Current techniques for fault analysis in photovoltaic (PV) systems plants involve either electrical performance measurements or image processing, as well as line infrared thermography for visual inspection. Deep convolutional neural networks (CNNs) are machine learning algorithms that perform tasks involving images, such as image classification and object recognition. However, to train a model effectively to recognize different patterns, it is crucial to have a sufficiently balanced dataset. Unfortunately, this is not always feasible owing to the limited availability of publicly accessible datasets for PV thermographic data and the unequal distribution of different faults in real-world systems. In this study, three data augmentation techniques—geometric transformations (GTs), generative adversarial networks (GANs), and the denoising diffusion probabilistic model (DDPM)—were combined with a CNN to classify faults in PV modules through thermographic images and identify the type of fault in 11 different classes (i.e., soiling, shadowing, and diode). Through the cross-validation method, the main results found with the Wasserstein GAN (WGAN) and DDPM networks combined with the CNN for anomaly classification achieved testing accuracies of 86.98% and 89.83%, respectively. These results demonstrate the effectiveness of both networks for accurately classifying anomalies in the dataset. The results corroborate the use of the diffusion model as a PV data augmentation technique when compared with other methods such as GANs and GTs.

This paper main goal is to detect anomalies in photovoltaic panels installed in power generation systems using thermographic images. It was used image processing techniques and pattern recognition by means of artificial neural network with supervised and adaptive training. In order to implement the anomaly detection system it was selected from the tutorial "Solar Cell Development" by FLIR systems, thermographic images of photovoltaic cells with defects. The language of technical computing, MATLAB©, was used to make the preprocessing of all the collected images and it was also used for the training of an ABNET network, a PSOM network and a MLP network. Through the cross-validation method, the hit rate of the ABNET’s network and PSOM’s network was 87.5% and for the MLP the hit rate was 96.0%. The results corroborate for the use of these neural networks on a bank of images that will be created through a Research and Development Project (R&D) in progress. Este trabalho tem como objetivo principal propor um método baseado em Redes Neurais Artificiais (RNA) para detectar anomalias em painéis fotovoltaicos instalados em sistemas de geração elétrica, tendo como base imagens termográficas. Assim, foram utilizadas técnicas de processamento de imagem e reconhecimento de padrões por meio de RNA com treinamento supervisionado e adaptativo. Para a implementação do sistema de detecção de anomalias, foram selecionadas imagens do tutorial “Solar Cell Development” da FLIR systems, sendo que estas imagens termográficas apresentam pontos de aquecimento em células específicas de determinado painel fotovoltaico. A ferramenta computacional Matlab© foi utilizada para realizar o pré-processamento do banco de imagens e para o treinamento de uma rede ABNET, uma rede PSOM e uma rede MLP. Por meio de validação cruzada, a taxa de acertos para a classificação dos padrões da rede ABNET e PSOM foi de 87,5% e para a rede MLP foi de 96%. Os resultados corroboram para utilização dessas redes em um banco de imagens que será construído por meio do Projeto de Pesquisa e Desenvolvimento (P&D) em andamento.