Fault Detection in Photovoltaic System Using Machine Learning Techniques and ANN

Ensuring the reliability of a photovoltaic (PV) system depends on effective fault classification and detection. Under various environmental circumstances, traditional fault detection methods may incorrectly identify operational PV systems as defective or vice versa. This paper proposes a machine learning and artificial neural network (ANN) model for effective fault diagnosis in PV systems. The dataset is pre-processed and fed into multiple models, including SVM, KNN, Random Forest, Naive Bayes, and ANN. This study uses metrics such as accuracy, precision, recall, and F1-score as performance measures. The results show that the ANN can reliably find faults and classify them in a wide range of environmental conditions by using its power to see intricate patterns. The findings show that the proposed model effectively improves fault diagnosis in PV systems.