Desenvolvimento de um método para monitorar a temperatura de módulos fotovoltaicos em uma usina solar por meio de imagens térmicas obtidas por drone

Abstract

In this dissertation, a method is presented that uses thermal images obtained th- rough the use of a drone, to monitor the temperature of photovoltaic modules in a solar plant, in order to make their inspection more agile, assist in the identification of defective modules and improve the performance of the photovoltaic system. Image processing and computer vision techniques were used to identify modules in thermal images of photo-voltaic panels, such as the Canny edge detector, the Hough transform and the projective transform. A strategy for fitting a curve to the temperature histogram, using the least squares method, was used to identify areas of excessive heat in the modules. These and other techniques were synthesized in the GNU Octave software environment, where a graphical interface was developed to use them. The proposed method was applied to a case study, in a 1 megawatt solar plant located in the east of the State of Sergipe. Some areas of the plant’s modules were shaded to verify the method’s ability to identify areas of excessive heat. The results prove the effectiveness of the method in identifying all of them.