Wavelet transform coherence for magnitude and phase spectrum prediction from high frequency transient signals: Partial discharge in transformers

• Published in: 2016 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES) pp. 1 - 6
• Main Authors: Suryavanshi, Hitesh, Velandy, Jeyabalan, Surendran, Jayalakshmi
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2016
• Subjects: Discharges (electric); Impulse testing; Insulation; Partial discharge; Partial discharges; Power transformer insulation; Signal processing analysis; spectrum analysis; Transformer; Transient signals; Wavelet transforms; Windings
• DOI: 10.1109/PEDES.2016.7914559

Identification of partial discharge (PD) is an important diagnostic tool for reliable decision of transformer during impulse test. Since, transformer is not accessible directly to investigate the pass or fail criteria of the transformer due to PD or any other major faults. Only the measured high frequency nature of current signals (winding responses) due to different impulse test sequences can help to decide the fault through any synchronization principles of whether two time domain or frequency domain responses are statistically independent or not. Extracting such actual information becomes difficult if the transformer is nonlinear and non stationary. Hence, to study the features of the measured non stationary nature of current signals due to PD, wavelet transform coherence analysis is performed. The spectrum of magnitude and phase are extracted from the measured current signals to identify the PD for showing the efficacy. To prove the feasibility of the analysis, 22 kV interleaved winding is utilized.