模糊聚类分析在低频振荡主导模式辨识中的应用

互联系统中，利用振荡曲线提取多机振荡信息，需要选择合适的曲线才能快速得到系统主导振荡模式。文章首先提出了一种基于模糊划分的迭代自组织数据分析技术的聚类方法；然后在一些基本假设的基础上，形成了模糊集合，并运用模糊聚类分析法将系统分区；最后用Prony分析算法从合适的低频振荡信号曲线中准确地提取区域主导振荡模式。通过对中国电力科学研究院8机36节点系统的算例仿真验证了该方法的可行性和有效性。

Application of Fuzzy Clustering Analysis in Identification of Low-Frequency Oscillations Dominant Mode

To extract multi-machine oscillation information of interconnected power grid, it is necessary to choose proper oscillation curves to quickly obtain dominant oscillation mode of power grid.For this reason, the authors propose a fuzzy clustering method in which the fuzzy partitioning based iterative self-organizing data analysis algorithm is used;then on the basis of several basic assumptions, a fuzzy set is formed and the power grid is partitioned by fuzzy clustering analysis;finally, by use of Prony method the regional dominant oscillation mode is accurately extracted from proper low frequency oscillation curve.The results of simulation by an 8-machine 36-bus system in Power System Analysis Software Package, which is developed by China Electric Power Research Institute, validate the feasibility and effectiveness of the proposed method.