基于GSA的肘形判据用于电力系统不良数据辨识

Application of GSA-based Elbow Judgment on Bad-data Detection of Power System

在分析GSA(gap statistic algorithm)数据挖掘技术应用于电力系统不良数据辨识的基础上,提出一种判断最佳聚类个数的肘形判据,该判据通过分析数据集的聚类离散度与聚类个数k的关系,按照各个k点的聚类离散度计算k处的肘形折角,并以最小肘形折角判断最佳聚类个数。将该判据与GSA相结合用于电力系统不良数据辨识。仿真结果表明:该方法不仅可以避免状态估计方法辨识的残差污染和残差淹没现象,而且可以克服单纯GSA辨识法在计算速度和辨识准确性方面的缺陷。对于大系统、数据量巨大的情况,该方法是一种快速高效的算法,具有很好的应用前景。

Based on bad data detection using GSA （Gap Statistic Algorithm） data mining method in power system, this paper propose the elbow criterion to estimate optimal clustering number. The criterion analyzes the relation between the degree of clustering dispersion and clustering number k of the data set firstly, then calculates the elbow angle at k and obtain the optimal clustering number based on the least elbow angle. Combined the criterion with GSA, bad data detection could be implemented efficiently, Computer results show that the integrated method not only can avoid residual pollution and residual submersion which would appear using traditional state estimate detection, but also is more accurate and rapid than GSA method. In the case of huge system and large amount of data, this method is a rapid and efficient algorithm, and has potential of good application.