Regional new energy power prediction method and device based on proxy scaling mechanism

The invention discloses a regional new energy power prediction method based on a proxy scaling mechanism. The method comprises the following steps: firstly, analyzing new energy output characteristics in a region by using an empirical orthogonal function method; then, a kmeans unsupervised clustering method is adopted to divide sub-regions, and a proxy new energy station is selected by using the correlation coefficient and prediction precision of the new energy station; according to the predicted power and the weight coefficient of the agent new energy station, upscaling prediction of the regional new energy power is completed; and finally completing regional new energy power prediction. According to the invention, the prediction precision of the regional new energy power is ensured, and the dependence of the regional prediction model on the data completeness and prediction precision of a single station is reduced. 本发明公开了一种基于代理缩放机制的区域新能源功率预测方法，包括如下步骤：首先使用经验正交函数法解析区域内新能源出力特征；然后采用kmeans无监督聚类法划分子区域，并利用新能源场站的相关系数