Identification of Partial Discharge Locations in Transformer Winding Using PSD Estimation

Abstract

This letter presents a new technique to identify partial discharge locations in transformer winding using electrical measurements. Useful features, which pinpoint the locations of partial discharges, are extracted from the periodogram of the secondary voltage of the transformer. These features are manipulated by a distance classifier to determine the locations of the partial discharges. Simulation results proof that the proposed method is efficient and can handle effectively transformer parameters deviations due to aging.Index Terms-Partial discharge, periodogram, power spectral density, power transformer.

Optimum acoustic sensor placement for partial discharge allocation in transformers

“…Each of these energy by-products, when measured, has advantages and disadvantages for identifying PD [6]. Based on these features, several works have been performed on the detection and diagnosis of the PD signals [7][8][9][10][11], and their localisation [12][13][14][15][16][17][18].…”

Identification of Partial Discharge Location in a Power Cable Using Fuzzy Inference System and Probabilistic Neural Networks

“…The bandwidth of capacitive coupler sensors are 1-20 MHz, and the sampling rate of the oscilloscope is 1 GSa/s. The periodogram method is employed to estimate the PSD for the PD signals [12], and the average power of the PSD can be calculated by the rectangle approximation integral method [27]. …”

“…Methods for identification of PD include the ( -q-n) patterns method [1][2][3], the neural network approach, the fuzzy classification method, neuron-fuzzy networks, support vector machines (SVMs), and the rise time of PD signals [4][5][6][7][8][9][10]. Tracking of source location of PD signals involves feature extraction using the arrival time method and estimation of the power spectral density (PSD) [11,12]. The methods mentioned above are used to track the source location of PD.…”

Phase determination of partial discharge source in a three-phase transmission line based on high-frequency equivalent circuit model and signal detection using HFCT

“…In accordance with the concept of power delivery, the average power of PD signals are also applied to located PD source [6]. In order to verify the proposed method, the signal can be expressed in an exponential Fourier series, as follows …”

“…The latter employs the acoustic emission technique, dissolved gas analysis and infrared imaging method [1,2]. The source location of PD signals can be tracked by feature extraction methods which include the conventional amplitude of PD signals [3], arrival time method [4,5], estimation of power spectral density (PSD) [6] and particle swarm optimization [7]. To identify phase determination of PD source, the methods have simultaneous detection in the three-phase power cable [8], and polarity and magnitude of PD are measured for a defect attached on different positions [9].…”