Effects of Wind Turbine Power Curve in Wind Speed Prediction Errors

Abstract

Wind power is the fastest growing renewable energy technology and nowadays is becoming a significant component of the energy mix. It is largely recognized that the use and importance of the wind power is much dependent on the ability to accurately predict the wind in advance. Statistical ARMA models are relatively simple and inexpensive forecasting tools. They do not require a huge amount of historical data and they improve the performance of the simplistic reference models. The performance evaluation indexes of the forecasting models are based in the conventional and widely used prediction error, which is defined as the difference between the measured wind speed at a certain time and the predicted wind speed for the same time. Usually, the computation of these errors does not take into account the highly non-linear relationship between wind speed and wind power. This is an important issue, since wind speed prediction is performed with a view to wind power production estimation, as far as power systems related studies are concerned. In this paper, the influence of wind power curve incorporation in the calculation of wind speed forecasting errors is discussed. Furthermore, the work performed in the scope of this paper confirms that ARMA models can improve significantly the performance of reference models, especially when the forecasting time horizon is enlarged.
Copyright © 2013 Praise Worthy Prize - All rights reserved.