Application of signal‐adapted wavelet to on‐line partial discharge measurement

Abstract

SUMMARYThis paper develops an approach for optimal selection of signal-adapted wavelet in de-noising. The new method employs Wu's elimination algorithm to construct compacted orthogonal wavelets. The threshold cost function is used as the objective function during choosing of the wavelet. The approach is validated by its application to denoising for on-line partial discharge (PD) measurement. The physical experiments and field tests show that the method improves the suppression of noise for on-line PD measurement with stronger robustness.

Islanding detection in multiple-DG microgrid by utility side current measurement

“…DWT is the multi‐resolution time‐frequency analyzer used for feature extraction in . The DWT has been used in power system analysis of different applications . The aim of the feature extraction is determination of a specific signature in the utility current used for distinguishing between islanding conditions and other events.…”

Investigation of suppressing PD white noise by compound information of complex wavelet transform

“…The amplitude is normalized to 1 and each pulse is numbered as I, II, III, IV as shown in Figure 1(a). For the definition of white noise, two groups of random numbers are produced as white noises by Gauss number algorithm [1][2][3][4][5][6][7], and the amplitudes of white noise respectively are 1 and 1.78. Then the original PD simulation signals are shown in Figure 1(a) and two contaminated signals of different SNR are shown in Figure 1(b) and 1(c).…”

“…It is a very powerful tool to detect insulation faults in electric power devices, especially to gas insulated substations (GIS) and power transformer [1]. One main problem for on-line UHF PD measurement is the suppression of external noise which may greatly reduce the sensitivity of the PD measurement and even corrupt meaningful information [2]. Although UHF techniques are immune to interferences and noises, the results are still not desirable when the noise levels are high.…”

“…WT is a new time-frequency analysis method with good spatial and spectral characteristics at the same time and is very suitable for processing the non-stationary signal. It is becoming popular for the denoising of PD signals and shows strong power [1][2][3][4][5][6][7]. However, it is found that the effectiveness of denoising depends on the mother wavelet chosen.…”

“…WT [2,[5][6][7] and wavelet packet [3][4] have obtained widespread application, especially for complex wavelet transform (CWT) proposed in the literature [1,[8][9][10] has not only the same mathematical principle of processing the non-stationary signal, but also it has better denoising effect. The CWT has particular phase-frequency characteristic that it could adjust the phase-frequency characteristic to make it more adaptive to the signal.…”