Analysis of Prediction Models for Wind Power Density, Case Study: Ercan Area, Northern Cyprus
This work focuses on the application of Multilayer Perceptron Neural Network (MLPNN), Radial Basis Function Neural Network (RPFNN) and Auto Regressive Integrated Moving Average (ARIMA) as predictive tools for the production of wind power density (WPD). The air temperature (AT), dew point (DP), atmospheric humidity (AH), pressure (P) and wind speed (WS) were used as the input variables for the models. Moreover, the performance of the models based on the R-squared value is presented. The results demonstrated that the MLPNN and ARIMA have the best accuracy for the prediction of WPD with the highest correlation coefficient of 0.99 compared to RPFNN. Consequently, it can be concluded that the MLPNN models developed in this study can be attractive for their incorporation in simulators.
KeywordsARIMA MLPNN RPFNN Wind power density
The authors would like to thank the Faculty of Engineering, particularly the Civil Engineering Department and Mechanical Engineering Department of Near East University for their support and encouragement.
- 6.Masseran, N., Razali, A.M., Ibrahim, K., Zaharim, A., Sopian, K.: The probability distribution model of wind speed over East Malaysia. Res. J. Appl. Sci. Eng. Technol. 6(10), 1774–1779 (2013)Google Scholar
- 8.Singh, V.: Application of artificial neural networks for predicting generated wind power. Int. J. Adv. Comput. Sci. Appl. 7(3), 250–253 (2016)Google Scholar