基于优化型SVM的高压断路器故障诊断方法研究
|
摘要
为准确评估高压断路器的运行状态,提出了一种基于分合闸线圈电流的断路器故障诊断方法。首先提取断路器分合闸线圈电流时间和电流值,并将其作为局部特征;然后对电流信号数据求取均值、标准差、峭度以及能量值并作为全局特征,由局部特征和全局特征组成综合特征;再利用PCA(主分量分析)法对特征矩阵进行降维处理,获得主分量特征;最后构建SVM(支持向量机)的训练样本集,获得SVM决策函数,实现断路器运行状态分类。利用该方法对断路器4种运行状态进行试验分析,结果表明该方法能够有效识别断路器不同故障状态。
In order to accurately evaluate the operation status of high voltage circuit breaker, a fault diagnosis method based on tripping and closing coil current is proposed. Firstly, the time and amplitude of the circuit breaker coil current are extracted as the local feature, and the mean, standard deviation, kurtosis and energy value of the current signal data are taken as the global characteristics to form comprehensive features. Then the principal component analysis(PCA) is used to reduce the dimension of the feature matrix, and the feature of the main component is obtained. Finally, the training sample set of SVM(support vector machine) is constructed to obtain decision function of SVM to classify the operation state of circuit breaker. Four different operating states of circuit breaker are tested and analyzed by this method. The results show that this method can accurately identify different fault states of circuit breaker.
In order to accurately evaluate the operation status of high voltage circuit breaker, a fault diagnosis method based on tripping and closing coil current is proposed. Firstly, the time and amplitude of the circuit breaker coil current are extracted as the local feature, and the mean, standard deviation, kurtosis and energy value of the current signal data are taken as the global characteristics to form comprehensive features. Then the principal component analysis(PCA) is used to reduce the dimension of the feature matrix, and the feature of the main component is obtained. Finally, the training sample set of SVM(support vector machine) is constructed to obtain decision function of SVM to classify the operation state of circuit breaker. Four different operating states of circuit breaker are tested and analyzed by this method. The results show that this method can accurately identify different fault states of circuit breaker.
引文
[1]LEE D,SOHN J,YOO W.Mass effect analysis of a coil spring for the performance evaluation of a gas insulated circuit breaker[J].International Journal of Precision Engineering and Manufacturing,2014(15):2543-2548.
[2]田书,康智慧.基于VMD算法的10 kV高压断路器机械故障诊断分析[J].电力科学与工程,2018,34(6):73-78.
[3]万书亭,肖珊珊,豆龙江,等.档杆故障下高压断路器弹簧操动机构应力分布研究[J].电力科学与工程,2017,33(12):55-60.
[4]陈朋永,赵书涛,李建鹏,等.基于EMD和SVM的高压断路器机械故障诊断方法研究[J].华北电力大学学报(自然科学版),2012,39(6):23-28.
[5]李春锋,孔海洋,王璇,等.基于PCA和聚类的断路器分合闸线圈电流研究[J].电力与能源,2016,37(1):32-36.
[6]赵钰,耿蒲龙,许春雨,等.基于控制回路的断路器永磁机构故障特征提取方法[J].高压电器,2017,53(2):39-46.
[7]杨为,柯艳国,朱胜龙,等.基于样条插值和线性拟合分析的高压断路器弹簧机构状态检测[J].高压电器,2017,53(6):147-153.
[8]周凯峰,李海涛,薛峰,等.GIS分合闸线圈电流特征分析及故障诊断[J].电气自动化,2017,39(2):95-97.
[9]李慧,祁力钧,张建华,等.基于PCA-SVM的棉花出苗期杂草类型识别[J].农业机械学报,2012,43(9):184-189.
[10]刘永斌,何清波,孔凡让,等.基于PCA和SVM的内燃机故障诊断[J].振动测试与诊断,2012,32(2):250-255.
[11]张勇,黄杰,徐科宇.基于PCA-LSSVR算法的WLAN室内定位方法[J].仪器仪表学报,2015,36(2):408-414.
[12]甘锡淞,李云,傅成华,等.基于信息融合和CS-SVM的变压器绕组变形故障诊断方法研究[J].电力系统保护与控制,2018,46(1):156-161.
[13]茆美琴,龚文剑,张榴晨,等.基于EEMD-SVM方法的光伏电站短期出力预测[J].中国电机工程学报,2013,33(34):17-24.
[14]孙曙光,张强,杜太行,等.基于分合闸线圈电流的万能式断路器故障诊断[J].仪器仪表学报,2018,39(2):130-140.
[15]刘轩,孙建平.风力发电机温升故障预警方法研究[J].电力科学与工程,2016,32(6):38-43.
[16]梁银林,刘庆.集成KPCA-SVM的汽轮发电机组故障诊断[J].电力科学与工程,2017,33(1):27-31.
[2]田书,康智慧.基于VMD算法的10 kV高压断路器机械故障诊断分析[J].电力科学与工程,2018,34(6):73-78.
[3]万书亭,肖珊珊,豆龙江,等.档杆故障下高压断路器弹簧操动机构应力分布研究[J].电力科学与工程,2017,33(12):55-60.
[4]陈朋永,赵书涛,李建鹏,等.基于EMD和SVM的高压断路器机械故障诊断方法研究[J].华北电力大学学报(自然科学版),2012,39(6):23-28.
[5]李春锋,孔海洋,王璇,等.基于PCA和聚类的断路器分合闸线圈电流研究[J].电力与能源,2016,37(1):32-36.
[6]赵钰,耿蒲龙,许春雨,等.基于控制回路的断路器永磁机构故障特征提取方法[J].高压电器,2017,53(2):39-46.
[7]杨为,柯艳国,朱胜龙,等.基于样条插值和线性拟合分析的高压断路器弹簧机构状态检测[J].高压电器,2017,53(6):147-153.
[8]周凯峰,李海涛,薛峰,等.GIS分合闸线圈电流特征分析及故障诊断[J].电气自动化,2017,39(2):95-97.
[9]李慧,祁力钧,张建华,等.基于PCA-SVM的棉花出苗期杂草类型识别[J].农业机械学报,2012,43(9):184-189.
[10]刘永斌,何清波,孔凡让,等.基于PCA和SVM的内燃机故障诊断[J].振动测试与诊断,2012,32(2):250-255.
[11]张勇,黄杰,徐科宇.基于PCA-LSSVR算法的WLAN室内定位方法[J].仪器仪表学报,2015,36(2):408-414.
[12]甘锡淞,李云,傅成华,等.基于信息融合和CS-SVM的变压器绕组变形故障诊断方法研究[J].电力系统保护与控制,2018,46(1):156-161.
[13]茆美琴,龚文剑,张榴晨,等.基于EEMD-SVM方法的光伏电站短期出力预测[J].中国电机工程学报,2013,33(34):17-24.
[14]孙曙光,张强,杜太行,等.基于分合闸线圈电流的万能式断路器故障诊断[J].仪器仪表学报,2018,39(2):130-140.
[15]刘轩,孙建平.风力发电机温升故障预警方法研究[J].电力科学与工程,2016,32(6):38-43.
[16]梁银林,刘庆.集成KPCA-SVM的汽轮发电机组故障诊断[J].电力科学与工程,2017,33(1):27-31.