摘要
电能质量扰动现象的准确分类是电能质量领域的热门课题.提出一种基于复阻抗和支持向量机的电能质量扰动分类方法.该方法首先从UCI(University of California,Irvine)数据库中分别提取出各电能质量扰动现象(电压暂降、电压暂升、电压中断、电压振荡、电压脉冲)的实际数据,通过Hilbert变换把扰动电压信号和扰动电流信号转换为相量形式,在此基础上得到复阻抗.接着通过复阻抗提取信号特征,组成特征向量,然后应用支持向量机分类器进行训练、测试和分类.最终对UCI数据库中大量实际扰动数据进行分类,分类取得了良好效果,此效果表明该方法具有一定的应用价值.
It is a popular topic classifying the power quality disturbances precisely in power quality domain. A new classification method was presented in this paper which was based on the complex impedance and support vector machine. Firstly, this method extracted five common power quality disturbances from the UCI database, including voltage sag, voltage swell, voltage interruption, voltage oscillation and voltage impulsion. In order to figure out the complex impedance, the phasor form of disturbing signal should be worked out firstly by using Hilbert transform. Then the features of complex impedance could be extracted for training, testing and classifying in the support vector machine classifier. It was proved that this method can classify the real power quality disturbances from UCI database accurately. So this method was feasible and of certain value in application.
出处
《安徽大学学报（自然科学版）》
CAS
北大核心
2016年第3期58-64,共7页
Journal of Anhui University(Natural Science Edition)
基金
国家自然科学基金资助项目(61172127)
高等学校博士学科点专项科研基金资助项目(20113401110006)
关键词
扰动分类
复阻抗
支持向量机
电能质量
disturbance classification
complex impedance
support vector machine
power quality
