基于多元振动序列共性特征的电抗器故障诊断  

Reactor fault diagnosis based on common feature of multivariate vibration sequences

作　　者：付铭 朱明 梅杰 张静 肖黎[2,3] 张宗喜 FU Ming;ZHU Ming;MEI Jie;ZHANG Jing;XIAO Li;ZHANG Zongxi(School of Electronic Information and Communication,Huazhong University of Science and Technology,Wuhan 430074,China;NARI Group(State Grid Electric Power Research Institute),Nanjing 211106,China;State Grid Electric Power Research Institute Wuhan NARI Group,Wuhan 430074,China;State Grid Sichuan Electric Power Research Institute,Chengdu 610041,China)

机构地区：[1]华中科技大学电子信息与通信学院,武汉430074 [2]南瑞集团(国网电力科学研究院)有限公司,南京211106 [3]国网电力科学研究院武汉南瑞有限责任公司,武汉430074 [4]国网四川省电力公司电力科学研究院,成都610041

出　　处：《电测与仪表》2025年第3期198-207,共10页Electrical Measurement & Instrumentation

基　　金：国家电网有限公司科技项目(52199919000A)。

摘　　要：针对当前特征选择算法只适用于单元振动序列的限制,提出了结合基于支持向量机递归特征消除算法(support vector machine recursive feature elimination algorithm,SVM-RFE)和遗传算法(genetic algorithm,GA)的多元振动序列特征选择算法SVM-RFE-GA。以某220 kV高压并联电抗器为研究对象搭建了机械故障模拟平台,设置了5种不同设备状态,在其表面24个采样位置采集不同设备状态的多元振动序列。实验从时域、频域和时频域出发构建特征池,利用SVM-RFE对单元振动序列中各个特征进行重要性排序和初步筛选,并记录每个位置的最高故障诊断准确率。再对初步筛选出的特征利用GA算法进一步优化选择,选出多元振动序列中具有最高准确率且数量最少的特征组合。实验结果表明,该方法可以选出多元振动序列的共性特征组合,同时该组合可以确保电抗器故障诊断准确率最高且特征数目最少。Aiming at the limitation that current feature selection algorithms are only available for univariate vibra-tion sequence,this paper proposes a multivariate vibration sequences feature selection algorithm named SVM-RFE-GA based on support vector machine recursive feature elimination algorithm(SVM-RFE)and genetic algorithm(GA).Taking a 220 kV high voltage shunt reactor as the research object,we build a mechanical fault simulation platform,set up 5 kinds of equipment states and collect multivariate vibration sequences of different equipment states at 24 sampling positions on its surface.We construct the feature pool from the time domain,frequency do-main and time-frequency domain.For single vibration sequences,we rank the features and select features prelimi-narily by SVM-RFE.Then,the preliminarily select features are further optimized by GA algorithm to select the fea-ture combination with the highest accuracy and the least number.The experimental result shows that the proposed method can select the common feature combination of multivariate vibration sequences,and the combination can en-sure the highest fault diagnosis accuracy and the minimum number of features.

关 键 词：高压并联电抗器 故障诊断 特征选择 递归特征消除算法 遗传算法 

分 类 号：TM472[电气工程—电器]