中文摘要：

场景分析是含风电电力系统处理风电随机性的典型方法,然而由于风电出力对电力系统运行的影响具有复杂的非线性,传统风电场景模型难以保证风电场景与电力系统优化运行保持一致。为此,不同于传统方法的先对风电场景聚类、再进行运行优化,而是先将风电出力样本代入无功/电压运行优化,对优化结果的无功/电压控制矢量进行场景聚类,再映射出风电的场景聚类,从而提出含大规模风电电力系统的无功/电压运行场景模型。考虑到K-means聚类方法难以确定聚类数的问题,通过聚类指标得到运行场景的最佳聚类数。将澳大利亚2个风电场实际数据接入到IEEE 30节点系统中,分别进行传统风电场景分析和所提出的运行场景分析,比较了系统网损和电压的概率特性,验证了所提出的运行场景分析方法的合理性和优越性。

英文摘要：

Scenario analysis is a typical approach to deal with wind power randomness in power systems with wind power integration. However, because impacts of wind power on power system operation are of complex nonlinearity, it is difficult for traditional wind power scene model to ensure optimal operation of power system keeping consistent with wind power scenes. Therefore, a reactive power/voltage operation scene model is proposed in this paper. Firstly, reactive power/voltage operation optimization is conducted with wind power integration. Then, scene clustering for optimal reactive power/voltage control variable vectors is performed to map out scene clustering of wind power. Obviously, the proposed model is different from traditional one where clustering is firstly finished for wind power scene without considering power system operations. Taking into account that it is difficult for K-means clustering method to determine cluster number, David-Bouldin Index （KDBI） is used to obtain optimal cluster number. Finally, actual data of two wind farms in Australia is connected to IEEE 30 node system, and the proposed operation scenario analysis is conducted and compared with traditional wind power scenario analysis in terms of probabilistic characteristics of power system active power loss and voltage. From simulation results, rationality and superiority of the proposed operation scenario clustering analysis method are verified.