Comparative Analysis of Machine Learning Algorithms for Transmission Line Fault Detection

Unforeseen failures in the electrical power transmission line may arise due to various unpredictable factors. The occurrence of power failures on transmission lines has the potential to cause significant damage to the existing power grid unless prompt detection and correction of faults are carried out. The primary objective of this paper is to address the detection and classification of faults occurring on electrical power transmission lines. Specifically, the study focuses on a dataset consisting of 12,000 data points, each characterized by six distinct features. To achieve this, artificial neural networks are employed as the chosen methodology. The classification of transmission line faults was performed using various machine learning algorithms, including Linear Regression, Support Vector Machine (SVM), Decision Tree, Random Forest, AdaBoost, Bagging Classifier, and XGBoost. The algorithms were evaluated based on parameters including accuracy, error rate, prediction speed, and training time. The findings suggest that tree-based classifiers have exhibited superior performance, achieving 100% accuracy in the classification of transmission line faults. On the other hand, Logistic Regression and Support Vector Machines (SVM) have yielded average results in this regard.