Abstract: The "hotspot" phenomenon of photovoltaic modules critically affect the photovoltaic power generation system. To mitigate high labor intensity and the low efficiency of traditional manual inspection, in this study, we propose an intelligent inspection method based on unmanned aerial vehicle (UAV) infrared vision for photovoltaic power station hotspots. This proposed method effectively improves the intelligence level of photovoltaic-power-station health status assessment. First, the state characteristics of the photovoltaic hotspots are analyzed, and a coupled relationship model of the hotspot infrared image, temperature, and severity is established. To improve the accuracy of hotspot detection, a target detection method based on hybrid intensive YOLO (MI-YOLO) is proposed, which enhances the expression ability of shallow feature information. Subsequently, an intelligent inspection scheme for the UAV is designed to obtain multidimensional information of hotspots in real-time. Finally, considering the recognition speed, recall rate, and accuracy rate as evaluation indicators, multiple sets of simulation experiments are performed to compare the influence of different activation functions, image size, image brightness, and false detection factors of "fake" targets. The experimental results show that MI-YOLO can effectively detect the location, quantity, category, and severity of photovoltaic module hotspots. Thus, it can fulfil the real-time requirements of power plant operation and maintenance, improve power generation efficiency, and reduce operation and maintenance costs.