Improving Wind Power Forecasts: Combination through Multivariate Dimension Reduction Techniques

Wind energy and wind power forecast errors have a direct impact on operational decision problems involved in the integration of this form into electricity system. As relationship between generated is highly nonlinear time-varying, given increasing number available forecasting techniques, it possible to use alternative models obtain more than one prediction for same hour horizon. To increase accuracy, combine different predictions better or dynamically select best each time period. Hybrid alternatives based combining few selected forecasts can be considered when large. One most popular ways estimate coefficients model its past errors. an alternative, we propose using multivariate reduction techniques Markov chain forecasts. The combination thus not directly We show that proposed strategies dimension provide competitive results terms Mean Square Error.