基于混合互信息的特征选择方法及其在静态电压稳定评价中的应用.PDF

26 卷 7 期 中 国 电 机 工 程 学 报 Vol.26 No.7 Apr. 2006 2006 年4 月 Proceedings of the CSEE 2006 Chin.Soc.for Elec.Eng. 文章编号：0258-8013 (2006) 07-0077-05 中图分类号：TM71 文献标识码：A 学科分类号：47040 基于混合互信息的特征选择方法及其 在静态电压稳定评估中的应用 皓，孙宏斌，张伯明，郭庆来 (清华大学电机系电力系统国家重点实验室，北京市 海淀区100084) Hybrid Mutual Information Based Feature Selection Method as Appied To Static Voltage Stability Assessment in Power Systems WANG Hao, SUN Hong-bin, ZHANG Bo-ming, GUO Qing-lai (State Key Lab of Power Systems, Department of Electrical Engineering, Tsinghua University, Haidian District, Beijing 100084, China) ABSTRACT: High dimensionality is one of the most 关键词：数据挖掘；特征选择；信息理论；互信息；静态电 troublesome difficulties encountered in power system analysis 压稳定；电力系统 and application. Feature selection is an effective dimension reduction method in data mining field. This paper proposes a 0 引言 new feature selection method, called HMI (Hybrid Mutual 电力系统安全性评估的速度和精度是一对矛 Information), which considers both the novel information a 盾，需要根据实际应用需求来精心处理。近年来， feature can provide and the relevance of the feature with respect to the target feature. In static voltage stability analysis 一些国内外学者试图利用数据挖掘、信息论等技术 application, HMI and other mutual information based feature 来在线实现大规模电力系统的安全评估，取得了不 selection methods are compared, the results show that HMI is [1-10] 少成果 。但在已有的研究中，大多是凭借专家 an effective feature selection method and also show the 经验，人工选择一定数量的特征属性作为数据挖掘 poten