基于支持向量机的动态电能质量扰动分类方法  

Classification method of dynamic power quality disturbances based on SVM

作　　者：宋晓芳[1] 陈劲操[2] 

机构地区：[1]国电自动化研究院,江苏南京210003 [2]南京理工大学动力工程学院,江苏南京210094

出　　处：《电力自动化设备》2006年第4期39-42,共4页Electric Power Automation Equipment

摘　　要：将支持向量机SVM(SupportVectorMachine)引入到动态电能质量分类问题中。在Matlab中编程建立了谐波、电压暂升、电压跌落、瞬时中断、电压波动、瞬变6种常见动态电能质量扰动数学模型,利用傅里叶变换和小波变换对产生的样本波形进行特征提取,产生训练和测试样本。给出了利用LIBSVM解决电能质量扰动分类问题的步骤,并根据分类结果对影响分类效果的参数进行了分析。对训练好的支持向量分类器进行测试,效果良好,当采用C-SVC,RBF核时调整参数可以得到最优分类效果,最高分类率可达到96.67%。The SVM（Support Vector Machine） method is introduced to classification of power quality disturbances. The concerned disturbances,including voltage sags,swells,interruptions,switching transients,flickers and harmonics,are modeled with Matlab. The features of sample waves are extracted by Fourier transform and wavelet transform to form the training and testing samples. The steps of disturbance classification using LIBSVM are described. The influence factors are analyzed according to the classification result, The trained support vector classifier is tested and validated effective. When using C-SVC and RBF kernel,parameters can be adjusted to achieve the optimal effect,while the maximal classification ratio reaches 96.67%.

关 键 词：动态电能质量 支持向量机 分类方法 多类分类 

分 类 号：TM71[电气工程—电力系统及自动化] TP181[自动化与计算机技术—控制理论与控制工程]