摘要
采用模糊数学的方法来辨识电力系统实时运行数据中的不良数据。利用基于模糊等价矩阵的聚类分析方法,以标准残差和相邻采样时刻的量测量差值作为特征值,通过寻找最佳阈值,对量测项目进行动态聚类,根据个别已知的良数据和'数以类聚'的原则,得到全良数据的分类,进而辨识出不良数据。最后分别对传统算例模型和某地区电网实时数据进行仿真分析,表明该方法能快速准确的辨识出不良数据,有效避免残差污染和残差淹没现象,更适合实际电网的计算要求。
Fuzzy mathematics method is used to identify the bad-data in power system real-time data. By use of clustering analysis method based on fuzzy equivalence matrix, and regarding the normalized residual difference and the difference value between adjacent sampling times data as the eigen values, measured items are clustered dynamically by searching the best threshold value. According to individual given good-data and＇like attracts like＇ principle, the good-dataset will be found, and thus the bad-data will be identified. Finally, the traditional instance model and power system real-time data in some place are simulated respectively. The results show that the proposed method not only can identify the bad-data quickly and exactly, but also can avoid residual pollution and residual submersion effectively. It is better to be used in the actual power systems.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2011年第21期1-6,11,共7页
Power System Protection and Control
关键词
电力系统
不良数据辨识
模糊等价矩阵
聚类分析
传递闭包
power system
bad-data identification
fuzzy equivalence matrix~ cluster analysis
transitive closure
作者简介
蒋德珑（1984-），男，硕士研究生，主要研究方向为电力系统稳定分析与控制；E-mail：terrific117@163．com
王克文（1964-），男，博士，教授，主要研究方向为电力系统稳定分析与控制、电力系统自动化等；
王祥东（1986-），男，硕士研究生，研究方向为电力系统稳定分析与控制。
