Article(id=1153375933345026441, tenantId=1146029695717560320, journalId=1146031654075715584, issueId=1153375932170621317, articleNumber=null, orderNo=null, doi=10.13234/j.issn.2095-2805.2024.6.318, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1640275200000, receivedDateStr=2021-12-24, revisedDate=1647100800000, revisedDateStr=2022-03-13, acceptedDate=1649174400000, acceptedDateStr=2022-04-06, onlineDate=1752915851198, onlineDateStr=2025-07-19, pubDate=1732896000000, pubDateStr=2024-11-30, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1752915851198, onlineIssueDateStr=2025-07-19, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1752915851198, creator=13701087609, updateTime=1752915851198, updator=13701087609, issue=Issue{id=1153375932170621317, tenantId=1146029695717560320, journalId=1146031654075715584, year='2024', volume='22', issue='6', pageStart='1', pageEnd='326', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1752915850917, creator=13701087609, updateTime=1753780959263, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1157004459600597839, tenantId=1146029695717560320, journalId=1146031654075715584, issueId=1153375932170621317, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1157004459600597840, tenantId=1146029695717560320, journalId=1146031654075715584, issueId=1153375932170621317, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=318, endPage=326, ext={EN=ArticleExt(id=1153375934993387921, articleId=1153375933345026441, tenantId=1146029695717560320, journalId=1146031654075715584, language=EN, title=Inverter Fault Diagnosis Method Based on CGAN-CNN under Sample Imbalance, columnId=1152281496422806390, journalTitle=Journal of Power Supply, columnName=Reliability and Diagnostics, runingTitle=null, highlight=null, articleAbstract=

The three-phase inverter is an important part of the motor drive system in an electric vehicle (EV). When a fault occurs, the fault sample size will be limited due to the short occurrence time, resulting in sample imbalance. To solve this problem, an inverter fault diagnosis method combining conditional generative adversarial network (CGAN) and convolutional neural network (CNN) is proposed in this paper. First, the phase current is taken as a fault sensitive signal, its frequency-domain characteristics are obtained by fast Fourier transform, and normalized preprocessing is carried out. Then, each sample is labeled and input into the CGAN model for countermeasure training to generate new samples in each fault mode. Finally, the CNN model is used to distinguish various fault modes of inverter. Through experimental research, it is found that the fault diagnosis accuracy based on CGAN-CNN can reach more than 98%, indicating that the proposed sample generation method is better than the traditional Smote and GAN methods. The results in this paper provide theoretical support for the intelligent operation and maintenance of new energy EVs.

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Quan SUN, Fei PENG, Hongsheng LI, Xianghai YU, Guodong SUN), CN=ArticleExt(id=1153376072377815652, articleId=1153375933345026441, tenantId=1146029695717560320, journalId=1146031654075715584, language=CN, title=样本不均衡下基于 CGAN-CNN 的逆变器故障诊断方法, columnId=1149830372659262229, journalTitle=电源学报, columnName=可靠性与诊断, runingTitle=null, highlight=null, articleAbstract=

三相逆变器是电动汽车电机驱动系统的重要部件,当出现故障时因发生时间较短导致故障样本规模受限,进而造成样本不均衡。为解决上述问题,提出1种融合条件生成对抗网络 CGAN(conditional generative adversarial network)与卷积神经网络 CNN(convolutional neural network)的逆变器故障诊断方法。首先将相电流作为故障敏感信号,经快速傅里叶变换 FFT(fast Fourier transform)得到其频域特征,并进行归一化预处理;然后将各样本添加标签后输入 CGAN 模型进行对抗训练,生成各故障模式下的新样本。最后,采用CNN 模型实现逆变器各类故障模式判别。实验结果表明,基于 CGAN-CNN 的故障诊断正确率可达98%以上,说明所提样本生成方法优于传统合成少数类过采样技术 SMOTE(synthetic minority over-sampling technique)方法和生成对抗网络 GAN(generative adversarial network)方法,可为新能源电动汽车智能运维提供理论支撑。

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=i5FHc1YeYey7ofD2nEJ54w==, magXml=qIl4YKi58a7q2Cxstp9WRw==, pdfUrl=null, pdf=AbkKN1N49/2pgTPnvs16vQ==, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=LqmsEE73+byLa2KQKPsCDQ==, mapNumber=null, authorCompany=null, fund=null, authors=

孙权(1987-),男,通信作者,博士,讲师。研究方向:功率变换器故障检测与健康状态评估、机器学习与数据挖掘。E-mail: sequel2005@163.com。

彭飞(1997-),男,硕士。研究方向:故障诊断与深度学习。E-mail: 812887688@qq.com。

李宏胜(1966–), 男,博士,教授。研究方向:数控技术、高性能伺服驱动、机器人控制与智能控制。E-mail: zdhxlhs@njit.edu.cn。

于翔海(1996-),男,硕士。研究方向:故障诊断与深度学习。E-mail: njityxh1996@163.com。

孙国栋(1992-),男,博士研究生。研究方向:电力电子系统故障诊断技术。E-mail: 734305447@qq.com。

, authorsList=孙权, 彭飞, 李宏胜, 于翔海, 孙国栋)}, authors=[Author(id=1154032483768193750, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=sequel2005@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1154032483839496924, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032483768193750, language=EN, stringName=Quan SUN, firstName=Quan, middleName=null, lastName=SUN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1154032483902411487, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032483768193750, language=CN, stringName=孙权, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 南京工程学院 自动化学院 南京 211167, bio={"img":"B/F3Ea3/5e75o0xyOGJDRQ==","content":"

孙权(1987-),男,通信作者,博士,讲师。研究方向:功率变换器故障检测与健康状态评估、机器学习与数据挖掘。E-mail: sequel2005@163.com。

"}, bioImg=B/F3Ea3/5e75o0xyOGJDRQ==, bioContent=

孙权(1987-),男,通信作者,博士,讲师。研究方向:功率变换器故障检测与健康状态评估、机器学习与数据挖掘。E-mail: sequel2005@163.com。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1154032482010780299, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=1, ext=[AuthorCompanyExt(id=1154032482014974604, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China), AuthorCompanyExt(id=1154032483009024689, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 南京工程学院 自动化学院 南京 211167)])]), Author(id=1154032483973714662, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=812887688@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1154032484468642555, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032483973714662, language=EN, stringName=Fei PENG, firstName=Fei, middleName=null, lastName=PENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1154032484560917249, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032483973714662, language=CN, stringName=彭飞, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 南京工程学院 自动化学院 南京 211167, bio={"content":"

彭飞(1997-),男,硕士。研究方向:故障诊断与深度学习。E-mail: 812887688@qq.com。

"}, bioImg=null, bioContent=

彭飞(1997-),男,硕士。研究方向:故障诊断与深度学习。E-mail: 812887688@qq.com。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1154032482010780299, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=1, ext=[AuthorCompanyExt(id=1154032482014974604, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China), AuthorCompanyExt(id=1154032483009024689, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 南京工程学院 自动化学院 南京 211167)])]), Author(id=1154032484628026118, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=zdhxlhs@njit.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1154032485131342614, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032484628026118, language=EN, stringName=Hongsheng LI, firstName=Hongsheng, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1154032485211034393, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032484628026118, language=CN, stringName=李宏胜, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 南京工程学院 自动化学院 南京 211167, bio={"content":"

李宏胜(1966–), 男,博士,教授。研究方向:数控技术、高性能伺服驱动、机器人控制与智能控制。E-mail: zdhxlhs@njit.edu.cn。

"}, bioImg=null, bioContent=

李宏胜(1966–), 男,博士,教授。研究方向:数控技术、高性能伺服驱动、机器人控制与智能控制。E-mail: zdhxlhs@njit.edu.cn。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1154032482010780299, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=1, ext=[AuthorCompanyExt(id=1154032482014974604, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China), AuthorCompanyExt(id=1154032483009024689, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 南京工程学院 自动化学院 南京 211167)])]), Author(id=1154032485265560349, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=njityxh1996@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1154032485588521766, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032485265560349, language=EN, stringName=Xianghai YU, firstName=Xianghai, middleName=null, lastName=YU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1154032485655630634, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032485265560349, language=CN, stringName=于翔海, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 南京工程学院 自动化学院 南京 211167, bio={"content":"

于翔海(1996-),男,硕士。研究方向:故障诊断与深度学习。E-mail: njityxh1996@163.com。

"}, bioImg=null, bioContent=

于翔海(1996-),男,硕士。研究方向:故障诊断与深度学习。E-mail: njityxh1996@163.com。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1154032482010780299, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=1, ext=[AuthorCompanyExt(id=1154032482014974604, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China), AuthorCompanyExt(id=1154032483009024689, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 南京工程学院 自动化学院 南京 211167)])]), Author(id=1154032485722739502, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=734305447@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1154032485798236978, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032485722739502, language=EN, stringName=Guodong SUN, firstName=Guodong, middleName=null, lastName=SUN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 College of Automation Engineering Nanjing University of Aeronautics and Astronautics Nanjing 211106 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1154032485848568629, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, authorId=1154032485722739502, language=CN, stringName=孙国栋, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 南京航空航天大学 自动化学院 南京 211106, bio={"content":"

孙国栋(1992-),男,博士研究生。研究方向:电力电子系统故障诊断技术。E-mail: 734305447@qq.com。

"}, bioImg=null, bioContent=

孙国栋(1992-),男,博士研究生。研究方向:电力电子系统故障诊断技术。E-mail: 734305447@qq.com。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1154032483097105077, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=2, ext=[AuthorCompanyExt(id=1154032483109687990, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032483097105077, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 College of Automation Engineering Nanjing University of Aeronautics and Astronautics Nanjing 211106 China), AuthorCompanyExt(id=1154032483688501970, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032483097105077, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 南京航空航天大学 自动化学院 南京 211106)])])], keywords=[Keyword(id=1154032486196695870, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, orderNo=1, keyword=Fault diagnosis), Keyword(id=1154032486280581952, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, orderNo=2, keyword=sample imbalance), Keyword(id=1154032486343496514, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, orderNo=3, keyword=sample generation), Keyword(id=1154032486452548425, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, orderNo=4, keyword=conditional generative adversarial network (CGAN)), Keyword(id=1154032486519657291, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, orderNo=5, keyword=convolutional neural network (CNN)), Keyword(id=1154032486590960462, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, orderNo=1, keyword=故障诊断), Keyword(id=1154032486674846543, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, orderNo=2, keyword=样本不均衡), Keyword(id=1154032486750344018, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, orderNo=3, keyword=样本生成), Keyword(id=1154032486821647188, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, orderNo=4, keyword=条件生成对抗网络), Keyword(id=1154032486968447831, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, orderNo=5, keyword=卷积神经网络)], refs=[Reference(id=1154032491179529166, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2021, volume=45, issue=null, pageStart=1644, pageEnd=1650, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=Vijayalakshmi S, Hubert Tony Raj L, Palaniyappan S, journalName=Materials Today: Proceedings, refType=null, unstructuredReference=Vijayalakshmi S, Hubert Tony Raj L, Palaniyappan S, et al. A review on multilevel H-bridge cascaded inductor less hybrid inverter for Electric vehicles with PWM control[J]. Materials Today: Proceedings, 2021. 45: 1644-1650., articleTitle=A review on multilevel H-bridge cascaded inductor less hybrid inverter for Electric vehicles with PWM control, refAbstract=null), Reference(id=1154032491255026639, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2019, volume=39, issue=1, pageStart=245, pageEnd=255, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=张建忠, 耿治, 徐帅, journalName=中国电机工程学报, refType=null, unstructuredReference=张建忠, 耿治, 徐帅. 基于T 型逆变器的 APF 故障诊断与容错控制[J]. 中国电机工程学报, 2019. 39(1): 245-255, 339., articleTitle=基于T 型逆变器的 APF 故障诊断与容错控制, refAbstract=null), Reference(id=1154032491347301329, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2019, volume=39, issue=1, pageStart=245, pageEnd=255, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=Zhang Jianzhong, Geng Zhi, Xu Shuai, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=Zhang Jianzhong, Geng Zhi, Xu Shuai. Fault diagnosis and fault tolerant control for APF based on T-type inverter[J]. Proceedings of the CSEE, 2019. 39(1): 245-255, 339 (in Chinese)., articleTitle=Fault diagnosis and fault tolerant control for APF based on T-type inverter, refAbstract=null), Reference(id=1154032491452158931, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2015, volume=62, issue=3, pageStart=1669, pageEnd=1678, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=Caseiro LMA, Mendes AMS, journalName=IEEE Transactions on Industrial Electronics, refType=null, unstructuredReference=Caseiro LMA, Mendes AMS. Real-time IGBT open-circuit fault diagnosis in three-level neutral-point-clamped voltage-source rectifiers based on instant voltage error[J]. IEEE Transactions on Industrial Electronics, 2015. 62(3): 1669-1678., articleTitle=Real-time IGBT open-circuit fault diagnosis in three-level neutral-point-clamped voltage-source rectifiers based on instant voltage error, refAbstract=null), Reference(id=1154032491544433621, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2018, volume=33, issue=4, pageStart=883, pageEnd=891, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=陈勇, 刘志龙, 陈章勇, journalName=电工技术学报, refType=null, unstructuredReference=陈勇, 刘志龙, 陈章勇. 基于电流矢量特征分析的逆变器开路故障快速诊断与定位方法[J]. 电工技术学报, 2018. 33(4): 883-891., articleTitle=基于电流矢量特征分析的逆变器开路故障快速诊断与定位方法, refAbstract=null), Reference(id=1154032491628319703, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2018, volume=33, issue=4, pageStart=883, pageEnd=891, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=Chen Yong, Liu Zhilong, Chen Zhangyong, journalName=Transactions of China Electrotechnical Society, refType=null, unstructuredReference=Chen Yong, Liu Zhilong, Chen Zhangyong. Fast diagnosis and location method for open-circuit fault in inverter based on current vector character analysis[J]. Transactions of China Electrotechnical Society, 2018. 33(4): 883-891 (in Chinese)., articleTitle=Fast diagnosis and location method for open-circuit fault in inverter based on current vector character analysis, refAbstract=null), Reference(id=1154032491733177305, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=6, issue=1, pageStart=52, pageEnd=60, url=null, language=null, rfNumber=[5], rfOrder=6, authorNames=Cheng Yongqi, Dong Weiguang, Gao Fengyang, journalName=Chinese Journal of Electrical Engineering, refType=null, unstructuredReference=Cheng Yongqi, Dong Weiguang, Gao Fengyang, et al. Open-circuit fault diagnosis of traction inverter based on compressed sensing theory[J]. Chinese Journal of Electrical Engineering, 2020. 6(1): 52-60., articleTitle=Open-circuit fault diagnosis of traction inverter based on compressed sensing theory, refAbstract=null), Reference(id=1154032491829646298, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2019, volume=31, issue=9, pageStart=4665, pageEnd=4674, url=null, language=null, rfNumber=[6], rfOrder=7, authorNames=Ji Zhendong, Liu Wei, journalName=Neural Computing and Applications, refType=null, unstructuredReference=Ji Zhendong, Liu Wei. Open-circuit fault detection for three-phase inverter based on backpropagation neural network[J]. Neural Computing and Applications, 2019. 31(9): 4665-4674., articleTitle=Open-circuit fault detection for three-phase inverter based on backpropagation neural network, refAbstract=null), Reference(id=1154032491888366556, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=56, issue=2, pageStart=284, pageEnd=289, url=null, language=null, rfNumber=[7], rfOrder=8, authorNames=陈石, 张兴敢, journalName=南京大学学报(自然科学), refType=null, unstructuredReference=陈石, 张兴敢. 基于小波包能量熵和随机森林的级联H 桥多电平逆变器故障诊断[J]. 南京大学学报(自然科学), 2020. 56(2): 284-289., articleTitle=基于小波包能量熵和随机森林的级联H 桥多电平逆变器故障诊断, refAbstract=null), Reference(id=1154032491955475422, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=56, issue=2, pageStart=284, pageEnd=289, url=null, language=null, rfNumber=[7], rfOrder=9, authorNames=Chen Shi, Zhang Xinggan, journalName=Journal of Nanjing University (Natural Science), refType=null, unstructuredReference=Chen Shi, Zhang Xinggan. Fault diagnosis for cascaded H-bridge multilevel inverter based on wavelet packet energy entropy and random forest[J]. Journal of Nanjing University (Natural Science), 2020. 56(2): 284-289 (in Chinese)., articleTitle=Fault diagnosis for cascaded H-bridge multilevel inverter based on wavelet packet energy entropy and random forest, refAbstract=null), Reference(id=1154032492051944416, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2021, volume=169, issue=null, pageStart=120796, pageEnd=null, url=null, language=null, rfNumber=[8], rfOrder=10, authorNames=Ebrahimi Shahabadi MS, Tabrizchi H, Kuchaki Rafsanjani M, journalName=Technological Forecasting and Social Change, refType=null, unstructuredReference=Ebrahimi Shahabadi MS, Tabrizchi H, Kuchaki Rafsanjani M, et al. A combination of clustering-based under-sampling with ensemble methods for solving imbalanced class problem in intelligent systems[J]. Technological Forecasting and Social Change, 2021. 169: 120796., articleTitle=A combination of clustering-based under-sampling with ensemble methods for solving imbalanced class problem in intelligent systems, refAbstract=null), Reference(id=1154032492140024801, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2019, volume=43, issue=1, pageStart=308, pageEnd=315, url=null, language=null, rfNumber=[9], rfOrder=11, authorNames=李元良, 丁坤, 陈富东, journalName=电网技术, refType=null, unstructuredReference=李元良, 丁坤, 陈富东, 等. 基于快速过采样主成分分析法的光伏阵列故障诊断[J]. 电网技术, 2019. 43(1): 308-315., articleTitle=基于快速过采样主成分分析法的光伏阵列故障诊断, refAbstract=null), Reference(id=1154032492223910883, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2019, volume=43, issue=1, pageStart=308, pageEnd=315, url=null, language=null, rfNumber=[9], rfOrder=12, authorNames=Li Yuanliang, Ding Kun, Chen Fudong, journalName=Power System Technology, refType=null, unstructuredReference=Li Yuanliang, Ding Kun, Chen Fudong, et al. Fault diagnosis method of PV array based on fast OS-PCA[J]. Power System Technology, 2019. 43(1): 308-315 (in Chinese)., articleTitle=Fault diagnosis method of PV array based on fast OS-PCA, refAbstract=null), Reference(id=1154032492320379876, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=46, issue=7, pageStart=2522, pageEnd=2529, url=null, language=null, rfNumber=[10], rfOrder=13, authorNames=刘云鹏, 和家慧, 许自强, journalName=高电压技术, refType=null, unstructuredReference=刘云鹏, 和家慧, 许自强, 等. 基于SVM SMOTE 的电力变压器故障样本均衡化方法[J]. 高电压技术, 2020. 46(7): 2522-2529., articleTitle=基于SVM SMOTE 的电力变压器故障样本均衡化方法, refAbstract=null), Reference(id=1154032492387488741, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=46, issue=7, pageStart=2522, pageEnd=2529, url=null, language=null, rfNumber=[10], rfOrder=14, authorNames=Liu Yunpeng, He Jiahui, Xu Ziqiang, journalName=High Voltage Engineering, refType=null, unstructuredReference=Liu Yunpeng, He Jiahui, Xu Ziqiang, et al. Equalization method of power transformer fault sample based on SVM SMOTE[J]. High Voltage Engineering, 2020. 46(7): 2522-2529 (in Chinese)., articleTitle=Equalization method of power transformer fault sample based on SVM SMOTE, refAbstract=null), Reference(id=1154032492475569127, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2015, volume=22, issue=5, pageStart=2931, pageEnd=2936, url=null, language=null, rfNumber=[11], rfOrder=15, authorNames=Pattanadech N, Yutthagowith P, journalName=IEEE Transactions on Dielectrics and Electrical Insulation, refType=null, unstructuredReference=Pattanadech N, Yutthagowith P. Fast curve fitting algorithm for parameter evaluation in lightning impulse test technique[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2015. 22(5): 2931-2936., articleTitle=Fast curve fitting algorithm for parameter evaluation in lightning impulse test technique, refAbstract=null), Reference(id=1154032492563649513, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=41, issue=2, pageStart=201, pageEnd=204, url=null, language=null, rfNumber=[12], rfOrder=16, authorNames=王圆方, journalName=软件, refType=null, unstructuredReference=王圆方. 基于层次聚类改进 SMOTE 的过采样方法[J]. 软件, 2020. 41(2): 201-204., articleTitle=基于层次聚类改进 SMOTE 的过采样方法, refAbstract=null), Reference(id=1154032492651729899, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2020, volume=41, issue=2, pageStart=201, pageEnd=204, url=null, language=null, rfNumber=[12], rfOrder=17, authorNames=Wang Yuanfang, journalName=Computer Engineering & Software, refType=null, unstructuredReference=Wang Yuanfang. Hybrid algorithm of aggregation hierarchy clustering and SMOTE for oversampling[J]. Computer Engineering & Software, 2020. 41(2): 201-204 (in Chinese)., articleTitle=Hybrid algorithm of aggregation hierarchy clustering and SMOTE for oversampling, refAbstract=null), Reference(id=1154032492723033072, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2014, volume=null, issue=null, pageStart=2672, pageEnd=2680, url=null, language=null, rfNumber=[13], rfOrder=18, authorNames=null, journalName=Proceedings of the 27th International Conference on Neural Information Processing Systems, refType=null, unstructuredReference=Goodfellow I J, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets[C]// Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal, Quebec, Canada: MIT Press, 2014: 2672-2680., articleTitle=Generative adversarial nets, refAbstract=null), Reference(id=1154032492790141941, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2019, volume=32, issue=3, pageStart=310, pageEnd=319, url=null, language=null, rfNumber=[14], rfOrder=19, authorNames=Wang Junliang, Yang Zhengliang, Zhang Jie, journalName=IEEE Transactions on Semiconductor Manufacturing, refType=null, unstructuredReference=Wang Junliang, Yang Zhengliang, Zhang Jie, et al. AdaBalGAN: An improved generative adversarial network with imbalanced learning for wafer defective pattern recognition[J]. IEEE Transactions on Semiconductor Manufacturing, 2019. 32(3): 310-319., articleTitle=AdaBalGAN: An improved generative adversarial network with imbalanced learning for wafer defective pattern recognition, refAbstract=null), Reference(id=1154032492874028024, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[15], rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Mirza M,Osindero S. Conditional generative adversarial nets[EB/OL]. (2014-11-06) [2020-06-10]. https://arxiv.org/abs/1411.1784., articleTitle=Conditional generative adversarial nets, refAbstract=null), Reference(id=1154032492932748281, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, doi=null, pmid=null, pmcid=null, year=2017, volume=6, issue=null, pageStart=6273, pageEnd=6285, url=null, language=null, rfNumber=[16], rfOrder=21, authorNames=Sun Quan, Wang Youren, Jiang Yuanyuan, journalName=IEEE Access, refType=null, unstructuredReference=Sun Quan, Wang Youren, Jiang Yuanyuan. A novel fault diagnostic approach for DC-DC converters based on CSA-DBN[J]. IEEE Access, 2017. 6: 6273-6285., articleTitle=A novel fault diagnostic approach for DC-DC converters based on CSA-DBN, refAbstract=null)], funds=[Fund(id=1154032490672018362, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, awardId=61901212, language=EN, fundingSource=National Natural Science Foundation of China(61901212), fundOrder=null, country=null), Fund(id=1154032490739127229, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, awardId=61901212, language=CN, fundingSource=国家自然科学基金资助项目(61901212), fundOrder=null, country=null), Fund(id=1154032490797847488, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, awardId=20KJA510007, language=EN, fundingSource=Natural Science Foundation of Jiangsu Higher Education Institutions of China(20KJA510007), fundOrder=null, country=null), Fund(id=1154032490856567747, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, awardId=20KJA510007, language=CN, fundingSource=江苏省高等学校自然科学研究重大资助项目(20KJA510007), fundOrder=null, country=null), Fund(id=1154032490927870918, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, awardId=XTCX201909, language=EN, fundingSource=Open Research Fund of Jiangsu Collaborative Innovation Center for Smart Distribution Network(XTCX201909), fundOrder=null, country=null), Fund(id=1154032491007562698, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, awardId=XTCX201909, language=CN, fundingSource=江苏省配电网智能技术与装备协同创新中心开放基金资助项目(XTCX201909), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1154032482010780299, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=1, ext=[AuthorCompanyExt(id=1154032482014974604, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Automation Nanjing Institute of Technology Nanjing 211167 China), AuthorCompanyExt(id=1154032483009024689, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032482010780299, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 南京工程学院 自动化学院 南京 211167)]), AuthorCompany(id=1154032483097105077, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, xref=2, ext=[AuthorCompanyExt(id=1154032483109687990, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032483097105077, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 College of Automation Engineering Nanjing University of Aeronautics and Astronautics Nanjing 211106 China), AuthorCompanyExt(id=1154032483688501970, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, companyId=1154032483097105077, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 南京航空航天大学 自动化学院 南京 211106)])], figs=[ArticleFig(id=1154032487887000431, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 1, caption=Schematic of GAN, figureFileSmall=n05ffqEXkcoCwCerMq85Og==, figureFileBig=I7NkdbWVM2tG2HbsqrkcKg==, tableContent=null), ArticleFig(id=1154032487958303601, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图1, caption=GAN 原理结构, figureFileSmall=n05ffqEXkcoCwCerMq85Og==, figureFileBig=I7NkdbWVM2tG2HbsqrkcKg==, tableContent=null), ArticleFig(id=1154032488021218165, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 2, caption=Schematic of CGAN, figureFileSmall=35IbKedCZ3p6IvJ2A+IvBQ==, figureFileBig=8R2uj7vaBZeh6LlGVUBQMA==, tableContent=null), ArticleFig(id=1154032488117687159, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图2, caption=CGAN 原理结构, figureFileSmall=35IbKedCZ3p6IvJ2A+IvBQ==, figureFileBig=8R2uj7vaBZeh6LlGVUBQMA==, tableContent=null), ArticleFig(id=1154032488188990331, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 3, caption=Schematic of CNN, figureFileSmall=szvZFoLK0BENvWiMapBANQ==, figureFileBig=4cCCcrJMKmZhtS1fapJOhQ==, tableContent=null), ArticleFig(id=1154032488264487806, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图3, caption=CNN 原理结构, figureFileSmall=szvZFoLK0BENvWiMapBANQ==, figureFileBig=4cCCcrJMKmZhtS1fapJOhQ==, tableContent=null), ArticleFig(id=1154032488331596673, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 4, caption=Structure of CNN network model, figureFileSmall=vd7qm9XrPDLqo1lH5W63GQ==, figureFileBig=6Pf2/9qkiNn9ksx0Mpt7jw==, tableContent=null), ArticleFig(id=1154032488419677060, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图4, caption=CNN 网络模型结构, figureFileSmall=vd7qm9XrPDLqo1lH5W63GQ==, figureFileBig=6Pf2/9qkiNn9ksx0Mpt7jw==, tableContent=null), ArticleFig(id=1154032488499368840, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 5, caption=Flow chart of fault diagnosis based on CGAN-CNN, figureFileSmall=iJnE1+9bJqKfIfMH0GBYLQ==, figureFileBig=E/2t3YNZfKpQvJBPG8o7pQ==, tableContent=null), ArticleFig(id=1154032488587449226, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图5, caption=CGAN-CNN 故障诊断流程, figureFileSmall=iJnE1+9bJqKfIfMH0GBYLQ==, figureFileBig=E/2t3YNZfKpQvJBPG8o7pQ==, tableContent=null), ArticleFig(id=1154032488667141003, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 6, caption=Experimental platform of BLDCM drive system fault detection, figureFileSmall=6bcfJ26sdf6Y970Ff1KUDw==, figureFileBig=G8ucyxOmHyk8Nn9Nf15Eng==, tableContent=null), ArticleFig(id=1154032488780387213, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图6, caption=BLDCM 驱动系统故障检测实验平台, figureFileSmall=6bcfJ26sdf6Y970Ff1KUDw==, figureFileBig=G8ucyxOmHyk8Nn9Nf15Eng==, tableContent=null), ArticleFig(id=1154032488948159376, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 7, caption=Waveforms in ${\mathrm{{FC}}}_{0},{\mathrm{{FC}}}_{4},{\mathrm{{FC}}}_{5}$ and ${\mathrm{{FC}}}_{6}$ under load ${}_{1}$, figureFileSmall=EHPmMOKUIZB706f6bAWwOw==, figureFileBig=lgc3rZPzcafuztsJkoGuCg==, tableContent=null), ArticleFig(id=1154032489011073937, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图7, caption=${\mathrm{{load}}}_{1}$ 情形下 ${\mathrm{{FC}}}_{0}$ 、 ${\mathrm{{FC}}}_{4}$ 、 ${\mathrm{{FC}}}_{5}$ 和 ${\mathrm{{FC}}}_{6}$ 状态下 ${i}_{\mathrm{a}}$ 的波形, figureFileSmall=EHPmMOKUIZB706f6bAWwOw==, figureFileBig=lgc3rZPzcafuztsJkoGuCg==, tableContent=null), ArticleFig(id=1154032489061405587, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 8, caption=Visual clustering of generated samples using t-SNE under load2, figureFileSmall=cXLylnzriTGEtxFP9Wt6Gw==, figureFileBig=aZ0EsMIIftqmfDxe2uehvA==, tableContent=null), ArticleFig(id=1154032489132708757, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图8, caption=${\mathrm{{load}}}_{2}$ 情形下生成样本 t-SNE 可视化聚类, figureFileSmall=cXLylnzriTGEtxFP9Wt6Gw==, figureFileBig=aZ0EsMIIftqmfDxe2uehvA==, tableContent=null), ArticleFig(id=1154032489216594838, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 9, caption=CNN model fault diagnosis confusion matrix in input modes of sample sets ${A}_{1},{A}_{2}$ and ${A}_{3}$, figureFileSmall=lXQ9K9hW2GcAyWgzVK/Edg==, figureFileBig=ZL5aG6BH38BaZg2ckkpIoQ==, tableContent=null), ArticleFig(id=1154032489279509400, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图9, caption=样本集 ${A}_{1}$ 、 ${A}_{2}$ 、 ${A}_{3}$ 输入方式下的 CNN 模型故障诊断混淆矩阵, figureFileSmall=lXQ9K9hW2GcAyWgzVK/Edg==, figureFileBig=ZL5aG6BH38BaZg2ckkpIoQ==, tableContent=null), ArticleFig(id=1154032489346618266, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 10, caption=Comparison of classification accuracy of CNN model in input modes of sample sets ${A}_{1},{A}_{2}$ and ${A}_{3}$, figureFileSmall=+aWVW4gHtZIJGeQxTxu+7A==, figureFileBig=0uElJNo1T+uw4ccP8T+vCw==, tableContent=null), ArticleFig(id=1154032489405338524, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图10, caption=3种样本集 ${A}_{1}$ 、 ${A}_{2}$ 和 ${A}_{3}$ 输入方式下的 CNN 模型分类准确率对比, figureFileSmall=+aWVW4gHtZIJGeQxTxu+7A==, figureFileBig=0uElJNo1T+uw4ccP8T+vCw==, tableContent=null), ArticleFig(id=1154032489459864478, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 11, caption=SMOTE-CNN, GAN-CNN and CGAN-CNN fault diagnosis confusion matrixes in input mode of sample set${A}_{4}$, figureFileSmall=UVIs8D4WgSNvVf1VGOZUIw==, figureFileBig=tt6Pqsb3yZJve1wUqGxCSA==, tableContent=null), ArticleFig(id=1154032489535361951, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图11, caption=样本集 ${A}_{4}$ 输入方式下的 SMOTE-CNN、GAN-CNN 和 CGAN-CNN 故障诊断混淆矩阵, figureFileSmall=UVIs8D4WgSNvVf1VGOZUIw==, figureFileBig=tt6Pqsb3yZJve1wUqGxCSA==, tableContent=null), ArticleFig(id=1154032489619248033, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Fig. 12, caption=Comparison of classification accuracy among SMOTE-CNN, GAN-CNN and CGAN-CNN in input mode of sample set ${A}_{4}$, figureFileSmall=JQWeMBCTgJan72OEIneJ2Q==, figureFileBig=LgQClnTYTmvMKj7DRteIQw==, tableContent=null), ArticleFig(id=1154032489753465764, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=图12, caption=样本集 ${A}_{4}$ 输入方式下的 SMOTE-CNN、GAN-CNN 和 CGAN-CNN 分类准确率对比, figureFileSmall=JQWeMBCTgJan72OEIneJ2Q==, figureFileBig=LgQClnTYTmvMKj7DRteIQw==, tableContent=null), ArticleFig(id=1154032489824768932, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Tab. 1, caption=Fault modes of three-phase full-bridge inverter, figureFileSmall=null, figureFileBig=null, tableContent=
故障编号 故障模式 故障编号 故障模式
${\mathrm{{FC}}}_{0}$ 正常状态 ${\mathrm{{FC}}}_{4}$ ${\mathrm{T}}_{4}$ 管开路
${\mathrm{{FC}}}_{1}$ ${\mathrm{T}}_{1}$ 管开路 ${\mathrm{{FC}}}_{5}$ ${\mathrm{T}}_{5}$ 管开路
${\mathrm{{FC}}}_{2}$ ${\mathrm{T}}_{2}$ 管开路 ${\mathrm{{FC}}}_{6}$ ${\mathrm{T}}_{6}$ 管开路
${\mathrm{{FC}}}_{3}$ ${\mathrm{T}}_{3}$ 管开路
), ArticleFig(id=1154032489904460710, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=表1, caption=三相全桥逆变器结构故障模式, figureFileSmall=null, figureFileBig=null, tableContent=
故障编号 故障模式 故障编号 故障模式
${\mathrm{{FC}}}_{0}$ 正常状态 ${\mathrm{{FC}}}_{4}$ ${\mathrm{T}}_{4}$ 管开路
${\mathrm{{FC}}}_{1}$ ${\mathrm{T}}_{1}$ 管开路 ${\mathrm{{FC}}}_{5}$ ${\mathrm{T}}_{5}$ 管开路
${\mathrm{{FC}}}_{2}$ ${\mathrm{T}}_{2}$ 管开路 ${\mathrm{{FC}}}_{6}$ ${\mathrm{T}}_{6}$ 管开路
${\mathrm{{FC}}}_{3}$ ${\mathrm{T}}_{3}$ 管开路
), ArticleFig(id=1154032489992541096, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Tab. 2, caption=Structure of CGAN model, figureFileSmall=null, figureFileBig=null, tableContent=
组成部分 结构 激活函数 优化器
判别器 卷积层 1 Relu Adam
Maxpool 池化层 1 卷积层 2 Maxpool 池化层 2 Relu
全连接层 1 Relu
全连接层 2 Sigmoid
生成器 全连接层 1
卷积层 1 Relu
卷积层 2 Relu
卷积层 3 Tanh
), ArticleFig(id=1154032490055455657, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=表2, caption=CGAN 模型结构, figureFileSmall=null, figureFileBig=null, tableContent=
组成部分 结构 激活函数 优化器
判别器 卷积层 1 Relu Adam
Maxpool 池化层 1 卷积层 2 Maxpool 池化层 2 Relu
全连接层 1 Relu
全连接层 2 Sigmoid
生成器 全连接层 1
卷积层 1 Relu
卷积层 2 Relu
卷积层 3 Tanh
), ArticleFig(id=1154032490139341738, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Tab. 3, caption=Parameters of CNN network model, figureFileSmall=null, figureFileBig=null, tableContent=
隐含层 卷积核 尺寸 卷积核 数量 卷积 步长 输出尺寸
输入层 ${1500}\times 1 \times 1$
卷积层 C1 $3 \times 1$ 8 1 ${1500}\times 1 \times 8$
池化层 P1 $2 \times 1$ 8 2 ${750}\times 1 \times 8$
卷积层 C2 $3 \times 1$ 16 1 ${750}\times 1 \times {16}$
池化层 P2 $2 \times 1$ 16 2 ${375}\times 1 \times {16}$
全连接层 ${6000}\times 1$
Softmax 7 1 7
), ArticleFig(id=1154032490235810734, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=表3, caption=CNN 网络模型参数, figureFileSmall=null, figureFileBig=null, tableContent=
隐含层 卷积核 尺寸 卷积核 数量 卷积 步长 输出尺寸
输入层 ${1500}\times 1 \times 1$
卷积层 C1 $3 \times 1$ 8 1 ${1500}\times 1 \times 8$
池化层 P1 $2 \times 1$ 8 2 ${750}\times 1 \times 8$
卷积层 C2 $3 \times 1$ 16 1 ${750}\times 1 \times {16}$
池化层 P2 $2 \times 1$ 16 2 ${375}\times 1 \times {16}$
全连接层 ${6000}\times 1$
Softmax 7 1 7
), ArticleFig(id=1154032490302919600, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=EN, label=Tab. 5, caption=Unbalanced proportion of each sample set, figureFileSmall=null, figureFileBig=null, tableContent=
样本集 不平衡比例
${A}_{1}$ $1 : 1 : 1 : 1 : 1 : 1$
${A}_{2}$ 2:2:2:2:2:2:1
${A}_{3}$ 10:10:10:10:10:10:3
${A}_{4}$ $1 : 1 : 1 : 1 : 1 : 1$
), ArticleFig(id=1154032490474886068, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1153375933345026441, language=CN, label=表5, caption=各样本集不平衡比例, figureFileSmall=null, figureFileBig=null, tableContent=
样本集 不平衡比例
${A}_{1}$ $1 : 1 : 1 : 1 : 1 : 1$
${A}_{2}$ 2:2:2:2:2:2:1
${A}_{3}$ 10:10:10:10:10:10:3
${A}_{4}$ $1 : 1 : 1 : 1 : 1 : 1$
)], attaches=null, journal=Journal(id=1046111678587809797, delFlag=0, nameCn=电源学报, nameEn=Journal of Power Supply, nameHistory1=null, nameHistory2=null, issn=2095-2805, eissn=, cn=12-1420/TM, coden=null, periodic=bio-monthly, language=CN, oaType=是, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=Mx+A2dn+ULnPHuEAI1LruQ==, journalPrice=null, startedYear=null, abbrevIsoEn=J Power Supp, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1759802942253, createdBy=null, updatedBy=18614031015, firstLetterCn=J, firstLetterEn=J, subjectCode=Engineering, subjectName=工程, subjectCodeEn=Engineering, subjectNameEn=null, picCn=Mx+A2dn+ULnPHuEAI1LruQ==, picEn=yHt2vwjzkDgqh+JDCfJKoQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1162453073839375337, language=CN, name=电源学报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.jops.cn/EN/home, createdTime=1755080010137, updatedTime=1755080010137, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.jops.cn/CN/column/column7.shtml, submissionAuthorUrl=https://mc03.manuscriptcentral.com/jops, submissionEditorUrl=https://mc03.manuscriptcentral.com/jops, submissionReviewUrl=https://mc03.manuscriptcentral.com/jops, submissionCeEditorUrl=https://mc03.manuscriptcentral.com/jops, submissionAeEditorUrl=https://mc03.manuscriptcentral.com/jops, option={"copyright":""}), JournalExt(id=1162453073902289898, language=EN, name=Journal of Power Supply, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.jops.cn/CN/home, createdTime=1755080010152, updatedTime=1755080010152, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.jops.cn/EN/column/column7.shtml, submissionAuthorUrl=https://mc03.manuscriptcentral.com/jops, submissionEditorUrl=https://mc03.manuscriptcentral.com/jops, submissionReviewUrl=https://mc03.manuscriptcentral.com/jops, submissionCeEditorUrl=https://mc03.manuscriptcentral.com/jops, submissionAeEditorUrl=https://mc03.manuscriptcentral.com/jops, option={"copyright":""})], databaseList=null, tenantJournalId=1146031654075715584, websiteList=[Website(id=1146832214672683008, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146031654075715584, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/dyxb/EN, language=EN, createTime=1751355707101, createBy=18614031015, updateTime=1753435268747, updateBy=18614031015, name=电源学报-英文站点, tplId=1146101810881728533, title=电源学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155559379819679852, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1146832214672683008, code=articleTextType, value=kx, createTime=1753436425404, updateTime=1753436425404, creator=18614031015, updator=18614031015), WebsiteProps(id=1155559379798708329, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1146832214672683008, code=banner, value=null, createTime=1753436425399, updateTime=1753436425399, creator=18614031015, updator=18614031015), WebsiteProps(id=1155559379781931112, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1146832214672683008, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=efYTu4aDDzS8GgTA1MjEKw==, createTime=1753436425396, updateTime=1753436425396, creator=18614031015, updator=18614031015), WebsiteProps(id=1155559379811291243, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1146832214672683008, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753436425402, updateTime=1753436425402, creator=18614031015, updator=18614031015), WebsiteProps(id=1155559379802902634, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1146832214672683008, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753436425400, updateTime=1753436425400, creator=18614031015, updator=18614031015)]), Website(id=1148243202240405915, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146031654075715584, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/dyxb/CN, language=CN, createTime=1751692112741, createBy=18614031015, updateTime=1753435242839, updateBy=18614031015, name=电源学报-中文站点, tplId=1146099689490845704, title=电源学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148618015060553758, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202240405915, code=articleTextType, value=kx, createTime=1751781475081, updateTime=1751781475081, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618015035387931, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202240405915, code=banner, value=null, createTime=1751781475075, updateTime=1751781475075, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618015022805018, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202240405915, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=efYTu4aDDzS8GgTA1MjEKw==, createTime=1751781475072, updateTime=1751781475072, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618015052165149, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202240405915, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751781475079, updateTime=1751781475079, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618015043776540, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202240405915, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751781475077, updateTime=1751781475077, creator=18614031015, updator=18614031015)])], journalTitle=电源学报, weixinUrl=null, journalUrl=http://www.jops.cn/CN/home, iacademicId=null, status=0, seqNo=null, journalTitleEn=Journal of Power Supply, journalPhotoCn=Mx+A2dn+ULnPHuEAI1LruQ==, journalPhotoEn=yHt2vwjzkDgqh+JDCfJKoQ==, journalFirstLetter=J, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=null, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/dyxb/CN/10.13234/j.issn.2095-2805.2024.6.318, detailUrlEn=https://castjournals.cast.org.cn/joweb/dyxb/EN/10.13234/j.issn.2095-2805.2024.6.318, pdfUrlCn=https://castjournals.cast.org.cn/joweb/dyxb/CN/PDF/10.13234/j.issn.2095-2805.2024.6.318, pdfUrlEn=https://castjournals.cast.org.cn/joweb/dyxb/EN/PDF/10.13234/j.issn.2095-2805.2024.6.318, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
样本不均衡下基于 CGAN-CNN 的逆变器故障诊断方法
收藏切换
PDF下载
孙权 1 , 彭飞 1 , 李宏胜 1 , 于翔海 1 , 孙国栋 2
电源学报 | 可靠性与诊断 2024,22(6): 318-326
收起
收藏切换
电源学报 | 可靠性与诊断 2024, 22(6): 318-326
样本不均衡下基于 CGAN-CNN 的逆变器故障诊断方法
全屏
孙权1 , 彭飞1 , 李宏胜1 , 于翔海1 , 孙国栋2
作者信息
  • 1 南京工程学院 自动化学院 南京 211167
  • 2 南京航空航天大学 自动化学院 南京 211106
  • 孙权(1987-),男,通信作者,博士,讲师。研究方向:功率变换器故障检测与健康状态评估、机器学习与数据挖掘。E-mail: sequel2005@163.com。

    彭飞(1997-),男,硕士。研究方向:故障诊断与深度学习。E-mail: 812887688@qq.com。

    李宏胜(1966–), 男,博士,教授。研究方向:数控技术、高性能伺服驱动、机器人控制与智能控制。E-mail: zdhxlhs@njit.edu.cn。

    于翔海(1996-),男,硕士。研究方向:故障诊断与深度学习。E-mail: njityxh1996@163.com。

    孙国栋(1992-),男,博士研究生。研究方向:电力电子系统故障诊断技术。E-mail: 734305447@qq.com。

Inverter Fault Diagnosis Method Based on CGAN-CNN under Sample Imbalance
Quan SUN1 , Fei PENG1 , Hongsheng LI1 , Xianghai YU1 , Guodong SUN2
Affiliations
  • 1 School of Automation Nanjing Institute of Technology Nanjing 211167 China
  • 2 College of Automation Engineering Nanjing University of Aeronautics and Astronautics Nanjing 211106 China
出版时间: 2024-11-30 doi: 10.13234/j.issn.2095-2805.2024.6.318
文章导航
收藏切换

三相逆变器是电动汽车电机驱动系统的重要部件,当出现故障时因发生时间较短导致故障样本规模受限,进而造成样本不均衡。为解决上述问题,提出1种融合条件生成对抗网络 CGAN(conditional generative adversarial network)与卷积神经网络 CNN(convolutional neural network)的逆变器故障诊断方法。首先将相电流作为故障敏感信号,经快速傅里叶变换 FFT(fast Fourier transform)得到其频域特征,并进行归一化预处理;然后将各样本添加标签后输入 CGAN 模型进行对抗训练,生成各故障模式下的新样本。最后,采用CNN 模型实现逆变器各类故障模式判别。实验结果表明,基于 CGAN-CNN 的故障诊断正确率可达98%以上,说明所提样本生成方法优于传统合成少数类过采样技术 SMOTE(synthetic minority over-sampling technique)方法和生成对抗网络 GAN(generative adversarial network)方法,可为新能源电动汽车智能运维提供理论支撑。

故障诊断  /  样本不均衡  /  样本生成  /  条件生成对抗网络  /  卷积神经网络

The three-phase inverter is an important part of the motor drive system in an electric vehicle (EV). When a fault occurs, the fault sample size will be limited due to the short occurrence time, resulting in sample imbalance. To solve this problem, an inverter fault diagnosis method combining conditional generative adversarial network (CGAN) and convolutional neural network (CNN) is proposed in this paper. First, the phase current is taken as a fault sensitive signal, its frequency-domain characteristics are obtained by fast Fourier transform, and normalized preprocessing is carried out. Then, each sample is labeled and input into the CGAN model for countermeasure training to generate new samples in each fault mode. Finally, the CNN model is used to distinguish various fault modes of inverter. Through experimental research, it is found that the fault diagnosis accuracy based on CGAN-CNN can reach more than 98%, indicating that the proposed sample generation method is better than the traditional Smote and GAN methods. The results in this paper provide theoretical support for the intelligent operation and maintenance of new energy EVs.

Fault diagnosis  /  sample imbalance  /  sample generation  /  conditional generative adversarial network (CGAN)  /  convolutional neural network (CNN)
孙权, 彭飞, 李宏胜, 于翔海, 孙国栋. 样本不均衡下基于 CGAN-CNN 的逆变器故障诊断方法. 电源学报, 2024 , 22 (6) : 318 -326 . DOI: 10.13234/j.issn.2095-2805.2024.6.318
Quan SUN, Fei PENG, Hongsheng LI, Xianghai YU, Guodong SUN. Inverter Fault Diagnosis Method Based on CGAN-CNN under Sample Imbalance[J]. Journal of Power Supply, 2024 , 22 (6) : 318 -326 . DOI: 10.13234/j.issn.2095-2805.2024.6.318
随着新能源汽车产业的飞速发展, 无刷直流电机BLDCM(brushless direct current motor)作为新能源汽车动力系统的重要部件, 已广泛应用于纯电动汽车、混合动力电动汽车中[1]。功率逆变器作为无刷直流电机驱动系统的核心组成部分, 其安全、稳定运行对于提高整个驱动系统的可靠性具有重大影响。
由于逆变器系统运行工况的复杂多变及其自身老化,功率开关管发生故障的可能性将会大幅增加[2]。短路和开路故障是功率开关管故障的主要类型[3]。当功率开关管发生短路时,由于整个过程比较短暂,对此故障的诊断可以直接通过硬件保护。 但是对于功率开关管的开路故障而言, 由于其缓慢的响应特性, 普通诊断方法无法直接诊断逆变器的开路故障, 同时长时间在故障条件下工作会使逆变器开关管发生二次故障[4]。因此,为了保证 BLDCM 驱动系统安全可靠的运行且减少经济损失, 开展逆变器系统的故障诊断具有重大意义。
随着机器学习的不断发展, 越来越多的机器学习方法被运用到故障诊断当中。逆变器故障诊断方法多基于对采集的敏感信号进行处理与分析。现有的故障诊断方法通常采用的模型有随机森林、BP 神经网络、支持向量机等; 文献[5]针对接触网供电系统牵引逆变器开路故障问题, 提出 1 种基于梯度信息最小二乘支持向量机的故障诊断新方法; 文献[6]为实现电力设备的实时故障检测, 应用 Park 变换获得三相电流基波幅度作为故障检测的特征变量,使用$\mathrm{{BP}}$ 神经网络结合简单的逻辑分析来定位故障开关设备;文献[7]对级联$\mathrm{H}$ 桥多电平逆变器的输出电压进行分解, 构建故障特征, 再通过主成分分析对故障成分进行维数约简, 最后采用参数优调后的随机森林进行故障诊断。上述模型研究的前提均立足于充足且均衡的样本。然而, 功率逆变器在实际工况运行中故障发生时间较短, 可用于诊断的参数少且易受负载变化的影响, 其故障样本数量较少, 因此在现场能够获得的正常状态的样本和故障样本的数量相差较大, 且不同类型的故障样本差距也较大, 样本存在不均衡问题。
用来解决样本不均衡问题的传统方法可以分为样本扩充和改良诊断模型两类。文献[8]针对大多数现实世界的数据集均存在数据样本在类中分布不均衡的问题, 提出 1 种新的基于聚类的欠采样方法来创建平衡数据集, 通过实验证明了所提方法的优越性; 文献[9]针对光伏阵列出现的组件阴影遮挡、短路与断路等故障, 提出 1 种基于快速过采样主成分分析算法的光伏阵列故障诊断方法来实现故障检测与故障识别; 文献[10]通过合成少数类过采样技术 SMOTE(synthetic minority over-sampling technique), 使得电力变压器故障样本均衡化, 并结合机器学习进行故障诊断。综上, 数据的下采样因除去了一部分样本, 导致了信息的丢失; 过采样通过样本的复制操作,易产生过拟合的现象[11]。样本扩充方法通过生成新样本来扩充小样本空间, 能有效避免采样方法存在的问题, 但是传统的样本扩充样本多运用 SMOTE 及其他改良模型[12],存在一定的局限性。一是在近邻选择时,存在一定的盲目性; 二是该算法无法克服非平衡数据集的数据分布, 容易产生分布边缘化问题。而改良诊断模型一般从自身出发,通过调整分类器敏感度来提高诊断精度, 但是提升的空间较小且难以获得最优权重。 尽管上述方法对不平衡故障分类任务进行了改进, 但难以生成高质量的数据样本。
2014 年, 生成对抗网络 GAN(generative adversarial network)由 Goodfellow I J 等提出[13]。随着对抗生成思想被不断完善, GAN 在图像生成、图像识别和故障诊断等领域有了较多应用[14]。GAN 能够在不依靠任何假设条件下通过无监督学习获得原始数据间的潜在特征规律并生成新的样本, 因此可用于解决三相逆变器开路故障诊断中的样本不平衡问题。基于此, 本文首先通过构建条件生成对抗网络 CGAN( conditional generative adversarial network)模型对原始逆变器故障样本进行数据增广, 达到平衡故障样本的目的; 再将增广后的故障样本输入至构建好的 CNN 进行训练, 最后通过 CNN 对样本的特征提取来获得对样本的分类精度, 并将其用于三相逆变器开路故障的诊断。
GAN 由生成器和判别器构成, 生成器用来混淆判别器的判定, 判别器用来将生成器生成的样本与原始样本进行区分, 两者在训练中相互博弈竞争,最终达到纳什平衡,获得符合真实数据分布的新样本, 具体原理结构如图1所示。
将生成器表示为$G$,判别器表示为$D$,该网络模型的训练目标函数可以表示为
$\mathop{\min }\limits_{G}\mathop{\max }\limits_{D}V\left({G, D}\right)= {E}_{x \sim {P}_{\mathrm{r}}}\left\lbrack {\ln D\left( x\right)}\right\rbrack +\\{E}_{z \sim {P}_{z}}\{\ln \{ 1 - D\left\lbrack {G\left( z\right)}\right\rbrack \}\}$
式中:$x$ 为真实样本;${P}_{\mathrm{r}}$ 为实际分布规律;$D\left( x\right)$ 为判别模型;$E$ 为期望;$z$ 为噪声;${P}_{\mathrm{z}}$ 为人工样本的分布规律;$G\left( z\right)$ 为生成函数。
利用式(1)训练$G$ 时,由于$G$ 训练仅与生成样本有关联, 即最小化生成样本被判定为生成样本的概率。该目标函数为
$\mathop{\min }\limits_{G}{L}_{G}= {E}_{z \sim {P}_{z}}\{\ln \{ 1 - D\left\lbrack {G\left( z\right)}\right\rbrack \}\}$
利用(1)式训练$D$ 时,根据输入的真实样本或者生成样本,通过$D$ 来最大化该样本被判定为该样本本身的概率, 即真实样本被判定为真实样本, 生成样本被判定为生成样本的概率。该目标函数为
$\mathop{\max }\limits_{D}{L}_{D}= {E}_{x \sim {P}_{\mathrm{r}}}\left\lbrack {\ln D\left( x\right)}\right\rbrack +{E}_{z \sim {P}_{\mathrm{z}}}\left\{{\ln \{ 1 - D\left\lbrack {G\left( z\right)}\right\rbrack \}}\right\}$
生成器和判别器相互对抗,进行不断迭代训练, 理论上可达到两者之间的纳什均衡点, 即生成器可生成判别器难以鉴别的新样本。
最基本的 GAN 存在一定的缺陷, 如训练不稳定、模式崩溃等。另外, 过于自由的生成模式使得训练过程和结果均不可控。为了解决上述问题, 基于 GAN 的衍生模型被不断提出, 其中包括文献[15] 所提出的 CGAN。CGAN 的原理机构如图2所示, CGAN 在生成器和判别器的输入中加入了条件信息, 而条件信息定义较广泛, 可以带有该输入特征的标签。CGAN 的提出使得 GAN 可以利用输入特征及该特征对应的标签进行训练, 并根据标签生成更加符合真实样本规律的生成样本。
在 GAN 的基础上, CGAN 增加的其他信息与$x\text{、}z$ 相结合,其目标函数可表示为
$\mathop{\min }\limits_{G}\mathop{\max }\limits_{D}V\left({G, D}\right)= {E}_{x \sim {P}_{\mathrm{r}}}\left\lbrack {\ln D\left({x \mid c}\right)}\right\rbrack +\\{E}_{z \sim {P}_{z}}\{\ln \{ 1 - D\left\lbrack {G\left({z \mid c}\right)\mid c}\right\rbrack \}\}$
式中,$c$ 为标签信息。本文中该信息为正常状态和各类型开关管开路故障的标签,分别令其为$c = 0$$c = 1\text{、}c = 2$$c = 3$ 等 7 种标签。
CNN 原理结构如图3所示, 其中该网络包含输入层、卷积层、池化层、全连接层和输出层。卷积层包含不同的卷积核, 隐式提取样本特征; 池化层对其特征进行降采样处理; 全连接层将降采样处理过后的特征进行综合,最后输出结果。
输入层到卷积层的函数可表示为
${\alpha }^{l}= \sigma \left({{\alpha }^{l - 1}* {W}^{l}+ {b}^{l}}\right)$
式中:$\sigma \left(\cdot \right)$ 为激活函数,采用 Relu 函数;$l$ 为网络层数;${\alpha }^{l}\text{、}{\alpha }^{l - 1}$ 分别为卷积层的输出和输入;$W$ 为权重;$b$ 为偏置;“*”为卷积符号。
池化层的输出函数可以表示为
${\alpha }^{l}= {p}_{\text{pool }}\left({\alpha }^{l - 1}\right)$
式中,${p}_{\mathrm{{pool}}}\left(\cdot \right)$ 为池化层输出。
全连接层的结构为 DNN 模型结构, 其输出可表示为
${\alpha }^{l}= \sigma \left({{\alpha }^{l - 1}{W}^{l}+ {b}^{l}}\right)$
输出层的函数可表示为
${\alpha }^{l}= \operatorname{Softmax}\left({{\alpha }^{l - 1}{W}^{l}+ {b}^{l}}\right)$
本文利用 CNN 的卷积层对三相电流信号进行特征提取, 利用池化层将这些提取的特征进行特征降维, 再通过全连接层将降维后的特征进行综合, 最后使用 Softmax 分类器作为输出层的激活函数,对样本进行分类处理。
在电机驱动系统[16] 中,功率开关管为${\mathrm{T}}_{1}\sim {\mathrm{T}}_{6}$, 续流二极管为${\mathrm{D}}_{1}\sim {\mathrm{D}}_{6}$,直流母线电压${V}_{\mathrm{{in}}}= {48}\mathrm{\;V}$。 根据三相绕组电流${i}_{\mathrm{a}}\text{、}{i}_{\mathrm{b}}\text{、}{i}_{\mathrm{c}}$ 获得故障数据。本文仅讨论 6 个功率开关管单独开路的状态, 包含正常工作状态共 7 种故障模式。三相全桥逆变器结构故障模式如表1所示。
本文所采用的 CGAN 模型包括 1 个生成器和 1 个判别器。判别器采用 1 个输入层、 2 个卷积层、 2 个池化层、 2 个全连接层和 1 个输出层, 而生成器采用 1 个输入层、 3 个卷积层、 1 个全连接层和 1 个输出层, 在卷积后利用 BatchNorm2d 进行归一化处理, 用来稳定网络。生成器和判别器的优化器均为 Adam。CGAN 模型结构如表2所示。
由于输入信号为$1 \times {1500}$,为了更好地提取样本特征,本文采用一维 CNN 结构。本文 CNN 借鉴了经典 AlexNet 网络结构思想, 即卷积核数量随着卷积层的深入而变多, 由于输入信号维度大于二维图像某一维度,因此常规的二维卷积核大小$3 \times 3$$5 \times 5\text{、}7 \times 7$ 不适用于该网络结构。经过多次实验,本文所涉及的 CNN 网络结构包含 1 个输入层、 2 个卷积层、2 个池化层、1 个全连接层和 Softmax 分类器。各卷积层分别包括 8 个$3 \times 1$ 的卷积核和 16 个$3 \times 1$ 的卷积核。各池化层分别采用$2 \times 1$$2 \times 1$ 的小窗口进行特征降维。Softmax 分类器输出 7 种识别概率向量。CNN 网络模型参数及结构分别如表3图4所示。
CGAN-CNN 故障诊断流程如图5所示, 具体步骤如下:
步骤 1 采集 DC-AC 驱动电路 A、B、C 三相绕组电流信号${i}_{\mathrm{a}}$${i}_{\mathrm{b}}$${i}_{\mathrm{c}}$ 作为故障敏感信号。
步骤 2 将采集到的敏感信号进行快速傅里叶变换, 得到故障样本的频率特性, 并截取适当的频率信号作为特征向量, 再对特征向量进行归一化处理得到故障数据样本。
步骤 3 CGAN 生成样本数据。将真实样本和噪声样本添加标签后输入 CGAN 模型, 确定生成器参数和判别器参数, 再开始通过博弈竞争机制不断迭代优化生成器和判别器模型的性能参数, 达到纳什平衡,最后根据原始数据生成新的数据。
步骤 4 CNN 故障诊断。构建 CNN 网络模型并初始化网络参数, 输入划分好的样本数据训练 CNN,通过不断训练调整网络参数与权重,最后输出故障分类诊断结果。
为验证本文所提方法的有效性,搭建 BLDCM 驱动系统故障检测实验平台, 如图6所示。该平台由直流电源、数据采集器、DC-AC 驱动电路、BLDCM 与风扇负载构成。BLDCM 额定转速为${800}\mathrm{r}/\mathrm{{min}}$,实验通过不断变化风扇的转速使得电机转速大概控制为${550}\text{、}{650}\text{、}{750}\mathrm{r}/\mathrm{{min}},3$ 种负载工况分别记为${\text{load}}_{1}$${\text{load}}_{2}$${\text{load}}_{3}$。分别对不同负载情形下 7 种故障模式三相电流${i}_{\mathrm{a}}\text{、}{i}_{\mathrm{b}}\text{、}{i}_{\mathrm{c}}$ 采样各 100 次,数据采样器的频率为${200}\mathrm{{kHz}}$,每次采样的采样点数目为$1 \times {10}^{5}$ 个。
图7为在${\mathrm{{FC}}}_{0}$${\mathrm{{FC}}}_{4}$${\mathrm{{FC}}}_{5}$${\mathrm{{FC}}}_{6}$ 这 4 种不同状态下${i}_{\mathrm{a}}$ 的波形。电机在${\operatorname{load}}_{1}$ 工况下采集的时域信号如图7(a)所示, 经过 FFT 处理的频率信号如图7(b)所示。
根据 2.1 节建立 CGAN 模型并初始化网络参数, 考虑到 loss 函数的收敛性及运行时间, 训练过程中设置 epoch 为 100 次。利用 CGAN 生成的数据为原始数据的 1 倍,即在${\operatorname{load}}_{1}$${\operatorname{load}}_{2}$${\operatorname{load}}_{3}$ 工况和 7 种状态下各 100 个样本。为了清晰地展示 CGAN 生成数据的能力,利用 t-SNE 对 load${}_{2}$ 工况生成的数据进行聚类可视化, 如图8所示。
图8清晰地展示了 CGAN 新生成样本的分布情况, 利用 t-SNE 对模型所提取的特征进行降维可视化, 在二维空间展示生成新样本, 由此可见 CGAN 模型较好地学习了 7 类带有标签样本的特征。
在正常工况下, 逆变器故障样本和正常状态下的样本相比, 会出现故障样本较少的情况, 导致样本不均衡。本文实验构造了样本不均衡情况:样本数据为在${\operatorname{load}}_{1}$ 工况下正常状态、${\mathrm{T}}_{1}$ 管开路、${\mathrm{T}}_{2}$ 管开路、${\mathrm{T}}_{3}$ 管开路、${\mathrm{T}}_{4}$ 管开路、${\mathrm{T}}_{5}$ 管开路和${\mathrm{T}}_{6}$ 管开路各 100 个样本,样本集${A}_{1}$ 为均衡样本即每种状态样本各 100 个; 再将${\mathrm{T}}_{6}$ 管开路故障样本分别减少至 50 个和 30 个,分别令其为样本集${A}_{2}$ 和样本集${A}_{3}$; 并将 CGAN、GAN 和 SMOTE 生成的${\mathrm{T}}_{6}$ 管开路故障样本对样本集${A}_{3}$ 进行填充,填充数量为 70 个,填充后为样本集${A}_{4}$。分别比较 CGAN、GAN 和 SMOTE 这 3 种方法填充后样本集${A}_{4}$ 的诊断精度, 以此验证不同数据增强算法在样本不均衡情景下提高逆变器故障诊断精度的有效性。每组进行 20 次有效诊断且在${\mathrm{{load}}}_{1}$ 情形中进行,各类型样本不平衡比例如表5所示。
将 4 种样本均分为训练集 80%,测试集 20%, 并输入至构建好的 CNN 诊断模型中进行分类识别故障诊断。3 种样本集${A}_{1}\text{、}{A}_{2}\text{、}{A}_{3}$ 的混淆矩阵如图9所示。3 种样本集输入方式下的 CNN 模型分类准确率如图10所示。
图9图10可知, 不平衡度越高的样本集, 故障诊断的准确率越低。根据不同平衡度样本的混淆矩阵可知, 类 7 故障样本大都被错分为类 2 的故障样本,由此可以得出${\mathrm{T}}_{6}$ 管开路故障的特征与${\mathrm{T}}_{1}$ 管开路故障的特征相似,造成了错误识别。
样本集${A}_{4}$ 通过 SMOTE、GAN 和 CGAN 这 3 种方法填充后, 数据的诊断精度混淆矩阵如图11所示, 样本集${A}_{4}$ 输入方式下的 SMOTE-CNN、GAN-CNN 和 CGAN-CNN 的分类准确率如图12所示。
图11图12可知: 利用 CGAN、GAN 和 SMOTE 填充样本集${A}_{4}$ 中的${\mathrm{T}}_{6}$ 管开路故障样本时, 故障诊断正确率有明显提升;利用 SMOTE 及 GAN 进行填充后的样本集${A}_{4}$ 的诊断精度达到了 95%以上,与填充前的相比, 被误判为类 2 的错误分类减少了很多; 而用 CGAN 进行填充的样本集${A}_{4}$ 的诊断精度接近 99%。说明 CGAN 在原始数据的模拟学习过程中, 生成的新样本特征与原始样本更加接近, 生成的样本质量更高,提高了故障诊断的准确率。
综上所述, 本文所提基于 CGAN-CNN 的逆变器开路故障诊断模型能够较好地识别逆变器的 7 种不同状态模式, 说明本文方法可以有效解决不均衡条件下逆变器的故障诊断问题。
本文采集三相逆变器电流信号在不同工况下的原始数据, 对原始数据进行 FFT 和归一化处理, 通过 CGAN 对原始数据进行增强, 将增加后的样本数据输入 CNN 进行故障识别, 比较不同样本集下的故障诊断结果, 证明本文方法的有效性和准确性。与其他方法相比,本文所提基于 CGAN-CNN 三相逆变器故障诊断方法的优势包括:
(1)在样本集不平衡的情况下,引入 CGAN 对原始数据进行增强, 将数据进行扩充, 再从中选取某一类别故障通过实验证明生成样本的有效性;
(2)引入 CNN 网络模型,使用双卷积层、双池化层,卷积核大小分别为$3 \times 1$$2 \times 1$,提取特征更加全面;
(3)在 CGAN 中,通过设置类别标签$c$,将纯无监督 GAN 变成有监督模型, 并通过对生成器设置梯度清零, 再反向传播进行迭代, 解决了 GAN 存在的生成器损失函数饱和及训练不稳定的问题。
  • 国家自然科学基金资助项目(61901212)
  • 江苏省高等学校自然科学研究重大资助项目(20KJA510007)
  • 江苏省配电网智能技术与装备协同创新中心开放基金资助项目(XTCX201909)
参考文献 引证文献
排序方式:
[1]
Vijayalakshmi S, Hubert Tony Raj L, Palaniyappan S, et al. A review on multilevel H-bridge cascaded inductor less hybrid inverter for Electric vehicles with PWM control[J]. Materials Today: Proceedings, 2021. 45: 1644-1650.
[2]
张建忠, 耿治, 徐帅. 基于T 型逆变器的 APF 故障诊断与容错控制[J]. 中国电机工程学报, 2019. 39(1): 245-255, 339.
Zhang Jianzhong, Geng Zhi, Xu Shuai. Fault diagnosis and fault tolerant control for APF based on T-type inverter[J]. Proceedings of the CSEE, 2019. 39(1): 245-255, 339 (in Chinese).
[3]
Caseiro LMA, Mendes AMS. Real-time IGBT open-circuit fault diagnosis in three-level neutral-point-clamped voltage-source rectifiers based on instant voltage error[J]. IEEE Transactions on Industrial Electronics, 2015. 62(3): 1669-1678.
[4]
陈勇, 刘志龙, 陈章勇. 基于电流矢量特征分析的逆变器开路故障快速诊断与定位方法[J]. 电工技术学报, 2018. 33(4): 883-891.
Chen Yong, Liu Zhilong, Chen Zhangyong. Fast diagnosis and location method for open-circuit fault in inverter based on current vector character analysis[J]. Transactions of China Electrotechnical Society, 2018. 33(4): 883-891 (in Chinese).
[5]
Cheng Yongqi, Dong Weiguang, Gao Fengyang, et al. Open-circuit fault diagnosis of traction inverter based on compressed sensing theory[J]. Chinese Journal of Electrical Engineering, 2020. 6(1): 52-60.
[6]
Ji Zhendong, Liu Wei. Open-circuit fault detection for three-phase inverter based on backpropagation neural network[J]. Neural Computing and Applications, 2019. 31(9): 4665-4674.
[7]
陈石, 张兴敢. 基于小波包能量熵和随机森林的级联H 桥多电平逆变器故障诊断[J]. 南京大学学报(自然科学), 2020. 56(2): 284-289.
Chen Shi, Zhang Xinggan. Fault diagnosis for cascaded H-bridge multilevel inverter based on wavelet packet energy entropy and random forest[J]. Journal of Nanjing University (Natural Science), 2020. 56(2): 284-289 (in Chinese).
[8]
Ebrahimi Shahabadi MS, Tabrizchi H, Kuchaki Rafsanjani M, et al. A combination of clustering-based under-sampling with ensemble methods for solving imbalanced class problem in intelligent systems[J]. Technological Forecasting and Social Change, 2021. 169: 120796.
[9]
李元良, 丁坤, 陈富东, 等. 基于快速过采样主成分分析法的光伏阵列故障诊断[J]. 电网技术, 2019. 43(1): 308-315.
Li Yuanliang, Ding Kun, Chen Fudong, et al. Fault diagnosis method of PV array based on fast OS-PCA[J]. Power System Technology, 2019. 43(1): 308-315 (in Chinese).
[10]
刘云鹏, 和家慧, 许自强, 等. 基于SVM SMOTE 的电力变压器故障样本均衡化方法[J]. 高电压技术, 2020. 46(7): 2522-2529.
Liu Yunpeng, He Jiahui, Xu Ziqiang, et al. Equalization method of power transformer fault sample based on SVM SMOTE[J]. High Voltage Engineering, 2020. 46(7): 2522-2529 (in Chinese).
[11]
Pattanadech N, Yutthagowith P. Fast curve fitting algorithm for parameter evaluation in lightning impulse test technique[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2015. 22(5): 2931-2936.
[12]
王圆方. 基于层次聚类改进 SMOTE 的过采样方法[J]. 软件, 2020. 41(2): 201-204.
Wang Yuanfang. Hybrid algorithm of aggregation hierarchy clustering and SMOTE for oversampling[J]. Computer Engineering & Software, 2020. 41(2): 201-204 (in Chinese).
[13]
Goodfellow I J, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets[C]// Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal, Quebec, Canada: MIT Press, 2014: 2672-2680.
[14]
Wang Junliang, Yang Zhengliang, Zhang Jie, et al. AdaBalGAN: An improved generative adversarial network with imbalanced learning for wafer defective pattern recognition[J]. IEEE Transactions on Semiconductor Manufacturing, 2019. 32(3): 310-319.
[15]
Mirza M,Osindero S. Conditional generative adversarial nets[EB/OL]. (2014-11-06) [2020-06-10]. https://arxiv.org/abs/1411.1784.
[16]
Sun Quan, Wang Youren, Jiang Yuanyuan. A novel fault diagnostic approach for DC-DC converters based on CSA-DBN[J]. IEEE Access, 2017. 6: 6273-6285.
2024年第22卷第6期
PDF下载
397
162
引用本文
BibTeX
文章信息
doi: 10.13234/j.issn.2095-2805.2024.6.318
  • 接收时间:2021-12-24
  • 首发时间:2025-07-19
  • 出版时间:2024-11-30
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2021-12-24
  • 修回日期:2022-03-13
  • 录用日期:2022-04-06
基金
National Natural Science Foundation of China(61901212)
国家自然科学基金资助项目(61901212)
Natural Science Foundation of Jiangsu Higher Education Institutions of China(20KJA510007)
江苏省高等学校自然科学研究重大资助项目(20KJA510007)
Open Research Fund of Jiangsu Collaborative Innovation Center for Smart Distribution Network(XTCX201909)
江苏省配电网智能技术与装备协同创新中心开放基金资助项目(XTCX201909)
作者信息
    1 南京工程学院 自动化学院 南京 211167
    2 南京航空航天大学 自动化学院 南京 211106
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/dyxb/CN/10.13234/j.issn.2095-2805.2024.6.318
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏