基于谐波电流差值变化率的滤波器故障在线监测研究  

Research on online monitoring of filter fault based on the change rate of harmonic current differences

作　　者：刘磊[1] 王渝红[1] 李兴源[1] 丁志林[2] 张彪[1] 欧林[3] 

机构地区：[1]四川大学电气信息学院,成都610065 [2]国网四川省电力公司检修公司,成都610041 [3]国网重庆市电力公司电力科学研究院,重庆401123

出　　处：《电测与仪表》2015年第20期1-6,共6页Electrical Measurement & Instrumentation

基　　金：国家高技术研究发展计划(863计划)项目(2011AA05A119);国网四川省电力公司科技项目(13H1134)

摘　　要：滤波器在电力系统中具有举足轻重的作用,为保证滤波器的安全稳定运行,应对滤波器进行有效的监测。本文在分析了滤波器的功能、结构以及谐波电流特点的基础上,提出了新的指标:谐波电流差值变化率,用于滤波器故障在线监测。运用支持向量机(SVM)进行故障分类,通过对不同核函数的分类效果的比较,确定了最优的核函数。通过对双调谐滤波器各元件发生故障时进行仿真,验证了该指标的有效性和SVM分类的准确性,从而说明了本文所提出的在线监测方法的有效性,对实际工程具有一定的价值。Filter has a decisive role in the power system. In order to ensure the safe and stable operation of the filter,effective monitoring is necessary. Based on the analysis of the function,structure and characteristics of harmonic currents of the filter,the changing rate of harmonic current differences is put forward to monitor filter fault on-line in this paper. The support vector machine（ SVM） is used for classifying fault. By comparing the classification results of different kernel functions,the optimal kernel function is determined. Through simulations of each component＇s failure of the double tuned filter,the validity of the index and accuracy of SVM are verified. The validity of the proposed online monitoring is proved. The index has a certain value for actual projects.

关 键 词：滤波器 谐波电流差值变化率 支持向量机 核函数 在线监测 

分 类 号：TM721.1[电气工程—电力系统及自动化]