An Improvement in Performance and Computational Cost of ANN Based Wind Speed Prediction System

Wind speed prediction is very important as it influences the wind energy and hence wind farm applications. An accurate wind speed prediction will be helpful in scheduling and management of the wind farms. In this work, empirical mode distribution (EMD) and ensemble empirical mode distribution (EEMD) have been used for decomposition of wind speed data into IMFs. An ANN has been proposed to predict the wind speed with IMFS given as the inputs. The accuracy of prediction by the ANN is further increased by using the actual wind speed at the previous instant along with the IMFs as the input to the ANN. This increase in accuracy in prediction was also accompanied by a reduction in the total computational cost.