Daily wind power profiles determination using clustering algorithms

Wind power will become a major contributor to electricity supply in many countries around the world over the next decades. Therefore, the wind farms owners must be equipped with tools that allow assisting them in managing and planning the wind power generation. With this work it is intended to extract from a database of wind power generation useful knowledge to support the wind farm owners in the management of their wind farms and also to improve their behaviour in the electricity market. To achieve the proposed goal, it was implemented a framework architecture based on the knowledge discovery process, involving several steps, specifically the pre-processing data phase, the application of the several clustering algorithms and finally the evaluation of the algorithms performance based on several clusters validity indices. In order to validate the proposed survey, a case study with a real database from a Portuguese wind farm is used.