Application of acoustic emission techniques and artificial neural networks to partial discharge classification

Abstract

Partial discharge (PD) detection, signal analysis and pattern identification, using acoustic emission measurements and the back-propagation (BP) artificial neural network (ANN) have been investigated. The measured signals were processed using three-dimensional patterns and short duration Fourier transforms (SDFT). Investigation indicates that using BP ANN with the SDFT components for classifying different PD patterns provides very good overall results.