Application of Gaussian mixture regression model for short-term wind speed forecasting

Abstract

Due to the complex stochastic behavior of the wind environment, there are significant challenges associated in integrating the wind power to the grid. In general, the wind power generation is highly dependent on accurate and reliable wind speed prediction. Therefore, the wind speed forecasting is very important as well greatly influences on the scheduling of a power system and the dynamic control of the wind turbine and resource planning. The main objective of this paper is to forecast the wind speed for short-term from the previous wind speed data. In this paper, a Gaussian Mixture Regression (GMR) model is applied for forecasting the wind-speed for short-term (Two-hours ahead). Moreover, in this work the real-world wind data sets were used in order to model training and testing. The GMR model is simulated in Matlab software. The proposed GMR model is evaluated in terms of Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE) in the task of short-term wind speed prediction. From the simulation results and numerical analysis, it can be stated that the prediction of the short-term wind speed is very accurate by using the GMR methods.