Wind Data Forecast using Artificial Neural Networks

The economic and environmental benefits of generation of wind energy generation has been attracting much attention, being one of the most promising power sources in Brazil. However, the associated uncertainty is usually large. Thus, wind data must be predicted to effectively mitigate the risks of wind power system operations. Recognizing this challenge, a new approach based on artificial neural networks is proposed for wind forecasting. The proposed method was evaluated using real data obtained from an anemometer station installed at Brazil northeastern region. Results demonstrate that the predicted data have a good match with the real wind speed.