;它有着比原信号增大一倍的幅值谱、和原信号相同的相位谱和瞬时频率。小波包克服了小波变换不能对高频序列进一步分解的缺陷,在时域和频域上皆具有良好的局部化特性,可以聚焦被分析信号不同频带的时域特性。通过对构造的零序电流解析信号进行小波包分解,提取在适当频带的信号相位谱,依据故障线路和非故障线路相位谱的不同进行故障选线。理论分析和仿真结果表明:这种基于零序电流解析信号的小波包分解所提取暂态信号相位谱的选线判据具有很高的灵敏性和可靠性。更多还原

【Abstract】 To extract the phase spectrum features of thepost-fault transient signals to locate fault line under singlephase to ground fault in distribution network, thedecom-position of wavelet packets of the analytical signals of zero sequence current is applied. The time dependent regularity of amplitudes, phases and the instantaneous frequencies of the signals can be better described by analytical expression of the signals. The analytical signals possess the same phase spectrum and instantaneous frequency as the original real signals, but its amplitude spectrum is double sized than the original realsignals. The wavelet packets possess good localizationcharacteristics in time-domain and frequency-domain. There-fore, the time-domain characteristics of the analyzedsignals at different frequency bands can be focused and thedefects of wavelet transform, i.e., the high frequency sequence cannot be further decomposed, is overcome. By decomposing the structured analytical signals of zero sequence current bywavelet packlets and extracting the phase spectrum of thesignals at proper frequency band, the fault line choosing can be carried out according to the differences of the phase spectrums between fault line and no-fault line. The results of analyses and simulation show that the criteria have satisfactory sensitivity and reliability.更多还原