Short-Term Wind Speed Forecasting Using Ensemble Learning

Wind speed forecasting plays a vital role in reliable operation and future planning ofwind turbines in smart grid to meet growing power demand. This article presents ensemble learning based short-term wind speed forecasting. Regression models developed using boosted trees and bagged trees in ensemble learning are used to predict the wind speed. The regression models are trained and tested with historical dataset of Rameswaram located in India. Compared with support vector regression (SVR), ensemble-based wind speed forecasting performs short-term wind speed forecasting effectively.