Enhancing solar PV reliability with hybrid local features and infrared thermography

Abstract

The widespread adoption of Solar Photovoltaic Systems (PVS) faces challenges due to operational and environmental stresses, which can lead to hotspots and potential failures. This study introduces a novel hybrid local features-based approach for monitoring PVS using infrared thermographs. This method is designed to be resistant to scaling, noise, rotation, and haze. The approach involves subdividing each thermograph into 5 × 5 pixel grids, extracting features, and clustering them using an unsupervised algorithm to reduce memory usage. A shallow classifier is then employed, achieving an impressive 98 % training accuracy and 96.8 % testing accuracy with 5-fold cross-validation. Additionally, the model demonstrates high performance across various metrics, with precision values of 92 %, 100 %, and 100 %; recall values of 100 %, 100 %, and 90 %; and F1 scores of 0.958, 1.0, and 0.947 for the faulty, healthy, and hotspot classes, respectively. A Simulink model was developed to assess the impact of environmental stresses on PVS power losses, showcasing the method's practical applicability. This approach significantly enhances the reliability and performance of PVS, aligning with global initiatives such as the United Nations’ Sustainable Development Goals, EU energy targets, the Kyoto Protocol, and the Paris Climate Accord. It ensures sustainable and reliable PVS operation over the next 25 years.