摘要
特征值分析法是研究大电网小扰动稳定问题最有效的方法,传统做法通过查看特征向量模态图,可以了解到振荡机组分群的情况。但是由于数据量较大,如果单靠人为看图的分析方式,势必会耗费大量的时间和精力,同时也无法满足在线分析的要求。为此,提出了基于数据聚类的电力系统小干扰稳定机组分群算法。该算法以特征向量之间的夹角作为主要判断依据,辅之以特征向量幅值比较作为挑选代表发电机的方法,从而可以快速辨别出某一振荡模式下的主要参与机组,以及发电机组分群情况,为抑制该模式振荡的发生提供依据,同时满足现代大电网在线小干扰稳定分析要求。仿真及试验结果验证了提出方法的正确性和有效性。
Among small signal stability analysis algorithms for large-scale power grid, eigenvalue analysis method is the most effective. The conventional practice is to analyze the oscillating generator group-dividing by the manual read-ing of the eigenvector modal diagram. However, the huge data volume leads to great waste of time and energy, and the online analysis requiremnt can not be met. Therefore, this paper proposes the online small signal stability genera- tor group-dividing algorithm for power system based on data clustering. Taking the angle between the eigenvectors as the main judgment basis, assisted by eigenvector amplitude comparison as the method to select representative genera- tors, the algorithm can quickly identify the main operating generators in a certain oscillation mode, as well as the gen-erator group-dividing information, which can provide the basis to inhibit the occurrence of such oscillation, and meet the small signal stability analysis requirements of modem large-scale power grid. Simulation and experimental results verify the correctness and effectiveness of the proposed method.
出处
《华东电力》
北大核心
2013年第11期2223-2228,共6页
East China Electric Power
基金
国家电网公司大电网重大专项(SGCC-MPLG001-2012)
国家重点基础研究发展计划资助(2013CB228203)~~
关键词
小干扰稳定
模态分析
机组分群
振荡模式
数据聚类
small signal stability
modal analysis
generator group-dividing
oscillation mode
data clustering
