Time-adaptive kernel density forecast: a new method for wind power uncertainty modeling

Abstract

This paper reports new contributions to the advancement of wind power uncertainty forecasting beyond the current state-of-the-art. A new kernel density forecast (KDF) method applied to the wind power problem is described. The method is based on the Nadaraya-Watson estimator, and a time-adaptive version of the algorithm is also proposed. Results are presented for different case-studies and compared with linear and splines quantile regression.