Study of Fault Detection on a 230kV Transmission Line Using Artificial Neural Network (ANN)

Z. I. Sobhuza, B. A. Thango

Research output: Contribution to conferencePaperpeer-review

2 Citations (Scopus)

Abstract

In the Southern African Development Community (SADC), transmission lines facilitate power utilities in distributing a considerable amount of electricity from generation stations to end users. Long conductors used in transmission lines can cover great distances. However, they are susceptible to various environmental and weather conditions which can have adverse effects such as power outages, and equipment malfunctions, posing a serious threat to the reliability of the system. Hence constant monitoring of the performance characteristics of this system is crucial. The main objective of this work is to propose an Artificial Intelligent (AI) based technique, which is the Artificial Neural Network (ANN) to detect and classify faults on a transmission line. In contrast to the most common approach viz. the protective relay system, a five-layer feed-forward-back propagation neural network architecture is proposed in this work to detect and classify faults on a 230kV transmission line system. A set of 12 fault conditions have been predefined viz. No fault, AB, AC, BC, ABC, AG, BG, CG, ABG, ACG, BCG and ABCG conditions. The results indicate that the proposed ANN approach with Levenberg-Marquardt (LM) algorithm, 5-Layers and TANSIG transfer function yield an output of 0.8927, 0.8882, 0.905 and 0.8938 in the training, validation, testing and overall accuracy respectively. To corroborate these results, a comparative study of the proposed network and other neural networks was also carried out.

Original languageEnglish
DOIs
Publication statusPublished - 2023
Event31st Southern African Universities Power Engineering Conference, SAUPEC 2023 - Johannesburg, South Africa
Duration: 24 Jan 202326 Jan 2023

Conference

Conference31st Southern African Universities Power Engineering Conference, SAUPEC 2023
Country/TerritorySouth Africa
CityJohannesburg
Period24/01/2326/01/23

Keywords

  • artificial neural network (ANN)
  • faults
  • Levenberg-Marquardt (LM) algorithm
  • Southern African Development Community (SADC)
  • transmission line

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Energy Engineering and Power Technology
  • Renewable Energy, Sustainability and the Environment
  • Electrical and Electronic Engineering
  • Safety, Risk, Reliability and Quality

Fingerprint

Dive into the research topics of 'Study of Fault Detection on a 230kV Transmission Line Using Artificial Neural Network (ANN)'. Together they form a unique fingerprint.

Cite this