Estimation of Amasya University solar power plant production data with artificial intelligence methods based on meteorological data

In recent years, while the demand for energy is increasing day by day, the need for renewable energy sources is increasing day by day with the decrease in fossil fuels and the onset of sensitivity to environmental pollution. Especially the use of solar energy has shown a rapid acceleration in recent times. In this study, solar energy has been investigated in general, and in practice, solar power plant production data has been estimated with artificial intelligence methods, multi-layer artificial neural networks (MLPNN) and adaptive fuzzy artificial neural network inference system (ANFIS), depending on meteorological data. It was determined that the best result was obtained with the Levenberg-Marquardt (trainlm) algorithm in the estimation made using the MLPNN method. In the estimation made using the ANFIS method, it was determined that the most successful result was produced by the hybrid algorithm and the gbellmf membership function type and the model in which the 4-4-4 membership function was used together for the input values. It has been seen from the scatter plots that the predicted values made with the created MLPNN and ANFIS models converge reasonably with the measured values. It was understood that the estimations were reliable and accurate in the regression curves made after the test results. The production data of the power plants in the energy sector can be calculated with the same methods.