Regression analysis for performance prediction according to climate data in solar power plants

In order to increase the efficiency of solar power plants, many studies are carried out to develop panel technology, determine suitable locations and analyze climate conditions. In this study, the output power of the 200kW solar power plant installed in Amasya University İpekköy Campus was estimated. Different methods were used to estimate the power produced depending on the climate data, and the performances of the methods were compared. Linear Regression, Decision Tree Regression and Gaussian Process Regression were used as estimation methods in the thesis. Results of analyzes performed using 5 and 10 fold cross validation were compared. The highest performance was obtained with Gaussian Process Regression.