Application of NBTree to selection of meteorological variables in wind speed prediction

This paper proposes a new method for selecting meteorological variables in wind speed prediction. The proposed method is based on the regression tree of data mining. In recent years, the power market becomes more deregulated and competitive. As a result, the distribution generation is introduced in power systems. As clean energy, power utilities are quite interested in wind power generation. However, it is difficult to deal with wind power generation due to the uncertainty of the wind power output. In this paper, an efficient method is proposed to construct rules of data and forecast the wind speed. This paper makes use of NBTree as the prediction model. It is a hybrid model of C4.5 and Nai¿ve Bayes (NB). The proposed method is successfully applied to real data for wind speed.