A Convolutional Neural Network-Based Approach for High-Accuracy Fault Diagnosis in Photovoltaic Systems

Abstract

The increasing demand for renewable energy has led to the widespread adoption of photovoltaic (PV) systems. However, the efficiency and reliability of these systems are often compromised due to various faults, which can significantly impact their performance. This paper presents a machine learning-based fault diagnosis system for solar panels, focusing on the detection and classification of faults in PV systems. We evaluate several machines learning models, including Random Forest, Decision Tree, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Logistic Regression, and Naive Bayes, using a dataset of grid-connected PV system faults. Our results demonstrate that the Convolutional Neural Network (CNN) model outperforms other models, achieving a validation accuracy of 99.8%. The proposed system offers a robust solution for maintaining the efficiency and reliability of PV systems, contributing to the broader goal of sustainable energy development.