Complete methodology on generating realistic wind speed profiles based on measurements

The wind speed represents the main exogenous signal applied to a Wind Energy Conversion System (WECS) and determines its behavior. The erratic variation of the wind speed, highly dependent on the given site and on the atmospheric conditions, makes the wind speed quite difficult to model. Moreover, wind modelling for medium and large time scales is poorly treated in the present literature. This paper presents methods for generating realistic wind speed profiles based on real measurements. The wind speed profile is divided in a low-frequency component (describing long term variations) and a turbulence component (corresponding to fast, high frequency variations). For modelling the low frequency component two methods are implemented and compared: one based on the Van der Hoven spectrum and one based on auto regressive models. The importance of the turbulence component is also analyzed and modelling methods based on white noise filtering are presented.