基于S变换和最小二乘支持向量机的电能质量扰动识别  

Power Quality Disturbances Identification Based on S-Transform and LS-SVM

作　　者：王学伟[1] 张宏财[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《电测与仪表》2009年第8期1-4,25,共5页Electrical Measurement & Instrumentation

摘　　要：采用S变换和最小二乘支持向量机相结合,构建了一种电能质量扰动识别的新方法。首先利用S变换对电能质量扰动信号进行时频分解;然后,从扰动信号S变换的结果中,提取扰动信号的特征向量,组成训练样本和测试样本;最后,使用最小输出编码的最小二乘支持向量机对扰动信号进行训练,实现电能质量扰动信号自动分类和识别。仿真结果表明,该方法识别准确率高,抗噪能力强,且训练时间很短,适用于电能质量扰动辨识系统。A new method based on S-transform time-frequency analysis and least square support vector machine （LS-SVM） is presented for power quality （PQ） disturbances identification. Firstly using S-transform analyzes PQ disturbances. Then feature components were extracted from the detecting outputs for training. Finally LS-SVM based on minimum output coding is used to classify and identify PQ disturbances. Simulation results indicate that the proposed method has an excellent performance on correct ratio and training speed, and strong resistances to nosies, so it is suitable for classification system for PQ disturbances.

关 键 词：电能质量 扰动识别 S变换 最小二乘支持向量机 

分 类 号：TM711[电气工程—电力系统及自动化]