Partial discharge classifications: Review of recent progress

Wavelet shrinkage methods are effective for de-noising of partial discharge (PD) detection. Base wavelets are related to distortion of PD signals de-noised by wavelet shrinkage methods. This paper presents a scale dependent wavelet selection scheme for de-noising of PD detection. The scale dependent wavelet selection scheme is called the energy based wavelet selection (EBWS) because an energy criterion is proposed for the scheme. In the proposed energy criterion, a base wavelet is selected as an optimal base wavelet if it can generate an approximation with the largest energy among all base wavelets for selection at each scale. PD high-frequency signals are simulated and PD ultra-high-frequency signals are obtained by experiments in laboratory for de-noising experiments and analysis. In comparison with the correlation-based wavelet selection (CBWS) scheme, the wavelet shrinkage, based on the EBWS, generates significantly smaller waveform distortion and magnitude errors of de-noised PD signals.

Scale dependent wavelet selection for de-noising of partial discharge detection

In this paper, a digital dc PD pulse detection system with bandwidth of 10 kHz - 40 MHz is introduced, which was developed using some artificial intelligence methodologies. Focus is made on digital detection, grouping and classification of random pulse signals generated by PD phenomena at dc voltage. Digital detection is developed only resorting to a band-pass filter, a high-speed digitizer (100 MS/s) and a PC with data processing software. Grouping is realized with feature extraction of pulse waveshapes using equivalent time-frequency method (ETFM), making the 2D parameters plane or 3D parameters space, then using the unsupervised clustering Fuzzy C-Means (FCM) method to achieve fast separation for pulse sequence. And classification resorts to least square support vector machine (LS-SVM) based on a fingerprint, which is derivate form 2D histograms of basic parameters, the discharge magnitude q and the time between discharges Deltat of each sub-group. Field application is made for typical defects of oil-paper insulation under dc voltage. At last, several methods to improve separability of the grouping technique are also given for some special cases, including threshold value grouping, marginal coordinates grouping based on 2D parameters plane and grouping using ETFM preprocessed by wavelet denosing. Experimental results show that the dc PD detection system developed with artificial intelligence methodologies is practical and effective.

Digital detection, grouping and classification of partial discharge signals at DC voltage

In this paper, we present results dealing with the non-uniform pollution carried out under 50 Hz applied voltage on a plane model simulating the 1512 L outdoor insulator largely used by the Algerian Company of Gas and Electric Power (SONELGAZ). Many configurations in non-uniform pollution are studied in the ENP's (Ecole Nationale Polytechnique d'Alger) High Voltage Laboratory in order to analyze the impact of polluted layer distribution on the insulator dielectric performances. The polluted solution has a conductivity of 1.2 mS/cm obtained with distilled water and NaCl. Our investigations are particularly focused on the on line monitoring of both position and width of the contaminated layer. The flashover voltage and the leakage current magnitude have been investigated in order to study the flashover process on this insulating surface. A video apparatus is used to reflect the parallel discharges behavior, appearing when the polluted layer reaches a critical width. Phase angle values between applied voltage and leakage current (LC) signals at the fundamental frequency (50 Hz) are calculated using the Fast Fourier Transform (FFT) spectral analysis. Phase angle measurements indicate that the equivalent impedance of the insulator behaves like RC circuit with a high capacitive effect engendered by the pre-established clean band. This effect decreases when electric discharges occur at a particular voltage level. The Discrete Wavelet Transform (DWT) is adopted for the leakage current decomposition in several time-frequency bands. The STD-MRA (Standard deviation-Multi Resolution Analysis) of these frequency bands is calculated and is employed to choose the most interesting details that detect both position and width increasing of the conducting layer. Reported results show that the pollution surface state and the severity of this conducting layer deposited on insulator surface could be identified from the STD-MRA representation of leakage current frequency bands. It was shown that the high frequency band of the leakage current increases before the final flashover when the polluted layer is located in the middle of the plane model. It was established that a good correlation has been found between the insulator state surface and details of the leakage current obtained through the DWT decomposition. In fact, these details provide relevant information on both position and width of the polluted layer nonuniformly distributed on the insulator surface.

Flashover process and frequency analysis of the leakage current on insulator model under non-uniform pollution conditions

Detection and diagnosis of partial discharge (PD) activity has been widely adopted in electrical plant condition monitoring. Analysis and detection of PD in practical applications is often hampered by noise in the signal. Recent research has shown that the discrete wavelet transform (DWT) is effective in extracting PD pulses from severe noise. One disadvantage, however, is that DWT does not reproduce accurate PD pulse magnitude and pulse shape after thresholding in the presence of strong noise. This paper presents the application of the second generation wavelet transform (SGWT), as an improved algorithm, to extraction of PD pulse from electrical noise. The paper begins with the description of the fundamental theory and structure of SGWT analysis and comparisons with DWT. The method is then applied to both simulated and real world PD data. Results prove that SGWT can significantly improve the effectiveness of PD denoising.

Second generation wavelet transform for data denoising in PD measurement

Recent research has shown that the wavelet transform (WT) can potentially be used to extract partial discharge (PD) pulses from severe noise. However, the method is more complex than the Fourier transform (FT), and requires expertise and experience to be applied to produce its best effect. The authors have previously published algorithms for selection of the most appropriate mother wavelet and for automatic determination of threshold values for applying the WT to PD measurement denoising. The present paper is to present an improved methodology to apply the discrete wavelet transform (DWT) with better denoising effect to PD measurement. Firstly the paper describes the structure of DWT's filter pairs. It then analyses the frequency bands of the wavelet coefficients in approximations and details, and energy distribution of a PD signal along each of the levels following the DWT. Finally a DWT-based denoising method is proposed and justified. Results prove that, with the proposed methodology, in conjunction with the algorithms proposed by the authors to select optimal mother wavelet and threshold values, significant improvement in denoising effect can be achieved.

X. Zhou

Papers

An improved methodology for application of wavelet transform to partial discharge measurement denoising