Investigation of the application of an automated monitoring system for detecting transmission cable deterioration in Nigeria: A case study of transmission cable lines between Offa and Oshogbo
This study investigates the application of an automated monitoring system for detecting the deterioration of transmission cable lines between Offa and Oshogbo in the South-West of Nigeria. By developing and implementing this advanced monitoring tool to detect transmission cable deteriorations, parti...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Results in Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2590123025002531 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study investigates the application of an automated monitoring system for detecting the deterioration of transmission cable lines between Offa and Oshogbo in the South-West of Nigeria. By developing and implementing this advanced monitoring tool to detect transmission cable deteriorations, particularly those caused by electrical overloading, a common issue in Nigeria and other developing nations, the problem of cable failure resulting from overloading can be minimized. This, in turn, will lead to an improved power supply and enhanced power system reliability in Nigeria. Neural networks were employed using MATLAB R2021a and Python, implemented in Visual Studio Code. Data from 132 kV overhead line (OHL) transmission cables, provided by the Transmission Company of Nigeria (TCN), were utilized in the study. The transmission cable data covered the connection between Offa and Oshogbo in the South-West region of Nigeria. Both forward propagation and backpropagation techniques were adopted for training Artificial Neural Networks (ANNs), and the gradient descent with momentum algorithm was employed for optimization. The best classification performance achieved a minimum mean squared error (MSE) of 0.074346 after 24 iterations (epochs) of training, validation, and testing. A generalization classification accuracy of 72.9% (accurately classified patterns) and a misclassification rate of 27.1% were recorded. The study successfully demonstrated the prediction of overload conditions using neural networks, with minimal errors illustrated through confusion matrices and performance plots. |
|---|---|
| ISSN: | 2590-1230 |