Investigation of Thermographic Images of Photovoltaic Modules using Deep Learning Models

Abstract

Recently Solar PV panel has important role in the power generation based on renewable energy. In this paper, presents a challenge faced by faults diagnosis using thermal image analysis of Photovoltaic (PV) Module is studied. In ancient time, compared to the modern technique of PV examination using thermal imaging, the manual PV inspection approach is frequently slower, riskier, and less accurate. The use of thermal photos has the advantage of being able to immediately identify the anomaly in PV array as well as offer other measurement parameters. This research on thermal image analysis will aid in the inspection of PV modules by offering a more accurate and cost-effective identification of PV defects. According to this study, deep learning approaches are currently being considered as a feasible classifier for image processing and computer vision. Various studies, on the other hand, employed the notion of deep learning to classify and detect thermographic images used to detect flaws in PV modules. The analysis of anomaly classification and parameter evaluation were presented and explored in this method. This study looks at how well Deep Neural Networks (DNNs) models perform when it comes to classifying abnormalities in images. DNNs models have high accuracy for implementing classification of anomalies.