基于广义S变换的短时电能质量扰动信号分类.pdfVIP

第33 卷 第5 期 电 网 技 术 Vol. 33 No. 5 2009 年3 月 Power System Technology Mar. 2009 文章编号：1000-3673 （2009 ）05-0022-06 中图分类号：TM714 文献标志码：A 学科代码：470·4051 基于广义 S 变换的短时电能质量扰动信号分类 易吉良，彭建春 （湖南大学 电气与信息工程学院，湖南省 长沙市 410082 ） Classification of Short-Time Power Quality Disturbance Signals Based on Generalized S-Transform YI Ji-liang ，PENG Jian-chun （College of Electrical and Information Engineering，Hunan University ，Changsha 410082，Hunan Province ，China ） ABSTRACT: Based on generalized S-transform, a 有独特的优势，但小波变换计算复杂，对母小波选 classification method for short-time power quality disturbance 择的依赖性较强，并且对噪声比较敏感[8] 。 signals is proposed. First, the generalized S-transform is S 变换作为小波变换和短时傅里叶变换的发 applied to short-time power quality disturbance signals to 展，具有直观的时频特性。文献[9-11]利用 S 变换 obtain modulus matrix; then five kinds of statistic eigenvalues are extracted from the modulus matrix; and then by use of 提取扰动信号的特征量，并分别用模糊模式识别和 decision tree, the subsumption of disturbance signals is 支持向量机作为分类器对扰动信号进行分类，取得 performed to implement automatic classification of short-time 了较好的效果，但分类器稍显复杂。文献[12]利用 power quality disturbance signals. Simulation results show that 决策树进行分类，算法简单，但对于电压凹陷和中 it is not sensitive to noises and its recognition rate is high, so it 断的分类仍采用时域方式实现。文献[13-14]分别用 is suitable to the analysis of power quality disturbance signals 矩阵相似度和专家系统进行分类，效果较好，但计 in actual power systems. 算量较大。 KEY WORDS: short-time power quality disturbance ；