Typical scene extraction method and system for wind power system planning

The invention discloses a typical scene extraction method and system for wind power system planning. According to the method, a Gaussian process regression algorithm is adopted to generate a couplingscene of the wind power system; the mutual coupling relationship of uncertain factors is considered; by sub-scheduling scheduling interval partitioning, typical scene extraction is carried out by using a central clustering algorithm in each sub-scheduling interval, and then connection is carried out by using a Cartesian product connection method, so that the technical defect of deviation of an extracted typical scene set due to poor performance of the clustering method on a high-dimensional data set is avoided, and the typical scene extraction precision is improved. 本发明公开了一种风电系统规划的典型场景提取方法及系统。本发明的方法采用高斯过程回归算法生成风电系统的耦合场景，考虑了不确定因素的相互耦合关系，通过子调度区间划分，在每个子调度区间内利用中心聚类算法进行典型场景提取，然后采用笛卡尔积连接法进行连接，避免了聚类方法在高维数据集上的表现较差，造成提取出的典型场景集合的偏差的技术缺陷，提高了典型场景提取的精度。