Title :
Comparison of feature extraction methods in partial discharge waveform recognition
Author :
Zheng, Z. ; Tan, K.
Author_Institution :
Tsinghua Univ., Beijing, China
Abstract :
Automated recognition of various types of partial discharge pulses based on the pulse waveform was investigated through application of an artificial neural network. Various feature extraction methods were applied, and the recognition efficiency was determined. The results indicate that the method based on the physical characteristics of the partial discharge, which employs expert prior knowledge, is most effective and computationally least intensive
Keywords :
feature extraction; neural nets; partial discharges; waveform analysis; artificial neural network; expert prior knowledge; feature extraction methods; partial discharge waveform recognition; physical characteristics; pulse waveform; recognition efficiency; Artificial neural networks; Feature extraction; Intelligent networks; Neurons; Oil insulation; Partial discharges; Pattern recognition; Pulse measurements; Stators; Testing;
Conference_Titel :
Electrical Insulation and Dielectric Phenomena, 2001 Annual Report. Conference on
Conference_Location :
Kitchener, Ont.
Print_ISBN :
0-7803-7053-8
DOI :
10.1109/CEIDP.2001.963547
