Image processing based fault classification in power systems with classical and intelligent techniques

This paper is devoted to develop interest of power system engineers in learning basic concepts of image processing and consequently using deep networks to solve problems of complex power system networks. To this end, we study fault classification in a power system through automation of equal area (E...

Full description

Saved in:
Bibliographic Details
Published in:Journal of intelligent & fuzzy systems 2022, Vol.43 (2), p.1921-1932
Main Authors: Sabih, Muhammad, Umer, Muhammad, Farooq, Umar, Gu, Jason, Balas, Marius M., Asad, Muhammad Usman, Qureshi, Khurram Karim, Khan, Irfan A., Abbas, Ghulam
Format: Article
Language:English
Subjects:
Algorithms
Classification
Double circuit
Image classification
Image processing
Multilayer perceptrons
Neural networks
Rotors
Short circuits
Transmission lines
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper is devoted to develop interest of power system engineers in learning basic concepts of image processing and consequently using deep networks to solve problems of complex power system networks. To this end, we study fault classification in a power system through automation of equal area (EAC) criterion. By considering EAC graphs as images and using classical image processing techniques, we successfully distinguish between different transient conditions including sudden change of input power as well as short circuit at the sending end and middle points of a single and double circuit transmission lines. In addition to classification, some parameters are also determined from EAC images such as initial rotor angle, clearing angle, and maximum rotor angle. Further, the use of deep networks is introduced to perform the same task of fault classification and a comparison is drawn with multilayer perceptron neural networks. Developed algorithms are tested in MATLAB as well as Pytorch environments.
ISSN:1064-1246
1875-8967
DOI:10.3233/JIFS-219293