Key points are not available for this paper at this time.
As a clean and sustainable renewable energy, photovoltaic (PV) power generation has been widely used around the world. The power generation efficiency of solar photovoltaic panels is affected by installation and environmental factors, so PV panels require regular cleaning, inspection and maintenance. However, traditional maintenance methods require a lot of labor and time. But in recent years, with the development of UAV technology and the advancement of target detection algorithms, a more efficient solution is provided. This study combines infrared target detection technology, uses drone aerial photography to collect infrared image data of PV power plant scenes, and uses spatial domain-based data enhancement and feature difference fusion (FDF) methods to train a fault detection model. The constructed neural network model can quickly, accurately and comprehensively acquire the types and locations of PV panel faults in PV power plants, significantly reducing labor and time costs. The experimental results show that the method has excellent performance in PV power plant fault detection and can effectively improve the operation and maintenance efficiency of PV power plants.
Si et al. (Fri,) studied this question.