基于OCSVM的火电厂一次风机异常检测
|
摘要
提出了一种基于单分类支持向量机的火电厂一次风机异常检测方法。首先,采用分帧加窗的预处理方法,对一次风机各测点的时域信号提取日平均时域特征,并结合测点原始数据构建特征向量;其次,将训练集特征向量送入单分类支持向量机分类器,构建异常检测模型;最后,利用测试集对模型进行性能评估。理论分析和实验结果表明:所提取的时域特征和构建的单分类异常检测模型对一次风机异常具有明显的辨识度,能有效对一次风机异常状态进行检测。
Aiming at the difficulty of fault diagnosis of primary fan,which is an important auxiliary equipment of thermal power plant,in complex working environment,a method of anomaly detection of primary fan based on one Class Support Vector Machine(OCSVM) is proposed in this paper.Firstly,the framing preprocessing method is used to extract the daily average time-domain features from the signals of each measuring point of primary fan,and the feature vector is constructed by combining the original data of the measured points.Secondly,the training set feature vectors are fed into the one-class SVM classifier to construct the anomaly detection model.Finally,the performance of the model is evaluated by using the test set.The theoretical analysis and experimental results show that the extracted time-domain features and the established oneclass anomaly detection model have obvious identification for primary fan anomalies.
Aiming at the difficulty of fault diagnosis of primary fan,which is an important auxiliary equipment of thermal power plant,in complex working environment,a method of anomaly detection of primary fan based on one Class Support Vector Machine(OCSVM) is proposed in this paper.Firstly,the framing preprocessing method is used to extract the daily average time-domain features from the signals of each measuring point of primary fan,and the feature vector is constructed by combining the original data of the measured points.Secondly,the training set feature vectors are fed into the one-class SVM classifier to construct the anomaly detection model.Finally,the performance of the model is evaluated by using the test set.The theoretical analysis and experimental results show that the extracted time-domain features and the established oneclass anomaly detection model have obvious identification for primary fan anomalies.
引文
[1]金秀章,张琨.基于MSET的一次风机故障预警[J].仪器仪表用户,2018,25(6):61-64,82
[2]马博洋.火电厂一次风机故障预警系统的应用研究[D].北京:华北电力大学,2016
[3]韩平,王天堃,孟永毅.基于LS-SVM的一次风机振动在线监测及故障预警研究[J].机电工程,2016,33(5):629-632
[4]党帅涛.基于图像信息和改进SVDD的绝缘子污秽度异常检测方法研究[D].成都:西南交通大学,2018
[5]曹晨曦,田友琳,张昱堃,等.基于统计方法的异常点检测在时间序列数据上的应用[J].合肥工业大学学报(自然科学版),2018,41(9):1284-1288
[6]林越,刘廷章,黄莉荣,等.基于双向KL距离聚类算法的变压器状态异常检测[J].广西师范大学学报(自然科学版),2018,36(4):20-26
[7]蒋华,季丰,王慧娇,等.改进Kmeans算法的海洋数据异常检测[J].计算机工程与设计,2018,39(10):3132-3136
[8]曾利云,肖云,张植豪.基于时间序列和聚类的挤压机能耗异常检测研究[J].机电工程技术,2018,47(9):32-36
[9]王婵,王慧泉,金仲和.皮纳卫星遥测数据异常检测聚类分析方法[J].哈尔滨工业大学学报,2018,50(4):110-116
[10]Bernhard Sch?lkopf. Support vector method for novelty detection[C]//International Conference on Neural Information Processing Systems. MIT Press, 1999
[11]DAVID M J TAX,ROBERT P W DUIN.Support vector domain description[J].Pattern Recognition Letters,1999,20(11-13):1191-1199
[12]Vapnik V.The nature of statistical learning theory[M].Springer,1995
[13]Platt J C.Fast Training of Support Vector Machines Using Sequential Minimal Optimization[M]//Advances in kernel methods. MIT Press, 1999
[2]马博洋.火电厂一次风机故障预警系统的应用研究[D].北京:华北电力大学,2016
[3]韩平,王天堃,孟永毅.基于LS-SVM的一次风机振动在线监测及故障预警研究[J].机电工程,2016,33(5):629-632
[4]党帅涛.基于图像信息和改进SVDD的绝缘子污秽度异常检测方法研究[D].成都:西南交通大学,2018
[5]曹晨曦,田友琳,张昱堃,等.基于统计方法的异常点检测在时间序列数据上的应用[J].合肥工业大学学报(自然科学版),2018,41(9):1284-1288
[6]林越,刘廷章,黄莉荣,等.基于双向KL距离聚类算法的变压器状态异常检测[J].广西师范大学学报(自然科学版),2018,36(4):20-26
[7]蒋华,季丰,王慧娇,等.改进Kmeans算法的海洋数据异常检测[J].计算机工程与设计,2018,39(10):3132-3136
[8]曾利云,肖云,张植豪.基于时间序列和聚类的挤压机能耗异常检测研究[J].机电工程技术,2018,47(9):32-36
[9]王婵,王慧泉,金仲和.皮纳卫星遥测数据异常检测聚类分析方法[J].哈尔滨工业大学学报,2018,50(4):110-116
[10]Bernhard Sch?lkopf. Support vector method for novelty detection[C]//International Conference on Neural Information Processing Systems. MIT Press, 1999
[11]DAVID M J TAX,ROBERT P W DUIN.Support vector domain description[J].Pattern Recognition Letters,1999,20(11-13):1191-1199
[12]Vapnik V.The nature of statistical learning theory[M].Springer,1995
[13]Platt J C.Fast Training of Support Vector Machines Using Sequential Minimal Optimization[M]//Advances in kernel methods. MIT Press, 1999