Short term wind speed forecasting based on cluster analysis and ANN in wind farms

The rapidly developing field of data mining and knowledge discovery in data is a blending of artificial intelligence, statistics, and human-computer interaction. The task of data mining is to determine the data pattern from a given set of data. This paper shows the usage of an algorithm of data mining based on clustering, in order to find different data categories in a wind farm data. Once these data categories have been identified, the most suitable prediction technique can be determined. The authors study several Artificial Neural Networks (ANN) topologies in order to establish the most accurate one for every found data category. The experiments were done using real data acquired from a monitoring sensor system.