Photovoltaic module dataset for automated fault detection and analysis in large photovoltaic systems using photovoltaic module fault detection
Description
The PVMD dataset has 3-category of 1000 images, which includes both permanent and temporal anomalies in solar cells of PV module such as hotspots, cracks, and shadings. The dataset was collected on September 5, 2024 at the Soshanguve South Campus, Tshwane University of Technology, South Africa using DJI Mavic 3 Thermal’s high-resolution thermal and visual imaging capabilities. The PVMD dataset comprises three folders, namely Hotspots, Cracks, and Shadings. The folder name is the name of the particular anomaly in it. The Hotspots folder comprises 350 thermal generated images; the Cracks folder comprises 350 thermal generated images; and the Shadings folder comprises 300 thermal generated images. The dataset collection aims to provide supervised and publicly available dataset good enough to support research method in providing a comprehensive and efficient approach to monitoring and maintaining large PV systems.