Wind Power Prediction Based on Similar Day Clustering Support Vector Machine

Wind power prediction technology is important to improve the reliability of grid-connected, the common statistic modeling method result is not satisfied because it lacks of effective pretreatment. This paper puts forward wind power prediction based on similar day clustering support vector machine, which catches the training data by similar day and modeling respectively, each model is used to predict specific similar days. Experiment on a wind farm shows the proposed method is effective.