Fault Detection and Fault Classification in Transmission Lines using Machine Learning

This paper focuses on the vital task of utilizing Machine Learning techniques for fault detection and fault classification in electrical power transmission lines. It employs two distinct datasets, one for fault detection and the other for fault classification. The study compares the effectiveness of the Sequential Neural Network (SNN) and K-Nearest Neighbors (KNN) for fault detection, and assesses ensemble classifier, specifically the Random Forest, Decision Tree and Extra Tree Classifier for fault classification. An additional Ensemble Classifier Algorithm that combines Random Forest, Decision Tree and Adaboost Classifier is introduced for fault classification, and performance of each Machine Learning model is evaluated. Simulated data is used for rigorous testing, showcasing the efficiency of this approach. The implementation of SNN and KNN for fault detection and the comparison of ensemble classifiers for fault classification, result in building a system with high accuracy and reliability in identifying and categorizing transmission line faults. Notably, SNN outperforms in fault detection, while the Ensemble Classifier with Random Forest, Decision Tree, and Extra Tree Classifier excels in fault classification. This research significantly contributes to enhancing the reliability of power transmission systems, leading to quicker responses and reduced downtime.