YOLOv8-based Photovoltaic Panel Cleaning Device

Abstract

In order to improve the efficiency of solar panels and solve the problems of high labour intensity, low efficiency and high safety risk in the cleaning process of photovoltaic panels, this paper designs a photovoltaic panel cleaning device based on YOLOv8 algorithm. Firstly, the YOLOv8 algorithm is used to collect image data of a large number of PV panel surface stains, identify and locate the stained areas in the image for data preprocessing. Then, the microcontroller with the main control chip model STM32F407IGH6 is selected to control the motion of the cleaning device so that it can complete the complete automatic cleaning function. Finally, the complete cleaning process of the photovoltaic panels is achieved by means of a roller brush on the cleaning device. The experimental results show that based on the YOLOv8 data processing results and PID control algorithm, the cleaning strategy can be dynamically adjusted according to the size and location of the target area, and the photovoltaic cleaning device can be accurately controlled to accurately and efficiently clean the dust and stains on the surface of the solar panels, which greatly improves the efficiency of solar energy cleaning.