Arvina Rizqi Nurul'aini, Mohammad Mahruf Alam, Rizky Ajie Aprilianto, Tole Sutikno, Tri Wahono, Mochammad Facta
Electric generator fault detection is crucial for ensuring the reliability and efficiency of power systems, but conventional detection methods often lack speed and accuracy. This study explores the use of machine learning algorithms such as SVM, Random Forest, KNN, Logistic Regression, and Naïve Bayes to classify generator faults based on voltage and current signal data recorded under both normal and faulty operating conditions. After data preprocessing and 5-fold GroupKFold cross-validation in training, the models were evaluated on test dataset using accuracy, precision, recall, and F1-score. The results show that Random Forest achieved the highest performance with 99.74% across all metrics, followed by SVM with 99.48% and Logistic Regression with 98.96%. KNN also demonstrated strong results with 97.65%, while Naïve Bayes lagged with 84.60%. These findings highlight the effectiveness of AI-based approaches, particularly Random Forest, in enhancing generator fault detection and supporting intelligent monitoring systems to improve reliability and reduce maintenance costs in power infrastructures. © 2025 IEEE.
Universitas Negeri Semarang, Department of Electrical Engineering, Semarang, Indonesia; Universitas Ahmad Dahlan, Faculty of Industrial Technology, Yogyakarta, Indonesia; Electronics Research Group, Embedded Systems and Power, Yogyakarta, Indonesia; Universitas Diponegoro, Department of Electrical Engineering, Semarang, Indonesia
