Wavelet transform and artificial intelligence based condition monitoring for GIS

T Lin, R K Aggarwal, C H Kim

Research output: Contribution to conferencePaper

7 Citations (Scopus)

Abstract

A condition monitoring (CM) system, which integrates wavelet analysis and artificial intelligent techniques to analyze the partial discharges (PD) patterns and classify the defective equipments in gas insulated switchgear (GIS), is presented here. The multi-resolution signal decomposition (MSD) attribute of the wavelet transform is employed to de-noise the PD signals measured on site, due to its spectrum decomposition capability. The performance of several well tried wavelets are investigated in the case of denoising, and this demonstrates that the combined use of the Biorthogonal and Daubechies type of wavelets achieves good de-noising results compared to other wavelet families. Different feature extraction methods are applied to the de-noised PD signals, to form pattern vectors which are further used as the inputs to a neural network based classifier, so as to identify the PD patterns and defective equipments in GIS. Several types of neural networks, both supervised and unsupervised (trained), are then evaluated with regard to their suitability as classifier according to training time and classification accuracy. Finally, the performance of the proposed CM system is ascertained by using a set of PD signals emanating from defects in a circuit breaker (CB), disconnect switch (DS), bus bar (BS) and insulation spacer (SP) in practical GISs in Korean 154 KV high voltage transmission networks.
Original languageEnglish
Pages191-195 Vol.1
Publication statusPublished - 2003
EventTransmission and Distribution Conference and Exposition, 2003 IEEE PES -
Duration: 1 Jan 2003 → …

Conference

ConferenceTransmission and Distribution Conference and Exposition, 2003 IEEE PES
Period1/01/03 → …

Fingerprint

Electric switchgear
Partial discharges
Condition monitoring
Wavelet transforms
Artificial intelligence
Gases
Classifiers
Neural networks
Decomposition
Wavelet analysis
Electric power transmission networks
Electric circuit breakers
Geographic information systems
Insulation
Feature extraction
Switches
Defects
Electric potential

Keywords

  • learning (artificial intelligence)
  • wavelet transforms
  • partial discharges patterns
  • feature extraction
  • neural network
  • power engineering computing
  • gas insulated switchgear
  • 154 kV
  • multiresolution signal decomposition
  • circuit breaker
  • disconnect switch
  • partial discharges
  • feature extraction methods
  • artificial intelligence
  • high voltage transmission networks
  • wavelet transform
  • condition monitoring
  • condition monitoring system
  • neural nets
  • bus bar
  • signal denoising
  • insulation spacer

Cite this

Lin, T., Aggarwal, R. K., & Kim, C. H. (2003). Wavelet transform and artificial intelligence based condition monitoring for GIS. 191-195 Vol.1. Paper presented at Transmission and Distribution Conference and Exposition, 2003 IEEE PES, .

Wavelet transform and artificial intelligence based condition monitoring for GIS. / Lin, T; Aggarwal, R K; Kim, C H.

2003. 191-195 Vol.1 Paper presented at Transmission and Distribution Conference and Exposition, 2003 IEEE PES, .

Research output: Contribution to conferencePaper

Lin, T, Aggarwal, RK & Kim, CH 2003, 'Wavelet transform and artificial intelligence based condition monitoring for GIS' Paper presented at Transmission and Distribution Conference and Exposition, 2003 IEEE PES, 1/01/03, pp. 191-195 Vol.1.
Lin T, Aggarwal RK, Kim CH. Wavelet transform and artificial intelligence based condition monitoring for GIS. 2003. Paper presented at Transmission and Distribution Conference and Exposition, 2003 IEEE PES, .
Lin, T ; Aggarwal, R K ; Kim, C H. / Wavelet transform and artificial intelligence based condition monitoring for GIS. Paper presented at Transmission and Distribution Conference and Exposition, 2003 IEEE PES, .
@conference{347d69f2284f4c71af915eb66afcecc9,
title = "Wavelet transform and artificial intelligence based condition monitoring for GIS",
abstract = "A condition monitoring (CM) system, which integrates wavelet analysis and artificial intelligent techniques to analyze the partial discharges (PD) patterns and classify the defective equipments in gas insulated switchgear (GIS), is presented here. The multi-resolution signal decomposition (MSD) attribute of the wavelet transform is employed to de-noise the PD signals measured on site, due to its spectrum decomposition capability. The performance of several well tried wavelets are investigated in the case of denoising, and this demonstrates that the combined use of the Biorthogonal and Daubechies type of wavelets achieves good de-noising results compared to other wavelet families. Different feature extraction methods are applied to the de-noised PD signals, to form pattern vectors which are further used as the inputs to a neural network based classifier, so as to identify the PD patterns and defective equipments in GIS. Several types of neural networks, both supervised and unsupervised (trained), are then evaluated with regard to their suitability as classifier according to training time and classification accuracy. Finally, the performance of the proposed CM system is ascertained by using a set of PD signals emanating from defects in a circuit breaker (CB), disconnect switch (DS), bus bar (BS) and insulation spacer (SP) in practical GISs in Korean 154 KV high voltage transmission networks.",
keywords = "learning (artificial intelligence), wavelet transforms, partial discharges patterns, feature extraction, neural network, power engineering computing, gas insulated switchgear, 154 kV, multiresolution signal decomposition, circuit breaker, disconnect switch, partial discharges, feature extraction methods, artificial intelligence, high voltage transmission networks, wavelet transform, condition monitoring, condition monitoring system, neural nets, bus bar, signal denoising, insulation spacer",
author = "T Lin and Aggarwal, {R K} and Kim, {C H}",
year = "2003",
language = "English",
pages = "191--195 Vol.1",
note = "Transmission and Distribution Conference and Exposition, 2003 IEEE PES ; Conference date: 01-01-2003",

}

TY - CONF

T1 - Wavelet transform and artificial intelligence based condition monitoring for GIS

AU - Lin, T

AU - Aggarwal, R K

AU - Kim, C H

PY - 2003

Y1 - 2003

N2 - A condition monitoring (CM) system, which integrates wavelet analysis and artificial intelligent techniques to analyze the partial discharges (PD) patterns and classify the defective equipments in gas insulated switchgear (GIS), is presented here. The multi-resolution signal decomposition (MSD) attribute of the wavelet transform is employed to de-noise the PD signals measured on site, due to its spectrum decomposition capability. The performance of several well tried wavelets are investigated in the case of denoising, and this demonstrates that the combined use of the Biorthogonal and Daubechies type of wavelets achieves good de-noising results compared to other wavelet families. Different feature extraction methods are applied to the de-noised PD signals, to form pattern vectors which are further used as the inputs to a neural network based classifier, so as to identify the PD patterns and defective equipments in GIS. Several types of neural networks, both supervised and unsupervised (trained), are then evaluated with regard to their suitability as classifier according to training time and classification accuracy. Finally, the performance of the proposed CM system is ascertained by using a set of PD signals emanating from defects in a circuit breaker (CB), disconnect switch (DS), bus bar (BS) and insulation spacer (SP) in practical GISs in Korean 154 KV high voltage transmission networks.

AB - A condition monitoring (CM) system, which integrates wavelet analysis and artificial intelligent techniques to analyze the partial discharges (PD) patterns and classify the defective equipments in gas insulated switchgear (GIS), is presented here. The multi-resolution signal decomposition (MSD) attribute of the wavelet transform is employed to de-noise the PD signals measured on site, due to its spectrum decomposition capability. The performance of several well tried wavelets are investigated in the case of denoising, and this demonstrates that the combined use of the Biorthogonal and Daubechies type of wavelets achieves good de-noising results compared to other wavelet families. Different feature extraction methods are applied to the de-noised PD signals, to form pattern vectors which are further used as the inputs to a neural network based classifier, so as to identify the PD patterns and defective equipments in GIS. Several types of neural networks, both supervised and unsupervised (trained), are then evaluated with regard to their suitability as classifier according to training time and classification accuracy. Finally, the performance of the proposed CM system is ascertained by using a set of PD signals emanating from defects in a circuit breaker (CB), disconnect switch (DS), bus bar (BS) and insulation spacer (SP) in practical GISs in Korean 154 KV high voltage transmission networks.

KW - learning (artificial intelligence)

KW - wavelet transforms

KW - partial discharges patterns

KW - feature extraction

KW - neural network

KW - power engineering computing

KW - gas insulated switchgear

KW - 154 kV

KW - multiresolution signal decomposition

KW - circuit breaker

KW - disconnect switch

KW - partial discharges

KW - feature extraction methods

KW - artificial intelligence

KW - high voltage transmission networks

KW - wavelet transform

KW - condition monitoring

KW - condition monitoring system

KW - neural nets

KW - bus bar

KW - signal denoising

KW - insulation spacer

M3 - Paper

SP - 191-195 Vol.1

ER -