局部放电离散谱干扰的小波包变换熵阈值抑制法

将信号的小波包变换看做通过等带宽的滤波器组，输出是对应频带上的分量，如某一频带上存在离散谱干扰，其小波包变换树节点的信息熵将会明显增大。同时，提出基于小波包分解和重构算法的熵阈值法。结果表明，这种方法具有良好的自适应性，无须事先确定离散谱干扰的数目及其中心频率，干扰抑制能力强，能准确提取局部放电脉冲的相位，对于单一放电类型，可以标定放电量的大小。

AN ENTROPY THRESHOLD METHOD BASED ON WAVELET PACKET DECOMPOSITION FOR DISCRETE SPECTRUM INTERFERENCE (DSI) REJECTION OF PD MONITORING

This paper presents the frequency domain division theory of binary wavelet decomposition and wavelet packet decomposition (WPD) with orthogonal wavelet base frame. Signal's WPD coefficients are treated as outputs of equivalent bandwidth filters with different center frequency. The corresponding WPD entropy values increase sharply when discrete spectrum interference (DSI), frequency spectrum of which is centered by several frequency points, exists in some frequency region. Based on WPD, hence this paper proposes an entropy threshold method (ETM) where entropy is used to determine whether partial discharge (PD) signals are interfered by DSI. Simulation and real data processing have shown that the proposed ETM works efficiently without pre-knowing knowledge of DSI information. This method can extract the phase of PD pulses accurately and calibrate quantity of single type discharge.