Close-range Hot Spot Detection based on K-neighbor Algorithm and Experiment on Hot Spot Effect

Abstract Abnormalities caused by the hot spot effect are very common in the application of photovoltaic modules, which seriously affect the life and power generation efficiency of photovoltaic modules. Therefore, the study of hot spot effect is of great significance to the development of photovoltaic industry in the future. By analyzing the hot spot effect and its causes, this paper lists possible factors that may affect the hot spot effect, uses the K-proximity algorithm for hot spot detection to classify the near-field infrared images obtained in photovoltaic power plants, and designs experiments to explore the influence of several factors on the hot spot effect, the influence characteristics of the shading area and the types of shading objects are deeply explored, and finally a certain summary is made.