A comprehensive analysis of an artificial neural-network-based fault direction discriminator

Author

Sidhu, T.S. ; Mital, L. ; Sachdev, Mohindar S.

Author_Institution

Dept. of Electr. & Comput. Eng., Univ. of Western Ontario, London, Ont., Canada

Volume

19

Issue

3

fYear

2004

fDate

7/1/2004 12:00:00 AM

Firstpage

1042

Lastpage

1048

Abstract

This paper presents a comprehensive analysis of an artificial neural network based fault direction discriminator. Individual outputs of neurons are investigated. A new technique based on backtracking search procedure is developed for analyzing the multilayer feedforward neural networks. The technique is applied to the fault direction discriminator´s individual neurons. Results and conclusions of this application are presented.

Keywords

feedforward neural nets; multilayer perceptrons; power engineering computing; power system faults; power system protection; power system relaying; relay protection; ANN-based relays; artificial neural-network; backtracking search procedure; fault direction discriminator; multilayer feedforward neural networks; neurons; power system protection; power system relaying; Artificial neural networks; Feedforward neural networks; Multi-layer neural network; Neural networks; Neurons; Power engineering and energy; Power system protection; Power system relaying; Protective relaying; Systems engineering and theory; ANN-based relays; artificial neural networks; fault direction discriminator; power system protection; power system relaying;

fLanguage

English

Journal_Title

Power Delivery, IEEE Transactions on

Publisher

ieee

ISSN

0885-8977

Type

jour

DOI

10.1109/TPWRD.2004.829118

Filename

1308325