Article(id=1149776902736536290, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1149776900194791454, articleNumber=null, orderNo=null, doi=10.12404/j.issn.1671-1815.2403506, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1715443200000, receivedDateStr=2024-05-12, revisedDate=1723219200000, revisedDateStr=2024-08-10, acceptedDate=null, acceptedDateStr=null, onlineDate=1752057775434, onlineDateStr=2025-07-09, pubDate=1744905600000, pubDateStr=2025-04-18, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1752057775434, onlineIssueDateStr=2025-07-09, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1752057775434, creator=13701087609, updateTime=1752057775434, updator=13701087609, issue=Issue{id=1149776900194791454, tenantId=1146029695717560320, journalId=1146123166801305609, year='2025', volume='25', issue='11', pageStart='4397', pageEnd='4826', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1752057774827, creator=13701087609, updateTime=1768456666677, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1218558837930512931, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1149776900194791454, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1218558837930512932, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1149776900194791454, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=4543, endPage=4550, ext={EN=ArticleExt(id=1149776903088857828, articleId=1149776902736536290, tenantId=1146029695717560320, journalId=1146123166801305609, language=EN, title=Fault Diagnosis Method of Centrifugal Pump Based on Siamese Networks under Small Sample Conditions, columnId=1156262732765717457, journalTitle=Science Technology and Engineering, columnName=Papers·Mechanical and Instrumental Industry, runingTitle=null, highlight=null, articleAbstract=

Conventional diagnostic methods that require a large amount of data support in practical engineering are difficult to effectively perform centrifugal pump fault diagnosis under small sample conditions. Therefore, the residual network (ResNet) in deep learning was combined with dilated convolution and extended into a siamese network to construct a dilated residual siamese network (DRSN). The dilated residual network was used as the feature extraction module of the siamese network, which enhanced the feature extraction ability of the model. Positive and negative sample pairs were constructed to extract more information from each sample, and make more effective use of limited data.The two sub-networks share parameters, the number of free parameters and lowering the risk of overfitting was reduced when the sample was insufficient. The proposed network model alleviated the problem of insufficient training samples, improved the efficiency of data utilization, and realized the fault classification of centrifugal pump under the condition of small samples. The research results show that even in the most sample-scarce situation, the accuracy of the model on the centrifugal pump test dataset can still reach 82.20%, which is at least 8.8 percentage points higher than other models.

, correspAuthors=Li-xiang DUAN, 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=Ke LI, Lai-bin ZHANG, Li-xiang DUAN, Hai-peng LIU, Xin-yue ZHANG), CN=ArticleExt(id=1149776937658311395, articleId=1149776902736536290, tenantId=1146029695717560320, journalId=1146123166801305609, language=CN, title=小样本下离心泵的孪生网络故障诊断方法, columnId=1156262732954461139, journalTitle=科学技术与工程, columnName=论文·机械、仪表工业, runingTitle=null, highlight=null, articleAbstract=

实际工程中需要大量数据支撑的常规诊断方法难以有效进行小样本条件下的离心泵故障诊断,为此,将深度学习中的残差网络(residual network,ResNet)与膨胀卷积相结合,并拓展为孪生网络,构建膨胀残差孪生网络(dilated residual siamese network,DRSN)。将膨胀残差网络作为孪生网络的特征提取模块,强化了模型的特征提取能力;构造正负样本对,从每个样本中提取更多的信息,更有效地利用有限的数据;两个子网络共享参数,减少自由参数的数量,降低样本不足时过拟合的风险。提出的网络模型缓解了训练样本不足的问题,提升了数据利用的效率,实现了小样本条件下的离心泵故障分类。研究结果表明:在样本最匮乏的情况下,该模型在离心泵试验数据集上的准确率仍能达到82.20%,相较其他模型,准确率至少提升了8.8个百分点。

, correspAuthors=段礼祥, authorNote=null, correspAuthorsNote=
* 段礼祥(1969—),男,汉族,四川泸州人,博士,教授。研究方向:安全监测与智能诊断工程、机械设备状态监测与故障诊断。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=ijaPKlFoSsklxRzpllNsPQ==, magXml=2kJ/3oH38weMRpV4upjg0Q==, pdfUrl=null, pdf=PlE5ddP+zWwJ1O3infLbtQ==, pdfFileSize=7157871, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=pqeJjaqc8K+sdDqwjj3JIA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=uJlhJnxAE6DcYuO8w7Givg==, mapNumber=null, authorCompany=null, fund=null, authors=

李科(1999—),男,汉族,安徽六安人,硕士研究生。研究方向:安全监测与智能诊断。E-mail:

, authorsList=李科, 张来斌, 段礼祥, 刘海鹏, 张馨月)}, authors=[Author(id=1218843909648729085, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=693496009@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1218843909803917327, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843909648729085, language=EN, stringName=Ke LI, firstName=Ke, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China
2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1218843909963300899, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843909648729085, language=CN, stringName=李科, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249
2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249, bio={"content":"

李科(1999—),男,汉族,安徽六安人,硕士研究生。研究方向:安全监测与智能诊断。E-mail:

"}, bioImg=null, bioContent=

李科(1999—),男,汉族,安徽六安人,硕士研究生。研究方向:安全监测与智能诊断。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1218843909204132815, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=1, ext=[AuthorCompanyExt(id=1218843909229298643, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China), AuthorCompanyExt(id=1218843909241881557, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249)]), AuthorCompany(id=1218843909380293602, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=2, ext=[AuthorCompanyExt(id=1218843909384487908, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China), AuthorCompanyExt(id=1218843909392876518, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249)])]), Author(id=1218843910181404721, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1218843910324011074, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843910181404721, language=EN, stringName=Lai-bin ZHANG, firstName=Lai-bin, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China
2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1218843910462423120, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843910181404721, language=CN, stringName=张来斌, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249
2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1218843909204132815, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=1, ext=[AuthorCompanyExt(id=1218843909229298643, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China), AuthorCompanyExt(id=1218843909241881557, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249)]), AuthorCompany(id=1218843909380293602, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=2, ext=[AuthorCompanyExt(id=1218843909384487908, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China), AuthorCompanyExt(id=1218843909392876518, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249)])]), Author(id=1218843910634389602, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=duanlx@cup.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1218843910776995953, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843910634389602, language=EN, stringName=Li-xiang DUAN, firstName=Li-xiang, middleName=null, lastName=DUAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, *, address=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China
2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1218843910919602309, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843910634389602, language=CN, stringName=段礼祥, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, *, address=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249
2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1218843909204132815, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=1, ext=[AuthorCompanyExt(id=1218843909229298643, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China), AuthorCompanyExt(id=1218843909241881557, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249)]), AuthorCompany(id=1218843909380293602, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=2, ext=[AuthorCompanyExt(id=1218843909384487908, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China), AuthorCompanyExt(id=1218843909392876518, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249)])]), Author(id=1218843911070597269, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1218843911200620708, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843911070597269, language=EN, stringName=Hai-peng LIU, firstName=Hai-peng, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3 China National Petroleum International Pipeline Co., Ltd., Beijing 102206, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1218843911318061236, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843911070597269, language=CN, stringName=刘海鹏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3 中油国际管道有限公司, 北京 102206, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1218843909501928430, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=3, ext=[AuthorCompanyExt(id=1218843909510317040, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909501928430, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 China National Petroleum International Pipeline Co., Ltd., Beijing 102206, China), AuthorCompanyExt(id=1218843909518705650, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909501928430, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 中油国际管道有限公司, 北京 102206)])]), Author(id=1218843911431307459, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1218843911578108113, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843911431307459, language=EN, stringName=Xin-yue ZHANG, firstName=Xin-yue, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China
2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1218843911720714464, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, authorId=1218843911431307459, language=CN, stringName=张馨月, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249
2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1218843909204132815, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=1, ext=[AuthorCompanyExt(id=1218843909229298643, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China), AuthorCompanyExt(id=1218843909241881557, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249)]), AuthorCompany(id=1218843909380293602, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=2, ext=[AuthorCompanyExt(id=1218843909384487908, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China), AuthorCompanyExt(id=1218843909392876518, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249)])])], keywords=[Keyword(id=1218843912014315773, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, orderNo=1, keyword=centrifugal pump), Keyword(id=1218843912102396167, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, orderNo=2, keyword=fault diagnosis), Keyword(id=1218843912224030997, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, orderNo=3, keyword=small sample), Keyword(id=1218843912358248738, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, orderNo=4, keyword=residual network), Keyword(id=1218843912458912050, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, orderNo=5, keyword=siamese network), Keyword(id=1218843912580546878, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, orderNo=1, keyword=离心泵), Keyword(id=1218843912697987405, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, orderNo=2, keyword=故障诊断), Keyword(id=1218843912823816537, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, orderNo=3, keyword=小样本), Keyword(id=1218843912941257059, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, orderNo=4, keyword=残差网络), Keyword(id=1218843913100640622, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, orderNo=5, keyword=孪生网络)], refs=[Reference(id=1218843918251246322, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2021, volume=40, issue=10, pageStart=41, pageEnd=49, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=焦瀚晖, 胡明辉, 江志农, journalName=振动与冲击, refType=null, unstructuredReference=焦瀚晖, 胡明辉, 江志农, 等. 基于补偿距离评估和一维卷积神经网络的离心泵故障快速智能识别方法[J]. 振动与冲击, 2021, 40(10): 41-49., articleTitle=基于补偿距离评估和一维卷积神经网络的离心泵故障快速智能识别方法, refAbstract=null), Reference(id=1218843918360298232, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2021, volume=40, issue=10, pageStart=41, pageEnd=49, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=Jiao Hanhui, Hu Minghui, Jiang Zhinong, journalName=Journal of Vibration and Shock, refType=null, unstructuredReference=Jiao Hanhui, Hu Minghui, Jiang Zhinong, et al. Rapid intelligent fault identification method for centrifugal pumps based on compensation distance evaluation and one-dimensional convolutional neural networks[J]. Journal of Vibration and Shock, 2021, 40(10): 41-49., articleTitle=Rapid intelligent fault identification method for centrifugal pumps based on compensation distance evaluation and one-dimensional convolutional neural networks, refAbstract=null), Reference(id=1218843918628733703, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=孙弋, journalName=基于深度学习的小样本条件下轴承故障诊断技术研究, refType=null, unstructuredReference=孙弋. 基于深度学习的小样本条件下轴承故障诊断技术研究[D]. 成都: 西南交通大学, 2021., articleTitle=null, refAbstract=null), Reference(id=1218843918754562830, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=Sun Yi, journalName=Research on bearing fault diagnosis technology under small sample conditions based on deep learning, refType=null, unstructuredReference=Sun Yi. Research on bearing fault diagnosis technology under small sample conditions based on deep learning[D]. Chengdu: Southwest Jiaotong University, 2021., articleTitle=null, refAbstract=null), Reference(id=1218843918880391958, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2006, volume=28, issue=4, pageStart=594, pageEnd=611, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=Li F F, Fergus R, Perona P, journalName=IEEE Transactions on Pattern Analysis and Machine Intelligence, refType=null, unstructuredReference=Li F F, Fergus R, Perona P, et al. One-shot learning of object categories[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(4): 594-611., articleTitle=One-shot learning of object categories, refAbstract=null), Reference(id=1218843919006221085, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=59, issue=12, pageStart=215, pageEnd=224, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=邵海东, 李伟, 刘翊, journalName=机械工程学报, refType=null, unstructuredReference=邵海东, 李伟, 刘翊, 等. 时变转速下基于双阈值注意力生成对抗网络和小样本的转子-轴承系统故障诊断[J]. 机械工程学报, 2023, 59(12): 215-224., articleTitle=时变转速下基于双阈值注意力生成对抗网络和小样本的转子-轴承系统故障诊断, refAbstract=null), Reference(id=1218843919148827425, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=59, issue=12, pageStart=215, pageEnd=224, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=Shao Haidong, Li Wei, Liu Yi, journalName=Journal of Mechanical Engineering, refType=null, unstructuredReference=Shao Haidong, Li Wei, Liu Yi, et al. Fault diagnosis of rotor-bearing system under time-varying speeds by using dual-threshold attention-embedded GAN and small samples[J]. Journal of Mechanical Engineering, 2023, 59(12): 215-224., articleTitle=Fault diagnosis of rotor-bearing system under time-varying speeds by using dual-threshold attention-embedded GAN and small samples, refAbstract=null), Reference(id=1218843919253685032, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2022, volume=81, issue=16, pageStart=22379, pageEnd=22405, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=Dev K, Ashraf Z, Muhuri P K, journalName=Multimedia Tools and Applications, refType=null, unstructuredReference=Dev K, Ashraf Z, Muhuri P K, et al. Deep autoencoder based domain adaptation for transfer learning[J]. Multimedia Tools and Applications, 2022, 81(16): 22379-22405., articleTitle=Deep autoencoder based domain adaptation for transfer learning, refAbstract=null), Reference(id=1218843919379514157, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[6], rfOrder=8, authorNames=丁鹏, journalName=有限数据下滚动轴承变工况寿命预测方法及应用研究, refType=null, unstructuredReference=丁鹏. 有限数据下滚动轴承变工况寿命预测方法及应用研究[D]. 南京: 东南大学, 2022., articleTitle=null, refAbstract=null), Reference(id=1218843919488566066, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[6], rfOrder=9, authorNames=Ding Peng, journalName=Research on limited data-driven life prediction methods and applications of rolling bearings under variable operating condition, refType=null, unstructuredReference=Ding Peng. Research on limited data-driven life prediction methods and applications of rolling bearings under variable operating condition[D]. Nanjing: Southeast University, 2022., articleTitle=null, refAbstract=null), Reference(id=1218843919631172407, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2020, volume=20, issue=30, pageStart=12268, pageEnd=12278, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=张卡, 宿东, 王蓬勃, journalName=科学技术与工程, refType=null, unstructuredReference=张卡, 宿东, 王蓬勃, 等. 深度学习技术在影像密集匹配方面的进展与应用[J]. 科学技术与工程, 2020, 20(30): 12268-12278., articleTitle=深度学习技术在影像密集匹配方面的进展与应用, refAbstract=null), Reference(id=1218843919744418622, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2020, volume=20, issue=30, pageStart=12268, pageEnd=12278, url=null, language=null, rfNumber=[7], rfOrder=11, authorNames=Zhang Ka, Su Dong, Wang Pengbo, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Zhang Ka, Su Dong, Wang Pengbo, et al. Progress and application of deep learning in image dense matching[J]. Science Technology and Engineering, 2020, 20(30): 12268-12278., articleTitle=Progress and application of deep learning in image dense matching, refAbstract=null), Reference(id=1218843919866053444, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=10, pageStart=3939, pageEnd=3951, url=null, language=null, rfNumber=[8], rfOrder=12, authorNames=邬娜, 王健, 杨建伟, journalName=科学技术与工程, refType=null, unstructuredReference=邬娜, 王健, 杨建伟, 等. 变工况及小样本情况下滚动轴承故障迁移学习方法综述[J]. 科学技术与工程, 2024, 24(10): 3939-3951., articleTitle=变工况及小样本情况下滚动轴承故障迁移学习方法综述, refAbstract=null), Reference(id=1218843919954133832, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=10, pageStart=3939, pageEnd=3951, url=null, language=null, rfNumber=[8], rfOrder=13, authorNames=Wu Na, Wang Jian, Yang Jianwei, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Wu Na, Wang Jian, Yang Jianwei, et al. Review of rolling bearing migration learning under variable operating conditions and small samples[J]. Science Technology and Engineering, 2024, 24(10): 3939-3951., articleTitle=Review of rolling bearing migration learning under variable operating conditions and small samples, refAbstract=null), Reference(id=1218843920084157262, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2022, volume=51, issue=4, pageStart=116, pageEnd=119, url=null, language=null, rfNumber=[9], rfOrder=14, authorNames=钱心筠, 王友仁, 赵亚磊, journalName=机械制造与自动化, refType=null, unstructuredReference=钱心筠, 王友仁, 赵亚磊. 基于孪生网络的行星齿轮箱故障诊断方法[J]. 机械制造与自动化, 2022, 51(4): 116-119., articleTitle=基于孪生网络的行星齿轮箱故障诊断方法, refAbstract=null), Reference(id=1218843920205792084, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2022, volume=51, issue=4, pageStart=116, pageEnd=119, url=null, language=null, rfNumber=[9], rfOrder=15, authorNames=Qian Xinyun, Wang Youren, Zhao Yalei, journalName=Machine Building & Automation, refType=null, unstructuredReference=Qian Xinyun, Wang Youren, Zhao Yalei. Method of planetary gearbox fault diagnosis based on siamese networks[J]. Machine Building & Automation, 2022, 51(4): 116-119., articleTitle=Method of planetary gearbox fault diagnosis based on siamese networks, refAbstract=null), Reference(id=1218843920319038297, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=16, authorNames=张青雨, journalName=基于深度胶囊网络的旋转机械故障智能诊断方法研究, refType=null, unstructuredReference=张青雨. 基于深度胶囊网络的旋转机械故障智能诊断方法研究[D]. 秦皇岛: 燕山大学, 2023., articleTitle=null, refAbstract=null), Reference(id=1218843920436478816, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=17, authorNames=Zhang Qingyu, journalName=Research on mechanical fault intelligent diagnosis based on deep capsule network, refType=null, unstructuredReference=Zhang Qingyu. Research on mechanical fault intelligent diagnosis based on deep capsule network[D]. Qihuangdao: Yanshan University, 2023., articleTitle=null, refAbstract=null), Reference(id=1218843920558113637, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=59, issue=19, pageStart=294, pageEnd=304, url=null, language=null, rfNumber=[11], rfOrder=18, authorNames=赵晓平, 彭澎, 张永宏, journalName=计算机工程与应用, refType=null, unstructuredReference=赵晓平, 彭澎, 张永宏, 等. 改进孪生网络在小样本轴承故障诊断中的应用[J]. 计算机工程与应用, 2023, 59(19): 294-304., articleTitle=改进孪生网络在小样本轴承故障诊断中的应用, refAbstract=null), Reference(id=1218843920704914289, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=59, issue=19, pageStart=294, pageEnd=304, url=null, language=null, rfNumber=[11], rfOrder=19, authorNames=Zhao Xiaoping, Peng Peng, Zhang Yonghong, journalName=Computer Engineering and Applications, refType=null, unstructuredReference=Zhao Xiaoping, Peng Peng, Zhang Yonghong, et al. Application of improved siamese neural network in small sample fault diagnosis of bearing[J]. Computer Engineering and Applications, 2023, 59(19): 294-304., articleTitle=Application of improved siamese neural network in small sample fault diagnosis of bearing, refAbstract=null), Reference(id=1218843920851714936, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=40, issue=9, pageStart=2851, pageEnd=2855, url=null, language=null, rfNumber=[12], rfOrder=20, authorNames=孙统风, 王康, 郝徐, journalName=计算机应用研究, refType=null, unstructuredReference=孙统风, 王康, 郝徐. 面向小样本学习的双重度量孪生神经网络[J]. 计算机应用研究, 2023, 40(9): 2851-2855., articleTitle=面向小样本学习的双重度量孪生神经网络, refAbstract=null), Reference(id=1218843920956572541, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2023, volume=40, issue=9, pageStart=2851, pageEnd=2855, url=null, language=null, rfNumber=[12], rfOrder=21, authorNames=Sun Tongfeng, Wang Kang, Hao Xu, journalName=Application Research of Computers, refType=null, unstructuredReference=Sun Tongfeng, Wang Kang, Hao Xu. Dual-metric siamese neural network for few-shot learning[J]. Application Research of Computers, 2023, 40(9): 2851-2855., articleTitle=Dual-metric siamese neural network for few-shot learning, refAbstract=null), Reference(id=1218843921057235843, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=4282, pageEnd=4291, url=null, language=null, rfNumber=[13], rfOrder=22, authorNames=Li B, Wu W, Wang Q, journalName=2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), refType=null, unstructuredReference=Li B, Wu W, Wang Q, et al. SiamRPN++: Evolution of siamese visual tracking with very deep networks[C]// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). California: Long Beach, 2019: 4282-4291., articleTitle=SiamRPN++: Evolution of siamese visual tracking with very deep networks, refAbstract=null), Reference(id=1218843921153704841, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=770, pageEnd=778, url=null, language=null, rfNumber=[14], rfOrder=23, authorNames=He K, Zhang X, Ren S, journalName=IEEE Conference on Computer Vision and Pattern Recognition, refType=null, unstructuredReference=He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]// IEEE Conference on Computer Vision and Pattern Recognition. Nevada: IEEE, 2016: 770-778., articleTitle=Deep residual learning for image recognition, refAbstract=null), Reference(id=1218843921279533967, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2020, volume=167, issue=null, pageStart=1, pageEnd=16, url=null, language=null, rfNumber=[15], rfOrder=24, authorNames=Kumar A, Gandhi C P, Zhou Y, journalName=Applied Acoustics, refType=null, unstructuredReference=Kumar A, Gandhi C P, Zhou Y, et al. Improved deep convolution neural network (CNN) for the identification of defects in the centrifugal pump using acoustic images[J]. Applied Acoustics, 2020, 167: 1-16., articleTitle=Improved deep convolution neural network (CNN) for the identification of defects in the centrifugal pump using acoustic images, refAbstract=null), Reference(id=1218843921396974483, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2017, volume=232, issue=6, pageStart=757, pageEnd=772, url=null, language=null, rfNumber=[16], rfOrder=25, authorNames=Kumar A, Kumar R, journalName=Proceedings of the Institution of Mechanical Engineers Part Journal of Engineering Tribology, refType=null, unstructuredReference=Kumar A, Kumar R. Oscillatory behavior based wavelet decomposition for bearing condition monitoring in the centrifugal pump[J]. Proceedings of the Institution of Mechanical Engineers Part Journal of Engineering Tribology, 2017, 232(6): 757-772., articleTitle=Oscillatory behavior based wavelet decomposition for bearing condition monitoring in the centrifugal pump, refAbstract=null), Reference(id=1218843921526997914, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2017, volume=108, issue=null, pageStart=119, pageEnd=133, url=null, language=null, rfNumber=[17], rfOrder=26, authorNames=Kumar A, Kumar R, journalName=Measurement, refType=null, unstructuredReference=Kumar A, Kumar R. Time-frequency analysis and support vector machine in automatic detection of defect from vibration signal of centrifugal pump[J]. Measurement, 2017, 108: 119-133., articleTitle=Time-frequency analysis and support vector machine in automatic detection of defect from vibration signal of centrifugal pump, refAbstract=null), Reference(id=1218843921631855516, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, doi=null, pmid=null, pmcid=null, year=2022, volume=165, issue=null, pageStart=1, pageEnd=14, url=null, language=null, rfNumber=[18], rfOrder=27, authorNames=Kumar A, Tang H, Vashishtha G, journalName=Mechanical Systems and Signal Processing, refType=null, unstructuredReference=Kumar A, Tang H, Vashishtha G, et al. Noise subtraction and marginal enhanced square envelope spectrum (MESES) for the identification of bearing defects in centrifugal and axial pump[J]. Mechanical Systems and Signal Processing, 2022, 165: 1-14., articleTitle=Noise subtraction and marginal enhanced square envelope spectrum (MESES) for the identification of bearing defects in centrifugal and axial pump, refAbstract=null)], funds=[Fund(id=1218843917991199454, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, awardId=ZLZX2020-05-02, language=CN, fundingSource=中国石油天然气集团有限公司战略合作科技专项(ZLZX2020-05-02), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1218843909204132815, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=1, ext=[AuthorCompanyExt(id=1218843909229298643, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China), AuthorCompanyExt(id=1218843909241881557, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909204132815, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 中国石油大学(北京)安全与海洋工程学院, 北京 102249)]), AuthorCompany(id=1218843909380293602, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=2, ext=[AuthorCompanyExt(id=1218843909384487908, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China), AuthorCompanyExt(id=1218843909392876518, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909380293602, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249)]), AuthorCompany(id=1218843909501928430, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, xref=3, ext=[AuthorCompanyExt(id=1218843909510317040, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909501928430, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 China National Petroleum International Pipeline Co., Ltd., Beijing 102206, China), AuthorCompanyExt(id=1218843909518705650, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, companyId=1218843909501928430, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 中油国际管道有限公司, 北京 102206)])], figs=[ArticleFig(id=1218843913499099526, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.1, caption=Residual learning module, figureFileSmall=bLWBmBV640ME1sfulISydw==, figureFileBig=3e3ma6bT7rwZuZyvv+q5HA==, tableContent=null), ArticleFig(id=1218843913620734352, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图1, caption=残差学习模块

F(x)为残差函数分支,包含两个权重层;x则为恒等映射;H(x)为网络的输出,为残差函数F(x)与恒等映射x之和

, figureFileSmall=bLWBmBV640ME1sfulISydw==, figureFileBig=3e3ma6bT7rwZuZyvv+q5HA==, tableContent=null), ArticleFig(id=1218843913750757785, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.2, caption=The receptive field growth brought about by dilated convolution, figureFileSmall=YRD5vzMv5YgY5TzLCF10vA==, figureFileBig=YuFjM5u3RJwC8/yxxSDU8g==, tableContent=null), ArticleFig(id=1218843913901752734, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图2, caption=膨胀卷积带来的感受野增长, figureFileSmall=YRD5vzMv5YgY5TzLCF10vA==, figureFileBig=YuFjM5u3RJwC8/yxxSDU8g==, tableContent=null), ArticleFig(id=1218843914006610347, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.3, caption=Dilated residual learning module, figureFileSmall=KoL9V0d45aI9h5lOmkV7kw==, figureFileBig=8kU3FnPzK6vRo2KzB9Sguw==, tableContent=null), ArticleFig(id=1218843914174382521, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图3, caption=膨胀残差学习模块, figureFileSmall=KoL9V0d45aI9h5lOmkV7kw==, figureFileBig=8kU3FnPzK6vRo2KzB9Sguw==, tableContent=null), ArticleFig(id=1218843914304405961, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.4, caption=DRN network architecture, figureFileSmall=zqqajY142/1CSLALD1Op4w==, figureFileBig=9MzM7pxHcw2CBrWTyB/x4A==, tableContent=null), ArticleFig(id=1218843914438623698, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图4, caption=DRN网络结构, figureFileSmall=zqqajY142/1CSLALD1Op4w==, figureFileBig=9MzM7pxHcw2CBrWTyB/x4A==, tableContent=null), ArticleFig(id=1218843914593812956, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.5, caption=Traditional siamese network architecture, figureFileSmall=JbWb7JdGvhVwGXdoXFGo6g==, figureFileBig=IeNeWKCkARextu4tyAKKSQ==, tableContent=null), ArticleFig(id=1218843914753196518, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图5, caption=传统孪生网络结构, figureFileSmall=JbWb7JdGvhVwGXdoXFGo6g==, figureFileBig=IeNeWKCkARextu4tyAKKSQ==, tableContent=null), ArticleFig(id=1218843914895802867, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.6, caption=DRSN network architecture, figureFileSmall=IzqPoJSMC+o9N+1/VqdRpg==, figureFileBig=XrdjZluP+oEumGSBMDKbbg==, tableContent=null), ArticleFig(id=1218843915017437695, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图6, caption=DRSN网络结构, figureFileSmall=IzqPoJSMC+o9N+1/VqdRpg==, figureFileBig=XrdjZluP+oEumGSBMDKbbg==, tableContent=null), ArticleFig(id=1218843915139072516, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.7, caption=Logic diagram of classifier operation, figureFileSmall=Tib3TEComSAzGL41Lz4BhA==, figureFileBig=FMjXpJbKDPF7dLJqr5GpKA==, tableContent=null), ArticleFig(id=1218843915264901647, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图7, caption=分类器工作逻辑图, figureFileSmall=Tib3TEComSAzGL41Lz4BhA==, figureFileBig=FMjXpJbKDPF7dLJqr5GpKA==, tableContent=null), ArticleFig(id=1218843915357176343, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.8, caption=Flow chart of centrifugal pump fault diagnosis based on DSRN, figureFileSmall=UGL60LNEWqBSUhs4aQvOMw==, figureFileBig=1fuivGKIzB6bT4Eon1gWVw==, tableContent=null), ArticleFig(id=1218843915474616865, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图8, caption=基于DSRN的离心泵故障诊断流程图, figureFileSmall=UGL60LNEWqBSUhs4aQvOMw==, figureFileBig=1fuivGKIzB6bT4Eon1gWVw==, tableContent=null), ArticleFig(id=1218843915608834603, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.9, caption=Schematic diagram of centrifugal pump device, figureFileSmall=DQ6VPC7HknlAGE36+vO+Bg==, figureFileBig=TloIB1OCh9/yZCudw9aVfA==, tableContent=null), ArticleFig(id=1218843915747246646, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图9, caption=离心泵装置示意图, figureFileSmall=DQ6VPC7HknlAGE36+vO+Bg==, figureFileBig=TloIB1OCh9/yZCudw9aVfA==, tableContent=null), ArticleFig(id=1218843915873075777, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.10, caption=Schematic diagram of sliding window sampling, figureFileSmall=NKJ0fjg9bhI11aJpRX4mmQ==, figureFileBig=AkO3KgV3zRdHlofQbYWKQw==, tableContent=null), ArticleFig(id=1218843915998904902, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图10, caption=滑窗取样示意图, figureFileSmall=NKJ0fjg9bhI11aJpRX4mmQ==, figureFileBig=AkO3KgV3zRdHlofQbYWKQw==, tableContent=null), ArticleFig(id=1218843916141511248, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.11, caption=Diagnostic accuracy of different feature extraction modules, figureFileSmall=UQQ35FqNcIYx07qCjb9G7A==, figureFileBig=yNaqpowHm30VRGiAf5IYxQ==, tableContent=null), ArticleFig(id=1218843916217008734, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图11, caption=不同特征提取模块的诊断准确率, figureFileSmall=UQQ35FqNcIYx07qCjb9G7A==, figureFileBig=yNaqpowHm30VRGiAf5IYxQ==, tableContent=null), ArticleFig(id=1218843916313477734, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.12, caption=Training situation of networks with different sample sets, figureFileSmall=OHgAylK9cW+94Pk7Hnwplg==, figureFileBig=30Pqtyy6JwIuX0lPSOlDJQ==, tableContent=null), ArticleFig(id=1218843916443501168, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图12, caption=不同样本集网络训练情况, figureFileSmall=OHgAylK9cW+94Pk7Hnwplg==, figureFileBig=30Pqtyy6JwIuX0lPSOlDJQ==, tableContent=null), ArticleFig(id=1218843916539970171, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.13, caption=Siamese network classification confusion matrix, figureFileSmall=ZCHoPIq84dbJ7o4gIG2zng==, figureFileBig=uIwBDxzBvOsZlATkKOLknA==, tableContent=null), ArticleFig(id=1218843916665799298, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图13, caption=孪生网络分类混淆矩阵, figureFileSmall=ZCHoPIq84dbJ7o4gIG2zng==, figureFileBig=uIwBDxzBvOsZlATkKOLknA==, tableContent=null), ArticleFig(id=1218843916846154384, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Fig.14, caption=Comparison of diagnostic effects of different methods, figureFileSmall=n943NjuLkB5ymRm7i1Yfhw==, figureFileBig=+1LFiQ9kttR6aiLF3q/jZg==, tableContent=null), ArticleFig(id=1218843916963594907, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=图14, caption=不同方法诊断效果对比, figureFileSmall=n943NjuLkB5ymRm7i1Yfhw==, figureFileBig=+1LFiQ9kttR6aiLF3q/jZg==, tableContent=null), ArticleFig(id=1218843917093618342, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Table 1, caption=

DRSN network architecture

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层 卷积核尺寸
及数量
左分支
(共享权重)
右分支
(共享权重)
输入
Conv1 Conv1 1×3,16, stride2
Conv2_x Conv2_x 1×3Max Pooling, stride2
1 × 3,32 , R d = 1 1 × 3,32 , R d = 2 1 × 3,32 , R d = 4×2
Conv3_x Conv3_x 1 × 3,64 , R d = 1 1 × 3,64 , R d = 2 1 × 3,64 , R d = 4×2
Conv4_x Conv4_x 1 × 3,128 , R d = 1 1 × 3,128 , R d = 2 1 × 3,128 , R d = 4×2
Global Max Pooling
Dense
Output_h1 Output_h2
L2-Distance_z
Dense
), ArticleFig(id=1218843917211058862, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=表1, caption=

DRSN网络结构

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层 卷积核尺寸
及数量
左分支
(共享权重)
右分支
(共享权重)
输入
Conv1 Conv1 1×3,16, stride2
Conv2_x Conv2_x 1×3Max Pooling, stride2
1 × 3,32 , R d = 1 1 × 3,32 , R d = 2 1 × 3,32 , R d = 4×2
Conv3_x Conv3_x 1 × 3,64 , R d = 1 1 × 3,64 , R d = 2 1 × 3,64 , R d = 4×2
Conv4_x Conv4_x 1 × 3,128 , R d = 1 1 × 3,128 , R d = 2 1 × 3,128 , R d = 4×2
Global Max Pooling
Dense
Output_h1 Output_h2
L2-Distance_z
Dense
), ArticleFig(id=1218843917324305080, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Table 2, caption=

Fault data sample set

, figureFileSmall=null, figureFileBig=null, tableContent=
故障类型 样本量 故障标签
训练集 测试集
叶轮断裂(BI) 500 100 0
叶轮堵塞(CI) 500 100 1
轴承内圈缺陷(IR) 500 100 2
轴承外圈缺陷(OR) 500 100 3
无缺陷(DF) 500 100 4
), ArticleFig(id=1218843917437551296, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=表2, caption=

故障数据样本集

, figureFileSmall=null, figureFileBig=null, tableContent=
故障类型 样本量 故障标签
训练集 测试集
叶轮断裂(BI) 500 100 0
叶轮堵塞(CI) 500 100 1
轴承内圈缺陷(IR) 500 100 2
轴承外圈缺陷(OR) 500 100 3
无缺陷(DF) 500 100 4
), ArticleFig(id=1218843917550797508, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=EN, label=Table 3, caption=

Training sample distribution

, figureFileSmall=null, figureFileBig=null, tableContent=
样本集 每类样
本量
样本总量 训练样本
对数量
验证样本
对数量
Dataset A 50 250 8 575 3 675
Dataset B 40 200 5 460 2 340
Dataset C 30 150 3 045 1 305
Dataset D 20 100 1 330 570
Dataset E 10 50 315 135
), ArticleFig(id=1218843917697598159, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149776902736536290, language=CN, label=表3, caption=

训练样本分布

, figureFileSmall=null, figureFileBig=null, tableContent=
样本集 每类样
本量
样本总量 训练样本
对数量
验证样本
对数量
Dataset A 50 250 8 575 3 675
Dataset B 40 200 5 460 2 340
Dataset C 30 150 3 045 1 305
Dataset D 20 100 1 330 570
Dataset E 10 50 315 135
)], attaches=null, journal=Journal(id=1146119176004939786, delFlag=0, nameCn=科学技术与工程, nameEn=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, issn=1671-1815, eissn=, cn=11-4688/T, coden=null, periodic=4, 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=UKU/O7GSka5polgCTkbIIw==, journalPrice=null, startedYear=null, abbrevIsoEn=Sci Technol Eng, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1754445529766, createdBy=null, updatedBy=13701087609, firstLetterCn=S, firstLetterEn=S, subjectCode=Natural Sciences, subjectName=自然科学, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=UKU/O7GSka5polgCTkbIIw==, picEn=5hwlULoNwcbj3xUmVi9MAQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1159791870395564357, language=CN, name=科学技术与工程, 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.stae.com.cn/jsygc/home, createdTime=1754445529793, updatedTime=1754445529793, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.stae.com.cn/jsygc/site/menus/20090429150146001, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1159791870441701702, language=EN, name=Science Technology and Engineering, 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.stae.com.cn/jsygc/home, createdTime=1754445529804, updatedTime=1754445529804, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1146123166801305609, websiteList=[Website(id=1148243202391400884, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, 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/kxjsygc/CN, language=CN, createTime=1751692112777, createBy=18614031015, updateTime=1753520965431, updateBy=18614031015, name=科学技术与工程-中文站点, tplId=1146099689490845704, title=科学技术与工程, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148622798802673703, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=articleTextType, value=kx, createTime=1751782615614, updateTime=1751782615614, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798781702180, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=banner, value=null, createTime=1751782615609, updateTime=1751782615609, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798769119267, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1751782615606, updateTime=1751782615606, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798794285094, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751782615612, updateTime=1751782615612, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798790090789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751782615611, updateTime=1751782615611, creator=18614031015, updator=18614031015)]), Website(id=1155914124811976731, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, 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/kxjsygc/EN, language=EN, createTime=1753521003206, createBy=18614031015, updateTime=1753521003206, updateBy=18614031015, name=科学技术与工程-英文站点, tplId=1146101810881728533, title=Science Technology and Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155914371227308235, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=articleTextType, value=kx, createTime=1753521061952, updateTime=1753521061952, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371210531016, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=banner, value=null, createTime=1753521061947, updateTime=1753521061947, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371202142407, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1753521061945, updateTime=1753521061945, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371223113930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753521061950, updateTime=1753521061950, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371218919625, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753521061949, updateTime=1753521061949, creator=18614031015, updator=18614031015)])], journalTitle=科学技术与工程, weixinUrl=null, journalUrl=null, iacademicId=null, status=0, seqNo=null, journalTitleEn=Science Technology and Engineering, journalPhotoCn=UKU/O7GSka5polgCTkbIIw==, journalPhotoEn=5hwlULoNwcbj3xUmVi9MAQ==, journalFirstLetter=S, 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/kxjsygc/CN/10.12404/j.issn.1671-1815.2403506, detailUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/10.12404/j.issn.1671-1815.2403506, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/PDF/10.12404/j.issn.1671-1815.2403506, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/PDF/10.12404/j.issn.1671-1815.2403506, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
小样本下离心泵的孪生网络故障诊断方法
收藏切换
PDF下载
李科 1, 2 , 张来斌 1, 2 , 段礼祥 1, 2, * , 刘海鹏 3 , 张馨月 1, 2
科学技术与工程 | 论文·机械、仪表工业 2025,25(11): 4543-4550
收起
收藏切换
科学技术与工程 | 论文·机械、仪表工业 2025, 25(11): 4543-4550
小样本下离心泵的孪生网络故障诊断方法
全屏
李科1, 2 , 张来斌1, 2, 段礼祥1, 2, * , 刘海鹏3, 张馨月1, 2
作者信息
  • 1 中国石油大学(北京)安全与海洋工程学院, 北京 102249
  • 2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249
  • 3 中油国际管道有限公司, 北京 102206
  • 李科(1999—),男,汉族,安徽六安人,硕士研究生。研究方向:安全监测与智能诊断。E-mail:

通讯作者:

* 段礼祥(1969—),男,汉族,四川泸州人,博士,教授。研究方向:安全监测与智能诊断工程、机械设备状态监测与故障诊断。E-mail:
Fault Diagnosis Method of Centrifugal Pump Based on Siamese Networks under Small Sample Conditions
Ke LI1, 2 , Lai-bin ZHANG1, 2, Li-xiang DUAN1, 2, * , Hai-peng LIU3, Xin-yue ZHANG1, 2
Affiliations
  • 1 College of Safety and Ocean Engineering, China University of Petroleum(Beijing), Beijing 102249, China
  • 2 Key Laboratory of Oil and Gas Production Safety and Emergency Technology, Ministry of Emergency Management, Beijing 102249, China
  • 3 China National Petroleum International Pipeline Co., Ltd., Beijing 102206, China
出版时间: 2025-04-18 doi: 10.12404/j.issn.1671-1815.2403506
文章导航
收藏切换

实际工程中需要大量数据支撑的常规诊断方法难以有效进行小样本条件下的离心泵故障诊断,为此,将深度学习中的残差网络(residual network,ResNet)与膨胀卷积相结合,并拓展为孪生网络,构建膨胀残差孪生网络(dilated residual siamese network,DRSN)。将膨胀残差网络作为孪生网络的特征提取模块,强化了模型的特征提取能力;构造正负样本对,从每个样本中提取更多的信息,更有效地利用有限的数据;两个子网络共享参数,减少自由参数的数量,降低样本不足时过拟合的风险。提出的网络模型缓解了训练样本不足的问题,提升了数据利用的效率,实现了小样本条件下的离心泵故障分类。研究结果表明:在样本最匮乏的情况下,该模型在离心泵试验数据集上的准确率仍能达到82.20%,相较其他模型,准确率至少提升了8.8个百分点。

离心泵  /  故障诊断  /  小样本  /  残差网络  /  孪生网络

Conventional diagnostic methods that require a large amount of data support in practical engineering are difficult to effectively perform centrifugal pump fault diagnosis under small sample conditions. Therefore, the residual network (ResNet) in deep learning was combined with dilated convolution and extended into a siamese network to construct a dilated residual siamese network (DRSN). The dilated residual network was used as the feature extraction module of the siamese network, which enhanced the feature extraction ability of the model. Positive and negative sample pairs were constructed to extract more information from each sample, and make more effective use of limited data.The two sub-networks share parameters, the number of free parameters and lowering the risk of overfitting was reduced when the sample was insufficient. The proposed network model alleviated the problem of insufficient training samples, improved the efficiency of data utilization, and realized the fault classification of centrifugal pump under the condition of small samples. The research results show that even in the most sample-scarce situation, the accuracy of the model on the centrifugal pump test dataset can still reach 82.20%, which is at least 8.8 percentage points higher than other models.

centrifugal pump  /  fault diagnosis  /  small sample  /  residual network  /  siamese network
李科, 张来斌, 段礼祥, 刘海鹏, 张馨月. 小样本下离心泵的孪生网络故障诊断方法. 科学技术与工程, 2025 , 25 (11) : 4543 -4550 . DOI: 10.12404/j.issn.1671-1815.2403506
Ke LI, Lai-bin ZHANG, Li-xiang DUAN, Hai-peng LIU, Xin-yue ZHANG. Fault Diagnosis Method of Centrifugal Pump Based on Siamese Networks under Small Sample Conditions[J]. Science Technology and Engineering, 2025 , 25 (11) : 4543 -4550 . DOI: 10.12404/j.issn.1671-1815.2403506
在离心泵故障诊断工程中,由于数据采集成本高昂、企业出于隐私和安全的考虑限制设备数据的访问和收集等原因,很难获得充足的监测数据,导致小样本问题[1-2]:常规方法在样本不足的情况下,会过度学习数据中的噪声等偶然特征,导致过拟合问题;常规方法直接学习原始的数据,对数据的利用效率较低,难以学习到有效的特征表示。
为了解决这个问题,小样本学习(few-shot learning, FSL)[3]应运而生,其核心理念在于通过分析并利用有限数据集来训练算法模型,以达到精确分类与识别的目标。通过学习少量样本的特征,将这些特征泛化到整个数据集中,从而实现对整体数据集的分类和识别。当前,小样本学习策略的研究受到了广泛关注。针对有限数据集处理的研究策略主要分为四类,分别为数据增强[4]、迁移学习(transfer learning, TL)[5]、元学习(meta-learning,ML)[6]和度量学习(metric learning,ML)[7]。由于实际工程中,采集的离心泵数据故障种类多,每种类别样本较少,而度量学习中的孪生网络(siamese network,SN)特别适用于以上情形,故考虑使用孪生网络。
孪生网络通过训练样本之间的相似性度量,进行反向更新优化,拉近相似样本的距离并分割不相似样本。孪生网络通常用于分类和聚类任务中,通过优化样本间的距离度量,提升模型的判别能力。通过精心设计的距离度量,孪生网络能够提升模型在处理复杂模式和细微差别方面的精确度。目前,故障诊断领域已经在轴承[8]、齿轮箱[9]、电机[10]等设备的小样本故障诊断中证实了孪生神经网络的可行性,但对离心泵的相关研究还相对匮乏。同时,目前针对孪生网络的改进大多集中在度量方法上,如改进自适应的网络度量方式[11]、使用自适应融合策略,整合不同级别特征的度量结果[12]等方法,而特征提取模块过于简单,难以有效学习到数据的表征[13],导致网络易陷入局部最优解问题,影响诊断准确性。因此,需要对孪生网络的特征提取模块进行深入研究和改进。
为解决上述问题,将深度学习中的残差网络(residual network,ResNet)[14]与膨胀卷积结合相结合,并拓展为孪生网络,构建膨胀残差孪生网络(dilated residual siamese network, DRSN),并用于小样本条件下离心泵的故障诊断。通过在残差神经网络中加入膨胀残差块,构建膨胀残差网络,维持特征细节信息的同时扩大感受野,再引入孪生网络结构,将膨胀残差网络作为孪生网络的特征提取模块,构建孪生膨胀残差网络,缓解训练样本不足的问题。
残差神经网络ResNet由多个残差学习模块组合而成,残差学习模块由两个部分组成,其结构如图1所示。
膨胀卷积方法是通过在卷积层引入膨胀率Rd(dilation rate,DR)的概念,来定义卷积操作时卷积核处理数据值的间距,即在卷积核的元素之间添加Rd-1个空格。标准卷积可以看作膨胀卷积在Rd=1时的特殊形式。图2(a)展示了标准卷积和膨胀卷积之间的差异;图2(b)是在第一层特征图上使用Rd=2的卷积得到的特征图,红点为卷积核的实际分布位置。可以看出膨胀卷积的感受野显著增大,但实际参与卷积计算的因子数量和计算复杂度并未改变。
基于ResNet与膨胀卷积,设计一种膨胀残差学习模块,在残差学习模块的基础上,将残差部分权重层中的普通卷积层替换为膨胀卷积层,结构如图3所示。
图3中,卷积层被替换为3个膨胀率Rd分别为1、2、4的3个膨胀卷积层。普通残差学习模块有两个权重层,而膨胀残差学习模块的权重层被拓展成3个。这样,单个模块的深度增加,整体网络的深度和参数量就会发生变化,因此需要对整体网络结构进行设计。
基于ResNet-18的网络结构,引入膨胀残差模块,并删减部分残差模块,以达到改进目的,具体为:原网络结构中的Conv1层不进行改动,替换Conv2_x、Conv3_x和Conv4_x层中的所有残差模块为图5所示的膨胀残差模块,并在Conv3_x层中添加Droput层,去除Conv5_x层。最终获得膨胀残差网络(dilated residual network, DRN),其结构图如图4所示,其中k3n32s1指卷积核尺寸为3,一层32个卷积核,步幅stride为1。
孪生网络结构是利用特征提取网络模块对输入样本对进行特征提取,然后度量两个特征向量在特征空间中的距离,判断特征向量是否属于同一类别。传统孪生网络结构如图5所示。
图5可知,孪生网络是由两条权值共享神经网络并行构成,每次将一对样本x1x2输入到特征提取网络中,提取两个特征向量,通过计算两个特征向量在特征空间中的距离,比较两个样本的相似度。
本文中采用对比损失函数(contrastive loss)进行训练,对比损失计算公式为
L[ω, (x1, x2, y)]= 1 2 N n = 1 Ny d ω 2+ 1 2(1-y)max(m- d ω 2,0)2
式(1)中: d ω 2为样本对特征向量之间的欧氏距离;y为两个样本是否匹配的标签,若y=1,说明两样本归属同一种类,y=0则说明两样本不是同一种类;m表示不相似的距离阈值,即两个不同类样本的距离在[0,m]范围内,超过[0,m]时,可理解为两个不相似样本欧式距离过远,其损失可忽略。
在传统的神经网络分类器中,softmax层被广泛用于多类分类任务。其工作原理是将神经网络最后一层的输出转换成概率分布,每个类别都有一个概率值,而分类任务就是选择具有最高概率的类别。然而,当面对小样本学习或者一些需要比较两个输入相似性的任务时,基于softmax的传统分类器便难以完成分类的任务。基于以上原因,本文选择创建专门基于孪生网络的分类器,满足故障样本的分类需求。
本文中创建的孪生网络分类器中,测试集的样本与支持集的样本进行一对一的比对。支撑集则选择能够代表类别特征的样本构成。这意味着支持集的样本数量可以远远少于传统分类任务中用到的样本数量,可以有效处理小样本问题。绘制分类器逻辑图如图7所示。
膨胀残差孪生网络诊断方法具体流程(图8)如下:
(1)数据采集与预处理:用加速度传感器采集离心泵系统的振动信号,滑窗截取振动信号生成振动信号预处理故障样本时域信息。
(2)样本对构建:在预处理故障样本的基础上构建数量接近的正负样本对,并一一对应添加样本对标签。
(3)孪生网络模型搭建:以DRN模型为特征提取网络,将其进一步拓展为孪生网络模型,将构建的样本对输入模型进行训练,提升模型度量能力。
(4)故障分类:将测试数据与支撑集一同输入到专门构建的分类器中,得到每个测试样本与支撑集各故障类型样本的相似度,进而完成分类。
采用印度朗格瓦尔圣朗格瓦尔工程技术学院的离心泵故障数据集[15-18]进行试验验证,试验装置如图9所示。
本试验选用靠近叶轮的轴承1上的加速度传感器监测的振动信号,采样频率为70 kHz。试验共模拟离心泵的5种运行状态:叶轮断裂(broken impeller,BI)、叶轮堵塞(clogged impeller,CI)、轴承内圈缺陷(bearing with inner race defect,IR)、轴承外圈缺陷(bearing with outer race defect, OR)以及无缺陷(defect free,DF)。通过安装不同类型的缺陷部件,模拟离心泵在运行过程中出现的不同故障状态。采用滑窗取样的方法划分原始数据,如图10所示。
从原始振动信号的起始点开始,每次采集1 024个数据点,采集完成后滑动窗口向后移动256个数据点继续采集。最终获取到每种模式500个训练集样本、100个测试样本,具体信息如表2所示。
为了探究训练样本数量对孪生网络模型性能的具体影响,本研究从表2展示的训练集中随机选取不同数量的样本,即每类故障分别选择抽取50、40、30、20、10个样本,形成5个不同的样本集,分别命名为Dataset A、B、C、D、E,旨在模拟故障训练样本量不足的情况,同时设定测试集的总样本量为500。利用这5种样本集合中的数据,构建正负样本对,并遵循7∶3的比例将其分为训练样本和验证样本,具体分布如表3所示。
为了分析孪生网络模型中特征提取模块对故障诊断性能的影响,选择本文提出的DRN与VGG16和ResNet18分别作为特征提取模块,构建孪生网络(以下分别简称DRSN、SVGG、RSN),在表3所示的不同训练集下测试了不同孪生网络的诊断准确率,结果如图11所示。
图11分析可知,SVGG在各规模样本集上的故障诊断性能均为最低,当训练样本250个时,SVGG的诊断准确率为87.20%当样本总量减少到50个时,准确率显著下滑至62.40%与之相比,RSN模型诊断效果得到了明显改善。尤其是在样本量较多的情况下,RSN的诊断准确率与DRSN非常相近。然而,随着样本总量的不断减少,RSN诊断准确率下降的趋势更加明显,当样本总量仅有50个时,准确率下降到了70.60%,而此时DRSN的准确率为82.20%,比SVGG、SDN分别高出了19.8个百分点和11.6个百分点。以上分析说明,DRN作为特征提取模块的孪生网络相比其他模型,能够提取有限样本的细节特征信息,实现小样本条件下的故障样本精确分类。
基于膨胀残差网络的孪生网络模型在不同训练样本量下的训练情况,包括准确率和损失函数随迭代次数的变化情况如图12所示。
可以看到,训练样本量会显著影响孪生网络模型训练收敛的速度,随着样本总量的减少,模型收敛需要训练的次数逐渐增加,训练样本的数量对孪生网络模型的训练收敛速度有显著影响,较少的样本量导致模型获取的信息量减少,准确率提升缓慢,模型收敛速度下降。
以DatasetA和DatasetE训练得到的模型为例,绘制了故障分类的混淆矩阵,如图13所示。分析可知:网络对标签0与标签1的故障类型的学习不够充分,剩余标签的故障受样本量影响较小。
尽管在极端小样本情况下(即使是样本最多的样本集也仅有300个样本),分类准确率有所下降,但与常规需要3 000甚至更多样本的训练相比,本试验中的模型仍能显示出相对较强的分类能力。这突显了DRSN模型在处理小样本问题时的优异泛化能力。
为了验证所提方法在小样本故障诊断领域的优势,将本文构建的DRSN模型与未拓展的DRN、ResNet-18、VGG16和CNN1D共4种方法进行对比试验。通过对比在不同轴承样本集下测试各模型的准确率表现,评估模型在故障诊断任务中的效果,试验结果如图14所示。
图14可知,随着数据集中样本个数的减少,各个模型分类准确率都出现了不同程度的下降。但是本文提出的DRSN在不同样本数量下的准确率始终高于其他模型,且数据量越小,优势越明显。当每类故障只有10个样本时,DRSN的准确率仍有82.20%,分别比对比模型高出8.8、12.8、20.6与24.8个百分点,提升十分显著。这说明DRSN比其他方法更适合小样本条件下的离心泵故障诊断。
(1)针对小样本情况下模型从有限样本中学习复杂特征的能力有限和数据利用效率低下,导致故障诊断准确率下降的问题,在DRN的基础上,引入孪生网络结构,提出了膨胀残差孪生网络。将少量样本通过构造样本对的方式进行扩充,训练网络模型的度量能力并构建了适用于孪生网络的故障样本分类器,讨论不同特征提取模块对小样本故障诊断的影响,最后通过离心泵数据集的小样本故障诊断试验,验证了模型的性能,为小样本故障诊断提供了新思路。
(2)本文中使用的DRN特征提取模块可以有效提取敏感特征,将不同域的信息互补,在小样本条件下显著优于其他特征提取模块,为实现孪生网络相似度准确的对比能力提供了保障。齿轮箱轴承试验中,在每类故障只有10样本的条件下,本文模型的准确率比使用VGG16、Resnet18作为特征提取网络的模型分别高出19.8个百分点和11.6个百分点。
(3)本文中使用的孪生网络模型结构能够在小样本条件下更好地利用数据信息学习不同故障类型的特征表示,并减少过拟合风险。在离心泵试验中,作为对比的其他神经网络会随着训练样本的减少出现性能的严重下滑,而本文模型的准确率在每类故障只有10样本的条件下,仍能保持82.20%的准确率,显著优于其他网络模型。表明本文所提模型能够在样本数量不足的情况下实现更准确的故障分类。
  • 中国石油天然气集团有限公司战略合作科技专项(ZLZX2020-05-02)
参考文献 引证文献
排序方式:
[1]
焦瀚晖, 胡明辉, 江志农, 等. 基于补偿距离评估和一维卷积神经网络的离心泵故障快速智能识别方法[J]. 振动与冲击, 2021, 40(10): 41-49.
Jiao Hanhui, Hu Minghui, Jiang Zhinong, et al. Rapid intelligent fault identification method for centrifugal pumps based on compensation distance evaluation and one-dimensional convolutional neural networks[J]. Journal of Vibration and Shock, 2021, 40(10): 41-49.
[2]
孙弋. 基于深度学习的小样本条件下轴承故障诊断技术研究[D]. 成都: 西南交通大学, 2021.
Sun Yi. Research on bearing fault diagnosis technology under small sample conditions based on deep learning[D]. Chengdu: Southwest Jiaotong University, 2021.
[3]
Li F F, Fergus R, Perona P, et al. One-shot learning of object categories[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(4): 594-611.
[4]
邵海东, 李伟, 刘翊, 等. 时变转速下基于双阈值注意力生成对抗网络和小样本的转子-轴承系统故障诊断[J]. 机械工程学报, 2023, 59(12): 215-224.
Shao Haidong, Li Wei, Liu Yi, et al. Fault diagnosis of rotor-bearing system under time-varying speeds by using dual-threshold attention-embedded GAN and small samples[J]. Journal of Mechanical Engineering, 2023, 59(12): 215-224.
[5]
Dev K, Ashraf Z, Muhuri P K, et al. Deep autoencoder based domain adaptation for transfer learning[J]. Multimedia Tools and Applications, 2022, 81(16): 22379-22405.
[6]
丁鹏. 有限数据下滚动轴承变工况寿命预测方法及应用研究[D]. 南京: 东南大学, 2022.
Ding Peng. Research on limited data-driven life prediction methods and applications of rolling bearings under variable operating condition[D]. Nanjing: Southeast University, 2022.
[7]
张卡, 宿东, 王蓬勃, 等. 深度学习技术在影像密集匹配方面的进展与应用[J]. 科学技术与工程, 2020, 20(30): 12268-12278.
Zhang Ka, Su Dong, Wang Pengbo, et al. Progress and application of deep learning in image dense matching[J]. Science Technology and Engineering, 2020, 20(30): 12268-12278.
[8]
邬娜, 王健, 杨建伟, 等. 变工况及小样本情况下滚动轴承故障迁移学习方法综述[J]. 科学技术与工程, 2024, 24(10): 3939-3951.
Wu Na, Wang Jian, Yang Jianwei, et al. Review of rolling bearing migration learning under variable operating conditions and small samples[J]. Science Technology and Engineering, 2024, 24(10): 3939-3951.
[9]
钱心筠, 王友仁, 赵亚磊. 基于孪生网络的行星齿轮箱故障诊断方法[J]. 机械制造与自动化, 2022, 51(4): 116-119.
Qian Xinyun, Wang Youren, Zhao Yalei. Method of planetary gearbox fault diagnosis based on siamese networks[J]. Machine Building & Automation, 2022, 51(4): 116-119.
[10]
张青雨. 基于深度胶囊网络的旋转机械故障智能诊断方法研究[D]. 秦皇岛: 燕山大学, 2023.
Zhang Qingyu. Research on mechanical fault intelligent diagnosis based on deep capsule network[D]. Qihuangdao: Yanshan University, 2023.
[11]
赵晓平, 彭澎, 张永宏, 等. 改进孪生网络在小样本轴承故障诊断中的应用[J]. 计算机工程与应用, 2023, 59(19): 294-304.
Zhao Xiaoping, Peng Peng, Zhang Yonghong, et al. Application of improved siamese neural network in small sample fault diagnosis of bearing[J]. Computer Engineering and Applications, 2023, 59(19): 294-304.
[12]
孙统风, 王康, 郝徐. 面向小样本学习的双重度量孪生神经网络[J]. 计算机应用研究, 2023, 40(9): 2851-2855.
Sun Tongfeng, Wang Kang, Hao Xu. Dual-metric siamese neural network for few-shot learning[J]. Application Research of Computers, 2023, 40(9): 2851-2855.
[13]
Li B, Wu W, Wang Q, et al. SiamRPN++: Evolution of siamese visual tracking with very deep networks[C]// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). California: Long Beach, 2019: 4282-4291.
[14]
He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]// IEEE Conference on Computer Vision and Pattern Recognition. Nevada: IEEE, 2016: 770-778.
[15]
Kumar A, Gandhi C P, Zhou Y, et al. Improved deep convolution neural network (CNN) for the identification of defects in the centrifugal pump using acoustic images[J]. Applied Acoustics, 2020, 167: 1-16.
[16]
Kumar A, Kumar R. Oscillatory behavior based wavelet decomposition for bearing condition monitoring in the centrifugal pump[J]. Proceedings of the Institution of Mechanical Engineers Part Journal of Engineering Tribology, 2017, 232(6): 757-772.
[17]
Kumar A, Kumar R. Time-frequency analysis and support vector machine in automatic detection of defect from vibration signal of centrifugal pump[J]. Measurement, 2017, 108: 119-133.
[18]
Kumar A, Tang H, Vashishtha G, et al. Noise subtraction and marginal enhanced square envelope spectrum (MESES) for the identification of bearing defects in centrifugal and axial pump[J]. Mechanical Systems and Signal Processing, 2022, 165: 1-14.
2025年第25卷第11期
PDF下载
324
136
引用本文
BibTeX
文章信息
doi: 10.12404/j.issn.1671-1815.2403506
  • 接收时间:2024-05-12
  • 首发时间:2025-07-09
  • 出版时间:2025-04-18
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-05-12
  • 修回日期:2024-08-10
基金
中国石油天然气集团有限公司战略合作科技专项(ZLZX2020-05-02)
作者信息
    1 中国石油大学(北京)安全与海洋工程学院, 北京 102249
    2 应急管理部油气生产安全与应急技术重点实验室, 北京 102249
    3 中油国际管道有限公司, 北京 102206

通讯作者:

* 段礼祥(1969—),男,汉族,四川泸州人,博士,教授。研究方向:安全监测与智能诊断工程、机械设备状态监测与故障诊断。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2403506
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