Performance analysis of machine learning algorithms for hybrid power generation prediction

This paper aims to analyze the energy potentials of Çanakkale and Balıkesir provinces and evaluate the performance of renewable energy forecasting models. The first section presents a comprehensive analysis of wind and solar energy potentials. ANOVA analysis is used to evaluate the effects of meteorological variables on energy production. In the wind power analysis, air density and wind speed were found to be critical, emphasizing the importance of these factors in determining wind energy potential. In the solar panel power analysis, temperature, sunshine duration, solar radiation, and ambient temperature were critical. Regression analysis showed that most of the variability in turbine power is caused by wind speed. In the second section, based on the results of the previous analysis, an evaluation of the performance of the machine learning models is presented. Comparisons between the actual data and the predicted values are made using metrics such as coefficient of determination, mean absolute error, and root mean squared error. In addition, a custom prediction model is used to identify the features that affect the model's outputs. The third section evaluates the regional impacts. Using six different forecasting models, a regional performance analysis is performed for wind turbine power and solar panel power in Çanakkale and Balıkesir provinces. The analysis shows that high coefficients of determination are obtained for both provinces. As a result, a comparison of model performances is made in the fourth section. The XGBoost algorithm showed high accuracy on the training set and performed equally well with the other algorithms on the test set. The Random Forest model performed the best in terms of MAE. This inferential analysis provides important information for understanding the energy potential of Çanakkale and Balıkesir provinces and evaluating the performance of machine learning models. The paper contributes to regional energy planning and sheds light on practical applications for more efficient use of renewable energy resources.