Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach

Convolutional Neural Network (CNN) is a Deep Learning algorithm that can take in an input signal, assign importance (learnable weights and biases) to various aspects of the signals and distinguish between them. The CNN algorithm trains a sample and obtains a CNN model capable of identifying differen...

Full description

Bibliographic Details
Published in:Journal of Advanced Research in Applied Sciences and Engineering Technology
Main Author: Omar A.M.S.; Samat A.A.A.; Faisal F.; Osman M.K.; Ibrahim M.N.; Hussain Z.
Format: Article
Language:English
Published: Penerbit Akademia Baru 2023
Online Access:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85174713230&doi=10.37934%2faraset.32.3.199216&partnerID=40&md5=8b103ceff2e1f68dc6fe5ff3209d2077
id 2-s2.0-85174713230
spelling 2-s2.0-85174713230
Omar A.M.S.; Samat A.A.A.; Faisal F.; Osman M.K.; Ibrahim M.N.; Hussain Z.
Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
2023
Journal of Advanced Research in Applied Sciences and Engineering Technology
32
3
10.37934/araset.32.3.199216
https://www.scopus.com/inward/record.uri?eid=2-s2.0-85174713230&doi=10.37934%2faraset.32.3.199216&partnerID=40&md5=8b103ceff2e1f68dc6fe5ff3209d2077
Convolutional Neural Network (CNN) is a Deep Learning algorithm that can take in an input signal, assign importance (learnable weights and biases) to various aspects of the signals and distinguish between them. The CNN algorithm trains a sample and obtains a CNN model capable of identifying different fault types to analyse transmission line faults. Fault analysis methods typically require feature extraction to synthesise the relevant and non-redundant information from the raw signals. However, these traditional methods are time-consuming and inconsistent, and they can produce biased results due to the reliance on human expertise and experience. Hence, this research focused on developing an intelligent system for fault detection in a three-phase transmission line using CNN. This research aims to develop a CNN model for automatic fault detection in a three-phase transmission line and evaluate the performance of the CNN model for analysing transmission line faults. The three-phase transmission line model was developed using MATLAB-Simulink. CNN was implemented to detect transmission line faults. The performance of CNN based on the signal segmentation approach was evaluated through three different types of data: ideal and noise-added signal data. The simulation result shows good performance accuracy of 99.11% for the ideal case, 99.36%, and 99.39% for 20 dB and 30 dB noise-added cases, respectively. The result shows that a higher noise value in transmission line fault current could increase the performance of CNN. In conclusion, the utilisation of CNN based on a signal segmentation approach for transmission line fault analysis has showed promising performance. © 2023, Penerbit Akademia Baru. All rights reserved.
Penerbit Akademia Baru
24621943
English
Article
All Open Access; Hybrid Gold Open Access
author Omar A.M.S.; Samat A.A.A.; Faisal F.; Osman M.K.; Ibrahim M.N.; Hussain Z.
spellingShingle Omar A.M.S.; Samat A.A.A.; Faisal F.; Osman M.K.; Ibrahim M.N.; Hussain Z.
Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
author_facet Omar A.M.S.; Samat A.A.A.; Faisal F.; Osman M.K.; Ibrahim M.N.; Hussain Z.
author_sort Omar A.M.S.; Samat A.A.A.; Faisal F.; Osman M.K.; Ibrahim M.N.; Hussain Z.
title Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
title_short Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
title_full Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
title_fullStr Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
title_full_unstemmed Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
title_sort Convolutional Neural Network for Transmission Line Fault Diagnosis Based on Signal Segmentation Approach
publishDate 2023
container_title Journal of Advanced Research in Applied Sciences and Engineering Technology
container_volume 32
container_issue 3
doi_str_mv 10.37934/araset.32.3.199216
url https://www.scopus.com/inward/record.uri?eid=2-s2.0-85174713230&doi=10.37934%2faraset.32.3.199216&partnerID=40&md5=8b103ceff2e1f68dc6fe5ff3209d2077
description Convolutional Neural Network (CNN) is a Deep Learning algorithm that can take in an input signal, assign importance (learnable weights and biases) to various aspects of the signals and distinguish between them. The CNN algorithm trains a sample and obtains a CNN model capable of identifying different fault types to analyse transmission line faults. Fault analysis methods typically require feature extraction to synthesise the relevant and non-redundant information from the raw signals. However, these traditional methods are time-consuming and inconsistent, and they can produce biased results due to the reliance on human expertise and experience. Hence, this research focused on developing an intelligent system for fault detection in a three-phase transmission line using CNN. This research aims to develop a CNN model for automatic fault detection in a three-phase transmission line and evaluate the performance of the CNN model for analysing transmission line faults. The three-phase transmission line model was developed using MATLAB-Simulink. CNN was implemented to detect transmission line faults. The performance of CNN based on the signal segmentation approach was evaluated through three different types of data: ideal and noise-added signal data. The simulation result shows good performance accuracy of 99.11% for the ideal case, 99.36%, and 99.39% for 20 dB and 30 dB noise-added cases, respectively. The result shows that a higher noise value in transmission line fault current could increase the performance of CNN. In conclusion, the utilisation of CNN based on a signal segmentation approach for transmission line fault analysis has showed promising performance. © 2023, Penerbit Akademia Baru. All rights reserved.
publisher Penerbit Akademia Baru
issn 24621943
language English
format Article
accesstype All Open Access; Hybrid Gold Open Access
record_format scopus
collection Scopus
_version_ 1809677579756503040