Analysis and applicability of Mersin region wind speed data with artificial neural networks

Abstract

In this study, wind speed data were analyzed in order to provide energy to the heating and electrical systems of a house with renewable energy sources in Mersin-Mut region. Three-year wind speed data is taken from the Turkey General Directorate of Meteorology in the region. Annual estimation was made with artificial neural networks using 28-day wind speed data for the studied area. Some of the wind data were used for training of the neural network, and some were used for testing. In the artificial neural network model, the most successful model was obtained by changing the number of neurons in the hidden layer. In the analysis made using eight neurons in the hidden layer, the lowest MAE and RMSE error values were calculated. While the number of neurons was eight, MAE and RMSE values were obtained as 0.4056 and 0.5403, respectively. In addition, analysis of wind data with WAsP software has been carried out for this region. Thus, the average instantaneous wind speed was determined according to the analysis studies.