MACHINE LEARNING APPROACH FOR SOILING DETECTION AND ANALYSIS ON SOLAR PANELS (Análise de modelos para detecção de sujidade e danos em placas fotovoltaicas)

Resumo

https://doi.org/10.5281/zenodo.20166284  

This work presents an innovative solution for monitoring solar panels by using images captured through surveillance cameras and drones, aiming to detect the level of dirtiness on photovoltaic modules. The proposed methodology employs machine learning techniques to analyze and categorize the degree of dirt accumulation on the panels, facilitating maintenance for system owners. By providing an automated and accurate assessment of the panels’ condition, the solution seeks to enhance the efficiency of solar energy generation, contributing to the optimal performance of photovoltaic systems and promoting energy sustainability for users.