Detecção de anomalias em módulos fotovoltaicos utilizando redes neurais convolucionais

Abstract

The solar photovoltaic energy sector plays a crucial role in advancing the global sustainable energy matrix, with predictions to reach an installed capacity of 2840 GW by 2030. However, the inherent challenges in the operation and maintenance of large photovoltaic plants demand increasingly proactive and efficient solutions to maximize energy efficiency and ensure the economic viability of these facilities. In this context, the application of computer vision techniques, such as Convolutional Neural Network (CNN) models for detecting anomalies in photovoltaic modules using infrared (IR) images, emerges as a promising solution to the challenges faced by the sector. The main objective of this work is to propose the application of a CNN model for the automatic detection of anomalies in solar photovoltaic plants, through the training and validation of four distinct experiments using a specific dataset for this purpose. By focusing on the use of artificial neural networks, this study aims not only to improve maintenance and operation practices in photovoltaic plants but also to contribute to the advancement of technical knowledge in the field of electrical engineering, paving the way for future practical and real-world applications of this technology.