Application of set pair analysis in wind speed interval prediction for wind farms

Wind farm output power is directly depended on the wind speed, and it is necessary to improve the prediction accuracy of the wind speed. Taking the difficulty of improving the wind speed point forecast accuracy into account, this paper presents a method for wind speed interval prediction based on the set pair analysis theory. This method uses the training data of wind direction, temperature, pressure, humidity, etc., which are the influential factors of wind speed. On the other hand, by considering the error distribution of the wind speed point forecast and the change rate of wind speed, the classified collection which the predictive value of the wind speed in a certain period in the future belongs to is determined. The upper and lower limits of the classified collection are taken as the upper and lower limits of the wind speed prediction interval, thus the wind speed interval prediction is realized. By using the training and prediction data that come from a wind farm in China, the effectiveness of the wind speed interval prediction based on the set pair analysis theory is verified. In fact, the proposed method can also be used for wind power prediction.