Efficient combination of deep learning models for solar panel damage and soiling detection

Abstract

Solar panels are essential for converting sunlight into electricity. Still, environmental factors can significantly compromise their efficiency and performance, particularly the accumulation of soiling on their surfaces or damage. This study proposes a hybrid model comprising an ensemble of deep-learning models to distinguish between soiled and damaged solar panels and their corresponding conditions. Our approach utilizes pre-trained deep learning models, fine-tuned for detecting soiling or damage on photovoltaic (PV) panels, to extract relevant features and build efficient classifiers. Introducing a post-processing ensemble model improves classification metrics compared to a single deep-learning model. Combining Convolutional Neural Networks and Vision Transformers in an ensemble model achieves the highest accuracy, with 96.3% for damage and soiling detection and 91.8% for damage and soiling type identification. These results significantly outperform one-tier deep learning models, which attain an accuracy of 87.7% when classifying all possible damage and soiling categories.