Statistical processing of wind speed data for energy forecast and planning

Abstract

This paper presents a statistical approach to manage wind speed sampled data in order to obtain the forecast of the wind energy potential of a given site. The proposed statistical method is the k-means clustering that allows to extract from a set of experimental measurements the sub-sets of useful data for describing the energy capability of the site. The wind speed distributions in different sites in Sicily, in the south of Italy, have been studied as case study. A suitable wind generator, matching the wind profile of the studied sites, has been selected for the evaluation of the producible energy. It is demonstrated that the use of the proposed method simplifies the problem of the wind plant energy assessment respect to the option of obtaining the desired information by managing a large amount of experimental observations. The proposed method represents a useful tool for an appropriate energy planning in distributed generation systems.