小波与分形理论的互补性及其在局部放电模式识别中的应用研究

研究应用于电气设备局部放电模式识别及故障诊断的放电特征量提取方法，是电气设备状态维护技术研究中的难题之一。该文从尺度变换的角度，研究了小波与分形理论的互补性；并从局部放电信号小波分解后的能量谱图提取放电特征，用于局部放电模式识别。研究结果表明：选用适当的小波函数和尺度函数，将局部放电信号的逼近信号能量谱和精细的结构能量谱的分形维数作为局部放电模式特征，能够有效地应用于局部放电模式识别中。该项研究结果具有较高的理论和应用价值。

RESEARCH ON COMPLEMENTARITY BETWEEN WAVELET AND FRACTAL THEORY AND RELEVANT APPLICATION IN PD PATTERN RECOGNITION

It is one of the difficult problems encoundtered in the research of condition maintenance techniqued of electrical equipment to extract partial discharge(PD) pattern features used for PD pattern recognition nad fault diagnosis.This paper studies the correlation between wavelet and fractal theory from point of view of the scale transformation and then brings forward the method to extract pattern features from the energy spectrum of decomposed PD signals.The study results show that fractal dimensions of approached signal and fine PD energy spectrum can be used as the PD pattern features with proper wavelet and scale functions.The research results are valuable in theory and application.