Wind Speed Prediction of Four Regions in Northern Cyprus Prediction Using ARIMA and Artificial Neural Networks Models: A Comparison Study

Wind speed data is one of the most critical factors affecting the operation of wind power farm systems. This paper examines the forecasting performance of Auto-Regressive Integrated Moving Average (ARIMA) and Artificial Neural Networks (ANN) models for predicting wind speeds in four regions in Northern Cyprus: Lefkoşa, Girne, Salamis, and Boğaz. For the application of the methodology, the meteorological measurements including wind speed, air temperature, humidity, sunshine duration, global solar radiation and rainfall values, from 1 January 2013 to 31 December 2016, were used. The obtained results demonstrated that the ANN model realizes the best accuracy for the prediction of the wind speeds with the highest R-squared value.