摘要
受道路颠簸、绝缘老化、接触不良等原因影响,电动汽车可能产生串联型电弧故障。基于干路电流的故障电弧检测方法会对电动汽车变速等工况产生误判。为准确地检测出电动汽车串联电弧故障,搭建了电动汽车串联型故障电弧实验平台,采集了不同速度、不同负载类型下的干路电流信号。通过变分模态分解(VMD)将干路电流信号分解为8个本征模态函数;其次,对电流信号进行了快速傅里叶变换(FFT),结合VMD的结果选择故障特征分量IMF1;对IMF1进行标准化处理,最后将处理后的IMF1分量输入支持向量机网格搜索(GS-SVM)模型进行故障电弧检测,使用十折交叉验证(CV)对模型进行准确率分析。开展了抗干扰实验,结果表明该模型抗干扰性较好,为研发电动汽车的故障电弧检测装置提供了一定的技术支持。
Electric vehicles may experience series arc faults due to factors such as bumpy roads,insulation aging,and poor contact.The fault arc detection method based on the main circuit current will cause misjudgment in working conditions such as speed change of electric vehicles.In order to accurately detect the series arc faults of electric vehicles,an experimental platform for series arc faults of electric vehicles is built,and the main circuit current signals under different speeds and different load types are collected.The main circuit current signal is decomposed into eight intrinsic mode functions through variational mode decomposition(VMD).Secondly,the fast Fourier transform(FFT)is performed on the current signal,and the fault characteristic component IMF1 is selected in combination with the results of VMD.The IMF1 is standardized.Finally,the processed IMF1 components are input into the support vector machine grid search(GSSVM)model for arc fault detection,and the accuracy of the model is analyzed using ten-fold cross-validation(CV).Anti-interference experiments are carried out.The results show that the model has anti-interference performance,providing certain technical support for the research and development of fault arc detection devices for electric vehicles.
作者
崔诗淼
王金龙
刘乙雁
Cui Shimiao;Wang Jinlong;Liu Yiyan(School of Electrical and Control Engineering,Liaoning University of Technology,Huludao 125105,China;Chengde Vocational College of Applied Technology,Scientific Research Center,Chengde 067000,China)
出处
《电力电子技术》
2026年第1期149-157,共9页
Power Electronics
基金
国家自然科学基金(52104160)
省教育厅理工类项目(LJ222410147064)。
关键词
串联故障电弧
电动汽车
变分模态分解
支持向量机网格搜索
series fault arc
electric vehicle
variational mode decomposition
support vector machine grid search
