Online partial discharge signal conditioning for φ – q – n representation under noisy condition in cable systems

Abstract

Almost all on-site measured partial discharge (PD) signals in cable networks suffer from the contributions of reflected PD pulses and noise which both decrease precision of measurement. This study explores an online digital signal conditioning procedure for PD signals being measured under noisy conditions in cable networks, that is, proposing a novel procedure for online preparing clean j-q-n representation of PD signals with neither directly applying de-noising methods and nor requiring voltage waveform (for calculation of j). In order to demonstrate the accuracy of the results, they are compared with noise-free conditions, where successes of the proposed method have been proved.

Modelling and experimental validation of EM wave propagation due to PDs in XLPE cables

“…3. This current produces an electric field according to (1), which has components in the a ρ and a z directions of the corresponding cylindrical coordinate system. This affirms the presence of electric fields in the directions a x , a y and a z during the occurrence of a discharge in the radial direction of the cable.…”

“…The power cable is one of the major components in a power system network. The existence of minor voids in the insulation is a pervasive defect leading to an extension of local degradation and complete failure of the insulation [1][2][3]. Partial discharge (PD) due to deformations is a major source of deterioration in cable insulation, which can result in cost-intensive repairs and possibly extended outages.…”

Feature extraction of GIS partial discharge signal based on S‐transform and singular value decomposition

“…Feature extraction methods are based on the PD analysis modes. In PRPD mode, the PD is described using a pulse count N , against charge amount Q and phase angle diagram [7, 8]. Then statistical parameters are extracted from the diagram [6].…”

“…Time–frequency analysis techniques such as fast Fourier transform (FFT), wavelet transform and wavelet packet transform (WPT) were adopted to characterise the UHF waveform. Based on these techniques, some feature parameters were extracted and used as input parameters of classifiers to realise PD classification and separation [3, 8–16]. In [9], WPT was used to decompose the PD UHF signals into multiple scales and a number of multi‐scale fractal dimensions and energy parameters were adopted as feature parameters.…”