Article(id=1203281641063035100, tenantId=1146029695717560320, journalId=1149653034449285133, issueId=1203281635908231645, articleNumber=null, orderNo=null, doi=10.16790/j.cnki.1009-9239.im.2025.06.015, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1716220800000, receivedDateStr=2024-05-21, revisedDate=1718726400000, revisedDateStr=2024-06-19, acceptedDate=null, acceptedDateStr=null, onlineDate=1764814299120, onlineDateStr=2025-12-04, pubDate=1750348800000, pubDateStr=2025-06-20, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1764814299120, onlineIssueDateStr=2025-12-04, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1764814299120, creator=13701087609, updateTime=1764814299120, updator=13701087609, issue=Issue{id=1203281635908231645, tenantId=1146029695717560320, journalId=1149653034449285133, year='2025', volume='58', issue='6', pageStart='1', pageEnd='148', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=0, articleOrder=1, issueType=-1, specialIssue=null, createTime=1764814297892, creator=13701087609, updateTime=1764815002353, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1203284590690545746, tenantId=1146029695717560320, journalId=1149653034449285133, issueId=1203281635908231645, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1203284590694740051, tenantId=1146029695717560320, journalId=1149653034449285133, issueId=1203281635908231645, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=122, endPage=130, ext={EN=ArticleExt(id=1203281641369219293, articleId=1203281641063035100, tenantId=1146029695717560320, journalId=1149653034449285133, language=EN, title=Multi-scale fusion prediction method of dissolved gas in power transformer oil considering spatio-temporal coupling relationship, columnId=1190369198472794288, journalTitle=Insulating Materials, columnName=Insulation Technology, runingTitle=null, highlight=

Multi-scale mining of the spatio-temporal coupling relationship of dissolved gases in oil is helpful to improve the prediction accuracy of dissolved gases in oil and provide a reliable theoretical basis for transformer operation and maintenance decisions. Thereby, a multi-scale fusion prediction method for dissolved gases in transformer oil considering spatio-temporal coupling information was proposed in this study. Firstly, the Res2Net was used to extract the multi-scale time characteristics of the dissolved gas data in oil, and the periodic time feature information of the characteristic gas under different frequencies was captured. Secondly, the implicit relationship between characteristic gases was captured by calculating mutual information, the correlation between different gases was described in the form of topological graphs, and the spatial information features were extracted by using graph convolutional neural network (GCN). Finally, multi-scale temporal information and spatial information were fused and spliced, and temporal convolution network (TCN) was used to predict the dissolved gas in oil. The proposed method was validated using online oil chromatography monitoring data from a 500 kV transformer. The results show that compared with the traditional prediction method, the Res2Net-GCN-TCN model can effectively improve the prediction accuracy of dissolved gas content in oil, and the average prediction accuracy is 98.68%.

, articleAbstract=

Multi-scale mining of the spatio-temporal coupling relationship of dissolved gases in oil is helpful to improve the prediction accuracy of dissolved gases in oil and provide a reliable theoretical basis for transformer operation and maintenance decisions. Thereby, a multi-scale fusion prediction method for dissolved gases in transformer oil considering spatio-temporal coupling information was proposed in this study. Firstly, the Res2Net was used to extract the multi-scale time characteristics of the dissolved gas data in oil, and the periodic time feature information of the characteristic gas under different frequencies was captured. Secondly, the implicit relationship between characteristic gases was captured by calculating mutual information, the correlation between different gases was described in the form of topological graphs, and the spatial information features were extracted by using graph convolutional neural network (GCN). Finally, multi-scale temporal information and spatial information were fused and spliced, and temporal convolution network (TCN) was used to predict the dissolved gas in oil. The proposed method was validated using online oil chromatography monitoring data from a 500 kV transformer. The results show that compared with the traditional prediction method, the Res2Net-GCN-TCN model can effectively improve the prediction accuracy of dissolved gas content in oil, and the average prediction accuracy is 98.68%.

, 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=Qianqian ZHANG, Min LI, Shaosheng GENG, Chunxin WANG, Jun XIE, Qing XIE), CN=ArticleExt(id=1203281648721834437, articleId=1203281641063035100, tenantId=1146029695717560320, journalId=1149653034449285133, language=CN, title=考虑时空耦合关系的电力变压器油中溶解气体多尺度融合预测方法, columnId=1190369198724452531, journalTitle=绝缘材料, columnName=绝缘技术, runingTitle=null, highlight=

对油中溶解气体的时空耦合关系进行多尺度挖掘,有助于提高油中溶解气体预测精度,为变压器运维决策提供可靠理论依据。为此,提出一种考虑时空耦合信息的变压器油中溶解气体多尺度融合预测方法。首先,利用Res2Net对油中溶解气体数据进行多尺度时间特征提取,捕捉特征气体不同频率的周期性时间特征信息。其次,通过计算互信息捕捉特征气体间隐性关系,以拓扑关系图的形式描述不同气体间关联性,并使用图卷积神经网络(GCN)提取空间信息特征。最后,将多尺度时间信息与空间信息进行融合拼接,采用时间卷积网路(TCN)对油中溶解气体进行预测,并使用某500 kV变压器油色谱在线监测数据对所提方法进行验证。结果表明:相比于传统预测方法,Res2Net-GCN-TCN模型可有效提高油中溶解气体含量预测精度,平均预测精度可达98.68%。

, articleAbstract=

对油中溶解气体的时空耦合关系进行多尺度挖掘,有助于提高油中溶解气体预测精度,为变压器运维决策提供可靠理论依据。为此,提出一种考虑时空耦合信息的变压器油中溶解气体多尺度融合预测方法。首先,利用Res2Net对油中溶解气体数据进行多尺度时间特征提取,捕捉特征气体不同频率的周期性时间特征信息。其次,通过计算互信息捕捉特征气体间隐性关系,以拓扑关系图的形式描述不同气体间关联性,并使用图卷积神经网络(GCN)提取空间信息特征。最后,将多尺度时间信息与空间信息进行融合拼接,采用时间卷积网路(TCN)对油中溶解气体进行预测,并使用某500 kV变压器油色谱在线监测数据对所提方法进行验证。结果表明:相比于传统预测方法,Res2Net-GCN-TCN模型可有效提高油中溶解气体含量预测精度,平均预测精度可达98.68%。

, correspAuthors=null, authorNote=null, correspAuthorsNote=
谢军(1988-),男(汉族),江苏江都人,副教授,博士,主要从事高电压试验技术、电力设备状态监测的研究工作。
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=ymxikwg8Aqu3+lxn6dFYwA==, magXml=W7/e8j8M4zlAos3+OKux5g==, pdfUrl=null, pdf=cMFVHciPDKRYOhLKN51fzQ==, pdfFileSize=10607332, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=ngoLz3u7BQydJCsrUZMNng==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=1YstNWV6MgLkjF93BXCtMA==, mapNumber=null, authorCompany=null, fund=null, authors=

张倩倩(2000-),女(汉族),吉林敦化人,硕士生,主要从事变压器状态监测的研究工作;

, authorsList=张倩倩, 李敏, 耿绍胜, 王春鑫, 谢军, 谢庆)}, authors=[Author(id=1203362944089305892, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, orderNo=0, 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=1203362944202552107, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362944089305892, language=EN, stringName=Qianqian ZHANG, firstName=Qianqian, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1203362944303215412, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362944089305892, language=CN, stringName=张倩倩, firstName=倩倩, middleName=null, lastName=张, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.华北电力大学 电力工程系,河北 保定 071003, bio={"content":"

张倩倩(2000-),女(汉族),吉林敦化人,硕士生,主要从事变压器状态监测的研究工作;

"}, bioImg=null, bioContent=

张倩倩(2000-),女(汉族),吉林敦化人,硕士生,主要从事变压器状态监测的研究工作;

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1203362943615349508, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943623738117, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China), AuthorCompanyExt(id=1203362943632126726, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.华北电力大学 电力工程系,河北 保定 071003)])]), Author(id=1203362944391295804, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, 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=1203362944491959107, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362944391295804, language=EN, stringName=Min LI, firstName=Min, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2. State Grid Baoding Power Supply Company, Baoding 071066, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1203362944584233798, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362944391295804, language=CN, stringName=李敏, firstName=敏, middleName=null, lastName=李, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.国网保定供电公司,河北 保定 071066, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1203362943753761555, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943762150164, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943753761555, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. State Grid Baoding Power Supply Company, Baoding 071066, China), AuthorCompanyExt(id=1203362943774733076, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943753761555, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.国网保定供电公司,河北 保定 071066)])]), Author(id=1203362945855107919, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, orderNo=2, 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=1203362945985131353, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362945855107919, language=EN, stringName=Shaosheng GENG, firstName=Shaosheng, middleName=null, lastName=GENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1203362946178069346, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362945855107919, language=CN, stringName=耿绍胜, firstName=绍胜, middleName=null, lastName=耿, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.华北电力大学 电力工程系,河北 保定 071003, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1203362943615349508, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943623738117, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China), AuthorCompanyExt(id=1203362943632126726, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.华北电力大学 电力工程系,河北 保定 071003)])]), Author(id=1203362946257761129, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, 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=1203362946383590257, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362946257761129, language=EN, stringName=Chunxin WANG, firstName=Chunxin, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1203362946446504823, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362946257761129, language=CN, stringName=王春鑫, firstName=春鑫, middleName=null, lastName=王, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.华北电力大学 电力工程系,河北 保定 071003, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1203362943615349508, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943623738117, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China), AuthorCompanyExt(id=1203362943632126726, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.华北电力大学 电力工程系,河北 保定 071003)])]), Author(id=1203362946576528255, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, 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=1203362946756883337, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362946576528255, language=EN, stringName=Jun XIE, firstName=Jun, middleName=null, lastName=XIE, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1203362946928849808, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362946576528255, language=CN, stringName=谢军, firstName=军, middleName=null, lastName=谢, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.华北电力大学 电力工程系,河北 保定 071003, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1203362943615349508, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943623738117, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China), AuthorCompanyExt(id=1203362943632126726, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.华北电力大学 电力工程系,河北 保定 071003)])]), Author(id=1203362947130176416, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, orderNo=5, 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=1203362947251811247, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362947130176416, language=EN, stringName=Qing XIE, firstName=Qing, middleName=null, lastName=XIE, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, address=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China
3. State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (NCEPU), Beijing 102206, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1203362947335697336, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, authorId=1203362947130176416, language=CN, stringName=谢庆, firstName=庆, middleName=null, lastName=谢, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, address=1.华北电力大学 电力工程系,河北 保定 071003
3.新能源电力系统全国重点实验室(华北电力大学),北京 102206, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1203362943615349508, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943623738117, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China), AuthorCompanyExt(id=1203362943632126726, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.华北电力大学 电力工程系,河北 保定 071003)]), AuthorCompany(id=1203362943887979291, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943963476764, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943887979291, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3. State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (NCEPU), Beijing 102206, China), AuthorCompanyExt(id=1203362943976059678, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943887979291, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.新能源电力系统全国重点实验室(华北电力大学),北京 102206)])])], keywords=[Keyword(id=1203362947599938504, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, orderNo=1, keyword=dissolved gas prediction in oil), Keyword(id=1203362947738350544, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, orderNo=2, keyword=Res2Net), Keyword(id=1203362947885151197, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, orderNo=3, keyword=graph convolution), Keyword(id=1203362948044534757, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, orderNo=4, keyword=temporal convolution), Keyword(id=1203362948174558190, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, orderNo=5, keyword=spatio-temporal information fusion), Keyword(id=1203362948283610103, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, orderNo=1, keyword=油中溶解气体预测), Keyword(id=1203362948413632512, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, orderNo=2, keyword=Res2Net), Keyword(id=1203362948497518597, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, orderNo=3, keyword=图卷积), Keyword(id=1203362948589793290, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, orderNo=4, keyword=时间卷积), Keyword(id=1203362948665290770, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, orderNo=5, keyword=时空信息融合)], refs=[Reference(id=1203362954960941289, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=56, issue=12, pageStart=43, pageEnd=53, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=周远翔, 李永印, 陈健宁, journalName=绝缘材料, refType=null, unstructuredReference=周远翔,李永印,陈健宁,等.基于局部放电相位图谱和油中溶解气体信息融合的油纸绝缘缺陷识别方法[J].绝缘材料,2023,56(12):43-53., articleTitle=基于局部放电相位图谱和油中溶解气体信息融合的油纸绝缘缺陷识别方法, refAbstract=null), Reference(id=1203362955053215981, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=56, issue=12, pageStart=43, pageEnd=53, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=ZHOUYuanxiang, LIYongyin, CHENJianning, journalName=Insulating Materials, refType=null, unstructuredReference=ZHOUYuanxiang, LIYongyin, CHENJianning, et al. Defect re-cognition method of oil-paper insulation based on information fusion of PRPD spectrum and dissolved gas data[J].Insulating Materials,2023,56(12):43-53., articleTitle=Defect re-cognition method of oil-paper insulation based on information fusion of PRPD spectrum and dissolved gas data, refAbstract=null), Reference(id=1203362955145490674, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2024, volume=57, issue=3, pageStart=96, pageEnd=101, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=张慧娟, 齐金平, 李鸿伟, journalName=绝缘材料, refType=null, unstructuredReference=张慧娟,齐金平,李鸿伟,等.基于隐马尔科夫模型的牵引变压器油纸绝缘剩余寿命预测[J].绝缘材料,2024,57(3):96-101., articleTitle=基于隐马尔科夫模型的牵引变压器油纸绝缘剩余寿命预测, refAbstract=null), Reference(id=1203362955267125494, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2024, volume=57, issue=3, pageStart=96, pageEnd=101, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=ZHANGHuijuan, QIJinping, LIHongwei, journalName=Insulating Materials, refType=null, unstructuredReference=ZHANGHuijuan, QIJinping, LIHongwei, et al. Residual life prediction of oil-paper insulation for traction transformers based on hidden Markov model[J]. Insulating Materials,2024,57(3):96-101., articleTitle=Residual life prediction of oil-paper insulation for traction transformers based on hidden Markov model, refAbstract=null), Reference(id=1203362955355205883, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=8, issue=null, pageStart=1360, pageEnd=1367, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=LIUJiaxin, ZHAOZijian, ZHONGYuanchen, journalName=Energy Reports, refType=null, unstructuredReference=LIUJiaxin, ZHAOZijian, ZHONGYuanchen, et al. Prediction of the dissolved gas concentration in power transformer oil based on sarima model[J]. Energy Reports,2022,8:1360-1367., articleTitle=Prediction of the dissolved gas concentration in power transformer oil based on sarima model, refAbstract=null), Reference(id=1203362955434897661, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2019, volume=10, issue=6, pageStart=1313, pageEnd=1322, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=LUShixiang, LINGuoying, QUEHuakun, journalName=International Journal of Machine Learning and Cybernetics, refType=null, unstructuredReference=LUShixiang, LINGuoying, QUEHuakun, et al. Grey relational analysis using gaussian process regression method for dissolved gas concentration prediction[J]. International Journal of Machine Learning and Cybernetics,2019,10(6):1313-1322., articleTitle=Grey relational analysis using gaussian process regression method for dissolved gas concentration prediction, refAbstract=null), Reference(id=1203362955514589441, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2017, volume=37, issue=12, pageStart=3643, pageEnd=3653,3694, url=null, language=null, rfNumber=[5], rfOrder=6, authorNames=肖怀硕, 李清泉, 施亚林, journalName=中国电机工程学报, refType=null, unstructuredReference=肖怀硕,李清泉,施亚林,等.灰色理论-变分模态分解和NSGA-Ⅱ优化的支持向量机在变压器油中气体预测中的应用[J].中国电机工程学报,2017,37(12):3643-3653,3694., articleTitle=灰色理论-变分模态分解和NSGA-Ⅱ优化的支持向量机在变压器油中气体预测中的应用, refAbstract=null), Reference(id=1203362955602669830, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2017, volume=37, issue=12, pageStart=3643, pageEnd=3653,3694, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=XIAOHuaishuo, LIQingquan, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=XIAOHuaishuo, LIQingquan. SHI Yalin,et al. Prediction of dissolved gases in oil for transformer based on grey theory-variational mode decomposition and support vector machine improved by NSGA-II[J]. Proceedings of the CSEE,2017,37(12):3643-3653,3694., articleTitle=SHI Yalin,, refAbstract=null), Reference(id=1203362955690750217, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2020, volume=46, issue=7, pageStart=2530, pageEnd=2538, url=null, language=null, rfNumber=[6], rfOrder=8, authorNames=黄新波, 蒋卫涛, 朱永灿, journalName=高电压技术, refType=null, unstructuredReference=黄新波,蒋卫涛,朱永灿,等.基于时间序列和支持向量机的变压器故障预测[J].高电压技术,2020,46(7):2530-2538., articleTitle=基于时间序列和支持向量机的变压器故障预测, refAbstract=null), Reference(id=1203362955808190736, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2020, volume=46, issue=7, pageStart=2530, pageEnd=2538, url=null, language=null, rfNumber=[6], rfOrder=9, authorNames=HUANGXinbo, JIANGWeitao, ZHUYongcan, journalName=High Voltage Engineering, refType=null, unstructuredReference=HUANGXinbo, JIANGWeitao, ZHUYongcan, et al. Transfor-mer fault prediction based on time series and support vector machine[J]. High Voltage Engineering,2020,46(7):2530-2538., articleTitle=Transfor-mer fault prediction based on time series and support vector machine, refAbstract=null), Reference(id=1203362955950797076, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2019, volume=134, issue=null, pageStart=93, pageEnd=101, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=SPEISERJ L, MILLERM E, TOOZEJ, journalName=Expert Systems with Applications, refType=null, unstructuredReference=SPEISERJ L,MILLERM E, TOOZEJ, et al. A comparison of random forest variable selection methods for classification prediction modeling[J]. Expert Systems with Applications,2019,134:93-101., articleTitle=A comparison of random forest variable selection methods for classification prediction modeling, refAbstract=null), Reference(id=1203362956068237593, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2018, volume=33, issue=19, pageStart=4594, pageEnd=4602, url=null, language=null, rfNumber=[8], rfOrder=11, authorNames=李恩文, 王力农, 宋斌, journalName=电工技术学报, refType=null, unstructuredReference=李恩文,王力农,宋斌,等.基于改进模糊聚类算法的变压器油色谱分析[J].电工技术学报,2018,33(19):4594-4602., articleTitle=基于改进模糊聚类算法的变压器油色谱分析, refAbstract=null), Reference(id=1203362956189872414, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2018, volume=33, issue=19, pageStart=4594, pageEnd=4602, url=null, language=null, rfNumber=[8], rfOrder=12, authorNames=LIEnwen, WANGLinong, SONGBin, journalName=Transactions of China Electrotechnical Society, refType=null, unstructuredReference=LIEnwen, WANGLinong, SONGBin, et al. Analysis of transformer oil chromatography based on improved fuzzy clustering algorithm[J]. Transactions of China Electrotechnical Society, 2018,33(19):4594-4602., articleTitle=Analysis of transformer oil chromatography based on improved fuzzy clustering algorithm, refAbstract=null), Reference(id=1203362956273758499, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=217, issue=null, pageStart=109176, pageEnd=null, url=null, language=null, rfNumber=[9], rfOrder=13, authorNames=ZENGWeiliang, CAOYunfei, FENGLutao, journalName=Electric Power Systems Research, refType=null, unstructuredReference=ZENGWeiliang, CAOYunfei, FENGLutao, et al. Hybrid CEEMDAN-DBN-ELM for online DGA serials and transformer status forecasting[J]. Electric Power Systems Research,2023,217:109176., articleTitle=Hybrid CEEMDAN-DBN-ELM for online DGA serials and transformer status forecasting, refAbstract=null), Reference(id=1203362956391199016, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2019, volume=39, issue=13, pageStart=3998, pageEnd=4008, url=null, language=null, rfNumber=[10], rfOrder=14, authorNames=刘云鹏, 许自强, 董王英, journalName=中国电机工程学报, refType=null, unstructuredReference=刘云鹏,许自强,董王英,等.基于经验模态分解和长短期记忆神经网络的变压器油中溶解气体浓度预测方法[J].中国电机工程学报,2019,39(13):3998-4008., articleTitle=基于经验模态分解和长短期记忆神经网络的变压器油中溶解气体浓度预测方法, refAbstract=null), Reference(id=1203362956483473708, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2019, volume=39, issue=13, pageStart=3998, pageEnd=4008, url=null, language=null, rfNumber=[10], rfOrder=15, authorNames=LIUYunpeng, XUZiqiang, DONGWangying, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=LIUYunpeng, XUZiqiang, DONGWangying, et al. Concentration prediction of dissolved gases in transformer oil based on empirical mode decomposition and long short-term memory neural networks[J]. Proceedings of the CSEE,2019,39(13):3998-4008., articleTitle=Concentration prediction of dissolved gases in transformer oil based on empirical mode decomposition and long short-term memory neural networks, refAbstract=null), Reference(id=1203362956546388271, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=38, issue=1, pageStart=117, pageEnd=130, url=null, language=null, rfNumber=[11], rfOrder=16, authorNames=杨童亮, 胡东, 唐超, journalName=电工技术学报, refType=null, unstructuredReference=杨童亮,胡东,唐超,等.基于SMA-VMD-GRU模型的变压器油中溶解气体含量预测[J].电工技术学报,2023,38(1):117-130., articleTitle=基于SMA-VMD-GRU模型的变压器油中溶解气体含量预测, refAbstract=null), Reference(id=1203362956621885746, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=38, issue=1, pageStart=117, pageEnd=130, url=null, language=null, rfNumber=[11], rfOrder=17, authorNames=YANGTongliang, HUDong, TANGChao, journalName=Transactions of China Electrotechnical Society, refType=null, unstructuredReference=YANGTongliang, HUDong, TANGChao, et al. Prediction of dissolved gas content in transformer oil based on SMA-VMD-GRU model[J]. Transactions of China Electrotechnical Society, 2023,38(1):117-130., articleTitle=Prediction of dissolved gas content in transformer oil based on SMA-VMD-GRU model, refAbstract=null), Reference(id=1203362956693188919, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2018, volume=32, issue=10, pageStart=87, pageEnd=94, url=null, language=null, rfNumber=[12], rfOrder=18, authorNames=吴汶倢, 何怡刚, 段嘉珺, journalName=电子测量与仪器学报, refType=null, unstructuredReference=吴汶倢,何怡刚,段嘉珺,等.油中溶解气体的灰色模型残差修正融合算法[J].电子测量与仪器学报,2018,32(10):87-94., articleTitle=油中溶解气体的灰色模型残差修正融合算法, refAbstract=null), Reference(id=1203362956772880699, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2018, volume=32, issue=10, pageStart=87, pageEnd=94, url=null, language=null, rfNumber=[12], rfOrder=19, authorNames=WUWenjie, HEYigang, DUANJiajun, journalName=Journal of Electronic Measurement and Instrumentation, refType=null, unstructuredReference=WUWenjie, HEYigang, DUANJiajun, et al. Residual-modified fusion algorithm of dissolved gasin transformer oil based on grey model[J]. Journal of Electronic Measurement and Instrumentation,2018,32(10):87-94., articleTitle=Residual-modified fusion algorithm of dissolved gasin transformer oil based on grey model, refAbstract=null), Reference(id=1203362956881932605, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2024, volume=50, issue=1, pageStart=210, pageEnd=220, url=null, language=null, rfNumber=[13], rfOrder=20, authorNames=马宏忠, 肖雨松, 孙永腾, journalName=高电压技术, refType=null, unstructuredReference=马宏忠,肖雨松,孙永腾,等.基于ICEEMDAN和时变权重集成预测模型的变压器油中溶解气体含量预测[J].高电压技术,2024,50(1):210-220., articleTitle=基于ICEEMDAN和时变权重集成预测模型的变压器油中溶解气体含量预测, refAbstract=null), Reference(id=1203362956953235777, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2024, volume=50, issue=1, pageStart=210, pageEnd=220, url=null, language=null, rfNumber=[13], rfOrder=21, authorNames=MAHongzhong, XIAOYusong, SUNYongteng, journalName=High Voltage Engineering, refType=null, unstructuredReference=MAHongzhong, XIAOYusong, SUNYongteng, et al. Prediction of dissolved gas concentration in transformer oil based on ICEEMDAN and time-varying weight integrated prediction model[J]. High Voltage Engineering,2024,50(1):210-220., articleTitle=Prediction of dissolved gas concentration in transformer oil based on ICEEMDAN and time-varying weight integrated prediction model, refAbstract=null), Reference(id=1203362957037121862, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2021, volume=70, issue=null, pageStart=1, pageEnd=11, url=null, language=null, rfNumber=[14], rfOrder=22, authorNames=MAXin, HUHao, SHANGYizi, journalName=IEEE Transactions on Instrumentation and Measurement, refType=null, unstructuredReference=MAXin, HUHao, SHANGYizi. A new method for transformer fault prediction based on multifeature enhancement and refined long short-term memory[J]. IEEE Transactions on Instrumentation and Measurement,2021,70:1-11., articleTitle=A new method for transformer fault prediction based on multifeature enhancement and refined long short-term memory, refAbstract=null), Reference(id=1203362957116813642, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=8, pageStart=2972, pageEnd=2981, url=null, language=null, rfNumber=[15], rfOrder=23, authorNames=刘展程, 王爽, 唐波, journalName=高电压技术, refType=null, unstructuredReference=刘展程,王爽,唐波.基于SSA-BiGRU-Attention模型的变压器油中溶解气体含量预测[J].高电压技术,2022,48(8):2972-2981., articleTitle=基于SSA-BiGRU-Attention模型的变压器油中溶解气体含量预测, refAbstract=null), Reference(id=1203362957192311119, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=8, pageStart=2972, pageEnd=2981, url=null, language=null, rfNumber=[15], rfOrder=24, authorNames=LIUZhancheng, WANGShuang, TANGBo, journalName=High Voltage Engineering, refType=null, unstructuredReference=LIUZhancheng, WANGShuang, TANGBo. Prediction of dissolved gas content in transformer oil based on SSA-BiGRU-Attention model [J]. High Voltage Engineering,2022,48(8):2972-2981., articleTitle=Prediction of dissolved gas content in transformer oil based on SSA-BiGRU-Attention model, refAbstract=null), Reference(id=1203362957272002899, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=37, issue=5, pageStart=1242, pageEnd=1251, url=null, language=null, rfNumber=[16], rfOrder=25, authorNames=赵洋, 王瀚墨, 康丽, journalName=电工技术学报, refType=null, unstructuredReference=赵洋,王瀚墨,康丽,等.基于时间卷积网络的短期电力负荷预测[J].电工技术学报,2022,37(5):1242-1251., articleTitle=基于时间卷积网络的短期电力负荷预测, refAbstract=null), Reference(id=1203362957339111767, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=37, issue=5, pageStart=1242, pageEnd=1251, url=null, language=null, rfNumber=[16], rfOrder=26, authorNames=ZHAOYang, WANGHanmo, KANGLi, journalName=Transactions of China Electrotechnical Society, refType=null, unstructuredReference=ZHAOYang, WANGHanmo, KANGLi, et al. Temporal convolution network-based short-term electrical load forecasting[J]. Transactions of China Electrotechnical Society,2022,37(5):1242-1251., articleTitle=Temporal convolution network-based short-term electrical load forecasting, refAbstract=null), Reference(id=1203362957402026331, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=45, issue=1, pageStart=35, pageEnd=68, url=null, language=null, rfNumber=[17], rfOrder=27, authorNames=吴博, 梁循, 张树森, journalName=计算机学报, refType=null, unstructuredReference=吴博,梁循,张树森,等.图神经网络前沿进展与应用[J].计算机学报,2022,45(1):35-68., articleTitle=图神经网络前沿进展与应用, refAbstract=null), Reference(id=1203362957477523808, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=45, issue=1, pageStart=35, pageEnd=68, url=null, language=null, rfNumber=[17], rfOrder=28, authorNames=WUBo, LIANGXun, ZHANGShusen, journalName=Chinese Journal of Computer, refType=null, unstructuredReference=WUBo, LIANGXun, ZHANGShusen, et al. Advances and applications in graph neural network[J]. Chinese Journal of Computer,2022,45(1):35-68., articleTitle=Advances and applications in graph neural network, refAbstract=null), Reference(id=1203362957586575715, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=35, issue=10, pageStart=10748, pageEnd=10761, url=null, language=null, rfNumber=[18], rfOrder=29, authorNames=CHENLing, CHENDonghui, SHANGZongjiang, journalName=IEEE Transactions on Knowledge and Data Engineering, refType=null, unstructuredReference=CHENLing, CHENDonghui, SHANGZongjiang, et al. Multi-scale adaptive graph neural network for multivariate time series forecasting[J]. IEEE Transactions on Knowledge and Data Engineering,2023,35(10):10748-10761., articleTitle=Multi-scale adaptive graph neural network for multivariate time series forecasting, refAbstract=null), Reference(id=1203362957691433319, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=11, issue=4, pageStart=1100, pageEnd=1114, url=null, language=null, rfNumber=[19], rfOrder=30, authorNames=LIAOWenlong, WANGShouxiang, BAK-JENSENB, journalName=Journal of Modern Power Systems and Clean Energy, refType=null, unstructuredReference=LIAOWenlong, WANGShouxiang, BAK-JENSENB, et al. Ultra-short-term interval prediction of wind power based on graph neural network and improved bootstrap technique[J]. Journal of Modern Power Systems and Clean Energy,2023,11(4):1100-1114., articleTitle=Ultra-short-term interval prediction of wind power based on graph neural network and improved bootstrap technique, refAbstract=null), Reference(id=1203362957796290923, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=51, issue=2, pageStart=132, pageEnd=140, url=null, language=null, rfNumber=[20], rfOrder=31, authorNames=欧阳福莲, 王俊, 周杭霞, journalName=电力系统保护与控制, refType=null, unstructuredReference=欧阳福莲,王俊,周杭霞.基于改进迁移学习和多尺度CNN-BiLSTM-Attention的短期电力负荷预测方法[J].电力系统保护与控制,2023,51(2):132-140., articleTitle=基于改进迁移学习和多尺度CNN-BiLSTM-Attention的短期电力负荷预测方法, refAbstract=null), Reference(id=1203362957859205485, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=51, issue=2, pageStart=132, pageEnd=140, url=null, language=null, rfNumber=[20], rfOrder=32, authorNames=OUYANGFulian, WANGJun, ZHOUHangxia, journalName=Power System Protection and Control, refType=null, unstructuredReference=OUYANGFulian, WANGJun, ZHOUHangxia. Short-term power load forecasting method based on improved hierarchical transfer learning and multi-scale CNN-BiLSTM-Attention[J].Power System Protection and Control,2023,51(2):132-140., articleTitle=Short-term power load forecasting method based on improved hierarchical transfer learning and multi-scale CNN-BiLSTM-Attention, refAbstract=null), Reference(id=1203362957913731440, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2021, volume=43, issue=2, pageStart=652, pageEnd=662, url=null, language=null, rfNumber=[21], rfOrder=33, authorNames=GAOShanghua, CHENGMingming, ZHAOKai, journalName=IEEE Transactions on Pattern Analysis and Machine Intelligence, refType=null, unstructuredReference=GAOShanghua, CHENGMingming, ZHAOKai, et al. Res2Net: a new multi-scale backbone architecture[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2021,43(2):652-662., articleTitle=Res2Net: a new multi-scale backbone architecture, refAbstract=null), Reference(id=1203362958014394741, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=21, pageStart=8496, pageEnd=8507, url=null, language=null, rfNumber=[22], rfOrder=34, authorNames=孙亦皓, 刘浩, 胡天宇, journalName=中国电机工程报, refType=null, unstructuredReference=孙亦皓,刘浩,胡天宇,等.基于时空关联特征与GCN-FEDformer的风速短期预测方法[J].中国电机工程报,2024,44(21):8496-8507., articleTitle=基于时空关联特征与GCN-FEDformer的风速短期预测方法, refAbstract=null), Reference(id=1203362958119252346, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=21, pageStart=8496, pageEnd=8507, url=null, language=null, rfNumber=[22], rfOrder=35, authorNames=SUNYihao, LIUHao, HUTianyu, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=SUNYihao, LIUHao, HUTianyu, et al. Short-term wind speed forecasting based on GCN and FEDformer[J]. Proceedings of the CSEE,2024,44(21):8496-8507., articleTitle=Short-term wind speed forecasting based on GCN and FEDformer, refAbstract=null), Reference(id=1203362959234937215, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=38, issue=17, pageStart=4683, pageEnd=4700, url=null, language=null, rfNumber=[23], rfOrder=36, authorNames=陈光宇, 袁文辉, 徐晓春, journalName=电工技术学报, refType=null, unstructuredReference=陈光宇,袁文辉,徐晓春,等.基于残差图卷积深度网络的电网无功储备需求快速计算方法[J].电工技术学报,2023,38(17):4683-4700., articleTitle=基于残差图卷积深度网络的电网无功储备需求快速计算方法, refAbstract=null), Reference(id=1203362959352377728, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=38, issue=17, pageStart=4683, pageEnd=4700, url=null, language=null, rfNumber=[23], rfOrder=37, authorNames=CHENGuangyu, YUANWenhui, XUXiaochun, journalName=Transactions of China Electrotechnical Society, refType=null, unstructuredReference=CHENGuangyu, YUANWenhui, XUXiaochun, et al. Fast calculation method for grid reactive power reserve demand based on residual graph convolutional deep network[J]. Transactions of China Electrotechnical Society,2023,38(17):4683-4700., articleTitle=Fast calculation method for grid reactive power reserve demand based on residual graph convolutional deep network, refAbstract=null), Reference(id=1203362959448846723, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2022, volume=200, issue=null, pageStart=751, pageEnd=766, url=null, language=null, rfNumber=[24], rfOrder=38, authorNames=ZHANJun, WUChengkun, YANGCanqun, journalName=Renewable Energy, refType=null, unstructuredReference=ZHANJun, WUChengkun, YANGCanqun, et al. Condition monitoring of wind turbines based on spatial-temporal feature aggregation networks[J]. Renewable Energy,2022,200:751-766., articleTitle=Condition monitoring of wind turbines based on spatial-temporal feature aggregation networks, refAbstract=null), Reference(id=1203362959536927109, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=38, issue=4, pageStart=104, pageEnd=112, url=null, language=null, rfNumber=[25], rfOrder=39, authorNames=王辉东, 姚海燕, 郭强, journalName=电力科学与技术学报, refType=null, unstructuredReference=王辉东,姚海燕,郭强,等.基于多尺度卷积神经网络的变压器故障诊断方法[J].电力科学与技术学报,2023,38(4):104-112., articleTitle=基于多尺度卷积神经网络的变压器故障诊断方法, refAbstract=null), Reference(id=1203362959729865096, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, doi=null, pmid=null, pmcid=null, year=2023, volume=38, issue=4, pageStart=104, pageEnd=112, url=null, language=null, rfNumber=[25], rfOrder=40, authorNames=WANGHuidong, YAOHaiyan, GUOQiang, journalName=Journal of Electric Power Science and Technology, refType=null, unstructuredReference=WANGHuidong, YAOHaiyan, GUOQiang, et al. A transformer fault diagnosis method based on multiscale 1DCNN[J]. Journal of Electric Power Science and Technology,2023,38(4):104-112., articleTitle=A transformer fault diagnosis method based on multiscale 1DCNN, refAbstract=null)], funds=[Fund(id=1203362954742837473, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, awardId=2023YFB2408203, language=CN, fundingSource=国家重点研发计划项目(2023YFB2408203), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1203362943615349508, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943623738117, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China), AuthorCompanyExt(id=1203362943632126726, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943615349508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.华北电力大学 电力工程系,河北 保定 071003)]), AuthorCompany(id=1203362943753761555, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943762150164, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943753761555, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. State Grid Baoding Power Supply Company, Baoding 071066, China), AuthorCompanyExt(id=1203362943774733076, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943753761555, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.国网保定供电公司,河北 保定 071066)]), AuthorCompany(id=1203362943887979291, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, xref=null, ext=[AuthorCompanyExt(id=1203362943963476764, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943887979291, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3. State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (NCEPU), Beijing 102206, China), AuthorCompanyExt(id=1203362943976059678, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, companyId=1203362943887979291, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.新能源电力系统全国重点实验室(华北电力大学),北京 102206)])], figs=[ArticleFig(id=1203362948900171805, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.1, caption=Comparison on the structure of bottleneck residual and Res2Net, figureFileSmall=lTqbxNHxMcMVQeROeAwefA==, figureFileBig=ZgFuP8pR07O6zfab8WSelQ==, tableContent=null), ArticleFig(id=1203362948996640802, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图1, caption=瓶颈残差与Res2Net结构对比图

(a) 瓶颈残差 (b) Res2Net

, figureFileSmall=lTqbxNHxMcMVQeROeAwefA==, figureFileBig=ZgFuP8pR07O6zfab8WSelQ==, tableContent=null), ArticleFig(id=1203362949135052840, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.2, caption=Schematic diagram of graph convolution process, figureFileSmall=k4wWpxFlJzo2UjXok+bZ0A==, figureFileBig=XxdauMt2RmUAKtaBw1lssg==, tableContent=null), ArticleFig(id=1203362950351401007, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图2, caption=图卷积过程示意图, figureFileSmall=k4wWpxFlJzo2UjXok+bZ0A==, figureFileBig=XxdauMt2RmUAKtaBw1lssg==, tableContent=null), ArticleFig(id=1203362950435287089, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.3, caption=Structure diagram of multi-scale spatio-temporal coupled information network, figureFileSmall=FsIYaAXBkunTnHpi3z7soA==, figureFileBig=i/rDcDNsnRB0ywgozL2BNQ==, tableContent=null), ArticleFig(id=1203362950548533305, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图3, caption=多尺度时空信息耦合网络结构图, figureFileSmall=FsIYaAXBkunTnHpi3z7soA==, figureFileBig=i/rDcDNsnRB0ywgozL2BNQ==, tableContent=null), ArticleFig(id=1203362950649196606, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.4, caption=Multivariate prediction flowchart based on spatio-temporal information fusion, figureFileSmall=ydhfKPi5buHbZnt7qNWWIg==, figureFileBig=R8bQH3hqJlg40keHixiWtw==, tableContent=null), ArticleFig(id=1203362950766637124, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图4, caption=基于时空信息融合的多元预测流程图, figureFileSmall=ydhfKPi5buHbZnt7qNWWIg==, figureFileBig=R8bQH3hqJlg40keHixiWtw==, tableContent=null), ArticleFig(id=1203362950934409293, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.5, caption=Transformer characteristic gas detection data chart, figureFileSmall=OFMjhqfdqNdR3oHKivIvoQ==, figureFileBig=BC5yfdvbelbi1PGlj9bDcA==, tableContent=null), ArticleFig(id=1203362951072821330, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图5, caption=变压器特征气体预测数据图, figureFileSmall=OFMjhqfdqNdR3oHKivIvoQ==, figureFileBig=BC5yfdvbelbi1PGlj9bDcA==, tableContent=null), ArticleFig(id=1203362951207039067, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.6, caption=Comparison on loss function values of models with different convolutional scales, figureFileSmall=EGzbxwM2mRXNXkrom8PY3w==, figureFileBig=NWY7WJbCAU1U3AY/ZqNjqg==, tableContent=null), ArticleFig(id=1203362951290925149, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图6, caption=不同卷积尺度数量的模型损失函数值对比图, figureFileSmall=EGzbxwM2mRXNXkrom8PY3w==, figureFileBig=NWY7WJbCAU1U3AY/ZqNjqg==, tableContent=null), ArticleFig(id=1203362951391588453, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.7, caption=Mutual information between feature variables and adjacency matrix, figureFileSmall=pzoc/jrUJ2b6Sa36JEBYEg==, figureFileBig=d2K7y9PUxO34XkBVsKDoEQ==, tableContent=null), ArticleFig(id=1203362951500640363, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图7, caption=特征变量间互信息与邻接矩阵, figureFileSmall=pzoc/jrUJ2b6Sa36JEBYEg==, figureFileBig=d2K7y9PUxO34XkBVsKDoEQ==, tableContent=null), ArticleFig(id=1203362951622275185, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.8, caption=RMSE values corresponding to different K values, figureFileSmall=ac1jONnop5lodAKXSLHsTQ==, figureFileBig=8TEvWm+bi+GDnIWiQ4fvuA==, tableContent=null), ArticleFig(id=1203362951689384056, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图8, caption=不同K值对应的RMSE, figureFileSmall=ac1jONnop5lodAKXSLHsTQ==, figureFileBig=8TEvWm+bi+GDnIWiQ4fvuA==, tableContent=null), ArticleFig(id=1203362951764881535, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.9, caption=Absolute error distribution map of GCN at different layers, figureFileSmall=NRqq4oEzG6SJFV9MsOQ37g==, figureFileBig=CsAFe80MuTDLu+Og+KVrdw==, tableContent=null), ArticleFig(id=1203362951924265093, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图9, caption=不同层GCN绝对误差分布图, figureFileSmall=NRqq4oEzG6SJFV9MsOQ37g==, figureFileBig=CsAFe80MuTDLu+Og+KVrdw==, tableContent=null), ArticleFig(id=1203362952024928398, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.10, caption=Prediction loss distribution map of different time windows, figureFileSmall=1lj96Rfsnxc1VHEzZ5VPBw==, figureFileBig=KQTMY67zIKWOVjIii7ABFg==, tableContent=null), ArticleFig(id=1203362952175923348, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图10, caption=不同时间窗预测损失分布图, figureFileSmall=1lj96Rfsnxc1VHEzZ5VPBw==, figureFileBig=KQTMY67zIKWOVjIii7ABFg==, tableContent=null), ArticleFig(id=1203362952406610080, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Fig.11, caption=Comparison on prediction performance of different spatiotemporal feature extraction models, figureFileSmall=ccIYWeLX+5KPXfEvARZErA==, figureFileBig=nPqqEpuFlg4b1pclnk2NZg==, tableContent=null), ArticleFig(id=1203362952570187948, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=图11, caption=不同时空特征提取模型预测效果对比图, figureFileSmall=ccIYWeLX+5KPXfEvARZErA==, figureFileBig=nPqqEpuFlg4b1pclnk2NZg==, tableContent=null), ArticleFig(id=1203362952666656946, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Table 1, caption=Comparison on the effects of different prediction methods, figureFileSmall=null, figureFileBig=null, tableContent=
预测模型RMSE/(μL/L)R2I/%
Res2Net-GCN-TCN0.092 40.957 598.68
SMA-VMD-GRU0.176 40.866 993.74
LSTM0.472 50.635 691.17
GRU0.456 00.651 491.33
), ArticleFig(id=1203362952763125943, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=表1, caption=

不同预测模型效果对比

, figureFileSmall=null, figureFileBig=null, tableContent=
预测模型RMSE/(μL/L)R2I/%
Res2Net-GCN-TCN0.092 40.957 598.68
SMA-VMD-GRU0.176 40.866 993.74
LSTM0.472 50.635 691.17
GRU0.456 00.651 491.33
), ArticleFig(id=1203362952872177851, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Table 2, caption=Results of ablation experiments, figureFileSmall=null, figureFileBig=null, tableContent=
模型RMSE/(μL/L)MAPE/%R2I/%
T0.257 26.810.712 593.19
R-T0.204 65.660.828 394.34
G-T0.164 63.750.884 996.25
R-G-T0.128 31.460.946 298.54
), ArticleFig(id=1203362953027367109, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=表2, caption=

消融实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型RMSE/(μL/L)MAPE/%R2I/%
T0.257 26.810.712 593.19
R-T0.204 65.660.828 394.34
G-T0.164 63.750.884 996.25
R-G-T0.128 31.460.946 298.54
), ArticleFig(id=1203362953123836108, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=EN, label=Table 3, caption=Comparison on the effects of different prediction methods, figureFileSmall=null, figureFileBig=null, tableContent=
预测模型RMSE/(μL/L)R2I/%t/s
Res2Net-GCN-TCN0.092 40.957 598.6828
Res2Net-GCN-LSTM0.172 00.875 593.1155
Res2Net-GCN-transformer0.163 70.859 694.03112
), ArticleFig(id=1203362953203527890, tenantId=1146029695717560320, journalId=1149653034449285133, articleId=1203281641063035100, language=CN, label=表3, caption=

不同预测方法效果对比

, figureFileSmall=null, figureFileBig=null, tableContent=
预测模型RMSE/(μL/L)R2I/%t/s
Res2Net-GCN-TCN0.092 40.957 598.6828
Res2Net-GCN-LSTM0.172 00.875 593.1155
Res2Net-GCN-transformer0.163 70.859 694.03112
)], attaches=null, journal=Journal(id=1146437660891000859, delFlag=0, nameCn=绝缘材料, nameEn=Insulating Materials, nameHistory1=null, nameHistory2=null, issn=1009-9239, eissn=null, cn=45-1287/TM, coden=null, periodic=0, language=CN, oaType=是, ccby=CC BY-NC-ND, 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=To3JhdgsOj1Q7eh5WnUN1g==, journalPrice=null, startedYear=null, abbrevIsoEn=null, journalRemark=null, publicationField=null, createdTime=1751261638140, updatedTime=1761735740591, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=I, firstLetterEn=I, subjectCode=Natural Sciences, subjectName=Natural Sciences, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=To3JhdgsOj1Q7eh5WnUN1g==, picEn=O7zoVkJ9hJoi8iGUSQYTxg==, jcr=null, cjcr=null, exts=[JournalExt(id=1190369230903152828, language=CN, name=绝缘材料, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1761735740638, updatedTime=1761735740638, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://jyct.cbpt.cnki.net/EditorDN/index.aspx?t=1, submissionEditorUrl=https://jyct.cbpt.cnki.net/EditorDN/index.aspx?t=3, submissionReviewUrl=https://jyct.cbpt.cnki.net/EditorDN/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1190369230945095869, language=EN, name=Insulating Materials, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1761735740648, updatedTime=1761735740648, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://jyct.cbpt.cnki.net/EditorDN/index.aspx?t=1, submissionEditorUrl=https://jyct.cbpt.cnki.net/EditorDN/index.aspx?t=3, submissionReviewUrl=https://jyct.cbpt.cnki.net/EditorDN/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1149653034449285133, websiteList=[Website(id=1189939819993166814, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1149653034449285133, 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/jycl/CN, language=CN, createTime=1761633361099, createBy=18614031015, updateTime=1761633401425, updateBy=18614031015, name=绝缘材料-中文, tplId=1146099689490845704, title=绝缘材料, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1189940292275991527, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=articleTextType, value=kx, createTime=1761633473700, updateTime=1761633473700, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292250825700, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=banner, value=null, createTime=1761633473694, updateTime=1761633473694, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292301157354, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=grayFlag, value=0, createTime=1761633473706, updateTime=1761633473706, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292238242787, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=logo, value=https://castjournals.cast.org.cn/joweb/jycl/CN/file/pic?fileId=MyqZAHzZT6tMetr2hjDKLQ==, createTime=1761633473691, updateTime=1761633473691, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292322128876, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=minRunFlag, value=0, createTime=1761633473711, updateTime=1761633473711, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292267602918, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jycl/CN/file/pic, createTime=1761633473698, updateTime=1761633473698, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292309545963, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=silenceFlag, value=0, createTime=1761633473708, updateTime=1761633473708, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292259214309, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1761633473696, updateTime=1761633473696, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292284380136, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=themeColor, value=null, createTime=1761633473702, updateTime=1761633473702, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940292292768745, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939819993166814, code=themeStyle, value=null, createTime=1761633473704, updateTime=1761633473704, creator=18614031015, updator=18614031015)]), Website(id=1189939820060275680, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1149653034449285133, 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/jycl/EN, language=EN, createTime=1761633361115, createBy=18614031015, updateTime=1761633397117, updateBy=18614031015, name=绝缘材料-英文, tplId=1146101810881728533, title=Insulating Materials, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1189940323548722161, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=articleTextType, value=kx, createTime=1761633481156, updateTime=1761633481156, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323519362030, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=banner, value=null, createTime=1761633481149, updateTime=1761633481149, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323569693684, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=grayFlag, value=0, createTime=1761633481161, updateTime=1761633481161, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323506779117, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=logo, value=https://castjournals.cast.org.cn/joweb/jycl/EN/file/pic?fileId=MyqZAHzZT6tMetr2hjDKLQ==, createTime=1761633481146, updateTime=1761633481146, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323586470902, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=minRunFlag, value=0, createTime=1761633481165, updateTime=1761633481165, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323540333552, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jycl/EN/file/pic, createTime=1761633481154, updateTime=1761633481154, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323578082293, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=silenceFlag, value=0, createTime=1761633481163, updateTime=1761633481163, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323531944943, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1761633481152, updateTime=1761633481152, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323557110770, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=themeColor, value=null, createTime=1761633481158, updateTime=1761633481158, creator=18614031015, updator=18614031015), WebsiteProps(id=1189940323565499379, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189939820060275680, code=themeStyle, value=null, createTime=1761633481160, updateTime=1761633481160, creator=18614031015, updator=18614031015)])], journalTitle=绝缘材料, weixinUrl=null, journalUrl=https://www.insulation.org.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Insulating Materials, journalPhotoCn=To3JhdgsOj1Q7eh5WnUN1g==, journalPhotoEn=O7zoVkJ9hJoi8iGUSQYTxg==, journalFirstLetter=I, 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=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/jycl/CN/10.16790/j.cnki.1009-9239.im.2025.06.015, detailUrlEn=https://castjournals.cast.org.cn/joweb/jycl/EN/10.16790/j.cnki.1009-9239.im.2025.06.015, pdfUrlCn=https://castjournals.cast.org.cn/joweb/jycl/CN/PDF/10.16790/j.cnki.1009-9239.im.2025.06.015, pdfUrlEn=https://castjournals.cast.org.cn/joweb/jycl/EN/PDF/10.16790/j.cnki.1009-9239.im.2025.06.015, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
考虑时空耦合关系的电力变压器油中溶解气体多尺度融合预测方法
收藏切换
PDF下载
张倩倩 1 , 李敏 2 , 耿绍胜 1 , 王春鑫 1 , 谢军 1 , 谢庆 1, 3
绝缘材料 | 绝缘技术 2025,58(6): 122-130
收起
收藏切换
绝缘材料 | 绝缘技术 2025, 58(6): 122-130
考虑时空耦合关系的电力变压器油中溶解气体多尺度融合预测方法
全屏
张倩倩1, 李敏2, 耿绍胜1, 王春鑫1, 谢军1, 谢庆1, 3
作者信息
  • 1.华北电力大学 电力工程系,河北 保定 071003
  • 2.国网保定供电公司,河北 保定 071066
  • 3.新能源电力系统全国重点实验室(华北电力大学),北京 102206
  • 张倩倩(2000-),女(汉族),吉林敦化人,硕士生,主要从事变压器状态监测的研究工作;

通讯作者:

谢军(1988-),男(汉族),江苏江都人,副教授,博士,主要从事高电压试验技术、电力设备状态监测的研究工作。
Multi-scale fusion prediction method of dissolved gas in power transformer oil considering spatio-temporal coupling relationship
Qianqian ZHANG1, Min LI2, Shaosheng GENG1, Chunxin WANG1, Jun XIE1, Qing XIE1, 3
Affiliations
  • 1. Department of Electrical Engineering, North China Electric Power University, Baoding 071003, China
  • 2. State Grid Baoding Power Supply Company, Baoding 071066, China
  • 3. State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (NCEPU), Beijing 102206, China
出版时间: 2025-06-20 doi: 10.16790/j.cnki.1009-9239.im.2025.06.015
文章导航
收藏切换

对油中溶解气体的时空耦合关系进行多尺度挖掘,有助于提高油中溶解气体预测精度,为变压器运维决策提供可靠理论依据。为此,提出一种考虑时空耦合信息的变压器油中溶解气体多尺度融合预测方法。首先,利用Res2Net对油中溶解气体数据进行多尺度时间特征提取,捕捉特征气体不同频率的周期性时间特征信息。其次,通过计算互信息捕捉特征气体间隐性关系,以拓扑关系图的形式描述不同气体间关联性,并使用图卷积神经网络(GCN)提取空间信息特征。最后,将多尺度时间信息与空间信息进行融合拼接,采用时间卷积网路(TCN)对油中溶解气体进行预测,并使用某500 kV变压器油色谱在线监测数据对所提方法进行验证。结果表明:相比于传统预测方法,Res2Net-GCN-TCN模型可有效提高油中溶解气体含量预测精度,平均预测精度可达98.68%。

油中溶解气体预测  /  Res2Net  /  图卷积  /  时间卷积  /  时空信息融合

Multi-scale mining of the spatio-temporal coupling relationship of dissolved gases in oil is helpful to improve the prediction accuracy of dissolved gases in oil and provide a reliable theoretical basis for transformer operation and maintenance decisions. Thereby, a multi-scale fusion prediction method for dissolved gases in transformer oil considering spatio-temporal coupling information was proposed in this study. Firstly, the Res2Net was used to extract the multi-scale time characteristics of the dissolved gas data in oil, and the periodic time feature information of the characteristic gas under different frequencies was captured. Secondly, the implicit relationship between characteristic gases was captured by calculating mutual information, the correlation between different gases was described in the form of topological graphs, and the spatial information features were extracted by using graph convolutional neural network (GCN). Finally, multi-scale temporal information and spatial information were fused and spliced, and temporal convolution network (TCN) was used to predict the dissolved gas in oil. The proposed method was validated using online oil chromatography monitoring data from a 500 kV transformer. The results show that compared with the traditional prediction method, the Res2Net-GCN-TCN model can effectively improve the prediction accuracy of dissolved gas content in oil, and the average prediction accuracy is 98.68%.

dissolved gas prediction in oil  /  Res2Net  /  graph convolution  /  temporal convolution  /  spatio-temporal information fusion
张倩倩, 李敏, 耿绍胜, 王春鑫, 谢军, 谢庆. 考虑时空耦合关系的电力变压器油中溶解气体多尺度融合预测方法. 绝缘材料, 2025 , 58 (6) : 122 -130 . DOI: 10.16790/j.cnki.1009-9239.im.2025.06.015
Qianqian ZHANG, Min LI, Shaosheng GENG, Chunxin WANG, Jun XIE, Qing XIE. Multi-scale fusion prediction method of dissolved gas in power transformer oil considering spatio-temporal coupling relationship[J]. Insulating Materials, 2025 , 58 (6) : 122 -130 . DOI: 10.16790/j.cnki.1009-9239.im.2025.06.015
油中溶解气体分析(dissolved gas analysis,DGA)是监测变压器潜伏性故障和追踪故障发展趋势的重要手段之一[1]。准确预测油中溶解气体的变化趋势可为变压器健康评估和可靠运行时间分析提供重要参考[2],为变压器运维决策提供可靠理论依据,对保障电力系统安全稳定运行具有重要意义。
目前常见的DGA预测方法主要有自回归模型[3]、高斯过程[4]、灰色理论[5]等统计方法,支持向量机[6]、随机森林[7]、聚类算法[8]等机器学习方法,以及深度置信网络(DBN)[9]、长短期记忆网络(LSTM)[10]、门控循环单元(GRU)[11]等深度学习方法。统计方法模型较为简单,在短时预测中具有优势。吴汶倢等[12]提出了一种灰色模型残差修正融合算法,将自适应回归模型和马尔可夫模型相结合对灰色模型残差进行修正,改善了灰色模型对油中溶解气体长时间跨度预测精度低的问题,但由于油中溶解气体具有非线性、非平稳的特性,预测效果欠佳。机器学习方法在处理DGA非平稳性预测问题具有较大优势。马宏忠等[13]提出了一种麻雀算法优化支持向量机(SVM)和门控循环网络组合预测方法,在一定程度上改善了SVM模型过于依赖参数寻优方法的问题。
上述方法并不能深层次挖掘油中溶解气体的时间分布特性,因此很难全面表征油中溶解气体变化趋势。深度学习方法能够从大量数据中自动提取数据间的复杂特征关系,且具有良好的泛化性,被广泛应用于油中溶解气体预测中。LSTM采用循环网络结构,善于捕捉时序序列中长短期的依赖关系,提高了特征气体的预测精度[14],但LSTM同样存在超参数影响预测精度的问题。GRU在LSTM结构上进行了简化,减少了模型的复杂性和参数数量[15],但受限于循环网络结构和门控单元,其梯度消失问题和训练内存消耗大的问题仍不可忽视。时间卷积网络(TCN)不同于循环架构,其反向传播路径与时序方向不同,只取决于网络深度,避免了梯度消失等问题[16],其强大的并行处理能力显著提高了时序序列预测效率。
然而目前DGA深度学习预测方法多针对单一特征气体进行单尺度时间特征提取并预测,忽略了特征气体间的强关联性,且多倾向于提取一种突出且共享的时间特征依赖关系,导致预测效果和鲁棒性较差。图卷积神经网络(GCN)能够充分挖掘不同特征气体之间的空间依赖关系[17-19],可以改善单一特征气体预测误差较大的问题。
充分挖掘油中溶解气体含量间的时空特征信息有助于提高特征气体预测的准确性,Res2Net作为一种新型多尺度网络模型在目标检测方面展现了里程碑式效果,将其多尺度特征提取能力应用到时间序列预测中,可以改善特征提取模型倾向于提取突出且共享的时间特征的问题[20]
基于以上分析,本文提出了Res2Net-GCN-TCN变压器油中溶解气体含量预测方法,同时考虑了多尺度时间信息和特征气体间关联信息,将时间、空间信息拼接进行融合预测。首先,使用Res2Net网络对油中溶解气体数据进行多尺度时间特征提取,捕捉特征气体不同频率的周期性时间特征信息。其次,通过计算特征提取后的新特征变量间的互信息大小,构建油中溶解气体间拓扑关系图,并使用GCN挖掘不同特征气体间的空间关联信息。最后,对时空特征信息融合拼接,利用TCN对油中溶解气体进行融合预测,并使用现场采集的某500 kV变压器油色谱在线监测数据对所提方法有效性进行验证。
针对油中溶解气体时序序列的非线性、非平稳特性,采用Res2Net网络对特征气体时序特征进行多尺度挖掘。Res2Net网络将多尺度金字塔模型和瓶颈残差结构融合,利用类似分层残差的结构,实现高效且低成本的多尺度特征提取。瓶颈残差结构与Res2Net具体网络结构[21]图1所示。
对油中溶解气体特征数据进行1×1卷积运算后,将数据按照时间顺序平均分为s个子集,子集定义为xii1,2,3s图1s取4。除x1外,其他子集均带有一个3×3卷积核,定义为Ki。输入特征沿传输路径进行运算,每经过一个3×3卷积过程,时间特征感受野将增加一个尺度,实现多尺度特征提取,经该结构输出的特征yi式(1)所示。
yi=xiKixiKixi+yi-1i=1i=22<is
最后,将所有特征yi再进行1×1卷积计算,实现多尺度特征融合,进而实现高效低成本地获取油中溶解气体的多尺度时间特征。
图卷积计算的本质是不断对自身以及临邻近信息聚合重组的过程[22]。图卷积过程可有效实现特征气体间的关系提取,图卷积计算公式如式(2)所示,图卷积计算过程[23]图2所示。
Xl+1==D˜-1/2A˜D˜-1/2X(l)
式(2)中:A为图的邻接矩阵,在本文中为根据互信息计算得到拓扑结构构建的(n, n)矩阵,n为节点数;A˜=A+I,即考虑自身信息的邻接矩阵;D为度矩阵,表示与该点相连的节点数量;D˜=D+I,即考虑自身的度矩阵;X为特征矩阵信息。
单层GCN只能获取一阶邻近信息,为扩大节点对周围节点的感受野,本文采用2层GCN网络结构,公式如式(3)所示。
Z=A^σ(A^XlW0)W1
式(3)中:A^=D˜-1/2A˜D˜-1/2σ()为激活函数;W为神经网络的权重。
为提高油中溶解气体预测的精度,构建多尺度时空特征耦合模型,本节将时间信息、空间信息进行融合拼接,得到时空耦合信息。首先,将油中溶解气体含量数据X输入至Res2Net进行多尺度时间特征提取,得到时间特征X';然后将时间特征X'输入至GCN进行空间特征提取,得到时空特征X。将时间特征X'和时空特征X进行拼接[24]得到最终时空耦合信息X0。此举最大程度保留了油中溶解气体时间特征信息,同时引入特征气体的空间信息进行辅助预测,多尺度时空信息耦合网络结构如图3所示。
采用TCN对油中溶解气体进行融合预测,TCN的因果卷积结构使其不存在信息泄露问题。相对于传统的循环神经网络,TCN具有强大的并行计算能力,且无门限结构,可有效提升时序序列预测的准确性和快速性。具体运算公式如式(4)所示。
F(t)=i=0K-1wixt-(K-1)d+di
式(4)中:wi为卷积核权重;K为卷积核的大小;d为膨胀因子;x为输入序列特征;F为输出结果;t为时序序列中的位置。
使用多尺度时空信息耦合网络对DGA数据进行特征提取得到时空耦合信息,将该信息输入至TCN网络得到最终的融合预测结果。
本文通过深度挖掘时空特征信息实现油中溶解气体的融合预测,具体流程如图4所示。
(1)数据预处理
采集变压器油中溶解气体数据,包含H2CH4C2H2C2H4C2H6COCO2、总烃(TH)共8类气体在T时刻内的测量值,对数据进行清洗后做归一化处理。
由于油中溶解气体数据为多元时序序列,不具有图结构,因此需要对时序序列进行图结构G=(VE)构建,其中V表示节点的集合,E表示边的集合。本文将每个特征气体看作一个节点,通过计算互信息的值来衡量两个变量共享信息的多少,构建油中溶解气体图拓扑结构,具体公式如式(5)所示。
IX, Y=HX-HXY=xXyYp(x, y)log2p(x, y)p(x)p(y)
式(5)中:xy为某特征气体含量值;p(x, y)为联合概率分布;p(x)p(y)为边缘概率分布;IX, Y为特征气体XY之间互信息的值;HX为特征气体X的熵;HXY为已知特征气体Y条件下,特征气体X的条件熵。
选取K个与变量X互信值最大的变量Y作为邻接节点,构建邻接矩阵,为图卷积计算提供拓扑关系图。
(2)模型训练
将油中溶解气体数据划分为训练集与测试集,确定训练的网络超参数,使用Res2Net和GCN实现时空特征信息融合提取,然后将特征信息输入至TCN进行特征气体预测,目标为油中溶解气体预测值与实际值误差最小,保存最大迭代次数内误差最小时的模型参数,损失函数如式(6)所示。
NLoss=y-y^2+λL2
式(6)中:y^为模型预测值;NLoss为模型损失函数;λ为超参数;正则化项L2可防止模型过拟合。
(3)油中溶解气体含量预测
利用训练好的深度学习模型对DGA测试集进行预测,计算预测模型的预测误差,对模型预测效果进行评估。
本文采用的数据样本集为某500 kV变压器从2022-04-01到2023-09-20在线采集的油中溶解气体数据,采样周期为12 h,共1 078组数据,特征气体监测数据如图5所示。经数据清洗选取800组数据作为训练集,另200组数据作为测试集。
为定量分析本文所提模型预测效果,选用均方根误差(RMSE)、平均绝对百分比误差(MAPE)、决定系数(R2)、预测精度(I)对模型进行评价。RMSE用以直观衡量模型预测准确度,其值越小,模型的预测值与实际值的偏差越小;MAPE用以衡量预测误差相对实际值的大小,MAPE越小,模型的预测值与实际值的偏差越小;R2用以衡量模型拟合数据程度,R2越大,拟合效果越好;I是评估模型预测性能的综合指标,I越接近1,预测效果越好。各指标计算方法具体公式如式(7)所示。
RMSE=1Ni=1Nyi-y^i2MAPE=1Ni=1Nyi-y^iyi×100%R2=1-i=1Nyi-y^i2i=1Nyi-y¯i2I=1-1Ni=1Nyi-y^iyi×100%
式(7)中:N为数据样本量;yi为油中溶解气体含量实际值;y^i为油中溶解气体含量预测值;y¯i为油中溶解气体含量实际数据的平均值。
(1)卷积尺度数量的选择
在多尺度时间特征提取过程中,子集s数量越多,则卷积尺度数量越多,为进一步分析卷积尺度数量对预测准确度的影响,验证多尺度卷积对提高油中溶解气体含量预测精度的有效性,本文构建5个不同卷积尺度数量的预测模型,模型在训练过程中的损失函数值对比图如图6所示。
图6可知,当卷积尺度数量为4和5时,预测误差最小,为降低计算成本,选取卷积尺度数量为4。
(2)邻近节点数量的选择
根据互信息计算结果构建邻接拓扑关系,选取最大的K个互信值作为邻接节点,K2,3,,7。特征变量间互信息与邻接矩阵如图7所示,从图中可以看出K值越大拓扑关系越稠密,K值越小拓扑关系越稀疏。
为确定特征气体间拓扑关系疏密对预测效果的影响,设置不同K值,利用GCN-TCN模型对特征气体含量进行预测并计算评价指标,得到不同K值对应的RMSE值如图8所示。由图8可知,当K小于4时,模型预测效果随K的增大逐渐提升;当K为4时预测效果最好;当K大于4时,模型预测效果随之下降,因此相邻节点的数量选择为4。
(3)GCN层数的选择
为进一步分析GCN层数对模型稳定性的影响,采用小提琴图和箱线图来分析特征气体预测结果与实际值的绝对误差分布情况,结果如图9所示。由图9可知,图卷积层数为2时的预测效果最佳,即2层图卷积的感受野已满足捕捉特征气体间有效信息的需要,继续叠加网络层数可能会导致过拟合,影响预测效果。
(4)预测时间窗宽度的选择
为确定预测时间窗宽对特征气体预测效果的影响,分别采用不同时间窗宽对特征气体进行单步预测和多步预测,预测结果如图10所示。
图10(a)可知,在单步预测时超前12步(6天)对特征气体进行预测效果较好。由图10(b)可知,在多步预测(5步)时超前30步(15天)更为合适。
(1)模型性能对比分析
为验证本文所提时空融合预测模型的性能,将本文模型与文献[11]所提模型以及传统模型LSTM、GRU进行预测对比分析,结果如表1所示。由表1可知,在油中溶解气体预测中,本文所提预测模型性能最佳,相较于传统时序预测模型LSTM,预测精度可提升7.51%。
(2)模型有效性验证
为验证本文所提模型Res2Net-GCN-TCN各模块的有效性,采用消融实验进行验证。分别采用TCN、Res2Net-TCN、GCN-TCN、Res2Net-GCN-TCN (分别简称为T、R-T、G-T、R-G-T)对特征气体进行时序预测,预测结果如表2所示。
表2可知,R-T模型在基准模型T上增加多尺度时间特征提取模块,使各指标有一定的提升,但由于T模型本身具备一定的时间特征挖掘能力,仅增加多尺度时间特征提取模块时提升效果较小。G-T在基准模型T上引入空间特征提取模块,充分挖掘特征气体间关联信息,可有效提升预测效果。R-G-T模型对特征气体时空耦合关系进行了多尺度深度挖掘,从时间、空间两个维度对特征信息进行融合补充,使预测效果显著提升。相较于T模型,R-G-T预测精度提升了5.35%,证明了时间、空间融合预测的有效性。
(3)不同时空特征提取模型有效性验证
为验证本文所选用时间、空间特征提取模型的有效性,采用传统模型替换本文所提模型的一个模块,以此对比预测效果。采用多尺度卷积MCNN[25]替换Res2Net验证多尺度时间特征提取模块的有效性、采用卷积CNN替换GCN验证空间提取模块的有效性。采用MCNN-GCN-TCN、Res2Net-CNN-TCN和Res2Net-GCN-TCN模型对油中溶解气体进行预测,预测结果如图11所示。
图11可知,采用Res2Net-GCN-TCN模型的预测结果与实际结果最为相近,进一步证明了本文所提时空特征提取模型的有效性。
(4)不同预测模型有效性验证
为验证TCN预测模型的有效性,分别采用LSTM、transformer替换TCN预测模块,使用Res2Net-GCN-LSTM、Res2Net-GCN-transformer以及本文所提模型对特征气体变化趋势进行预测,结果如表3所示。由表3可知,得益于TCN模型强大的并行处理能力,本文所提模型可以实现高精度、低时间成本的油中溶解气体预测。
为及时准确地预测变压器油中溶解气体变化趋势,改善油中溶解气体时序序列非线性、非平稳特性影响预测精度的问题,提出了一种考虑时空耦合关系的油中溶解气体多尺度融合预测方法,并使用某500 kV变压器在线监测数据进行实例分析,得出以下结论:
(1)本文提出了一种多尺度时空信息融合预测方法,从时间、空间两个维度对DGA数据进行特征提取和融合,可进一步提升油中溶解气体的预测精度,对及时发现变压器内部潜伏性故障和制定运维决策具有重要意义。
(2)采用某变压器油中溶解气体在线监测数据进行实验对比分析,证明了Res2Net-GCN-TCN模型的预测性能优于传统预测模型,预测精度可提升7.51%,平均预测精度可达98.68%。
(3)通过消融实验和对比实验分析,结果表明本文选用的Res2Net、GCN进行时空特征提取的性能以及TCN在并行预测方面的性能均优于其他传统网络,验证了Res2Net-GCN-TCN模型各模块的有效性。
  • 国家重点研发计划项目(2023YFB2408203)
参考文献 引证文献
排序方式:
[1]
周远翔,李永印,陈健宁,等.基于局部放电相位图谱和油中溶解气体信息融合的油纸绝缘缺陷识别方法[J].绝缘材料,2023,56(12):43-53.
ZHOUYuanxiang, LIYongyin, CHENJianning, et al. Defect re-cognition method of oil-paper insulation based on information fusion of PRPD spectrum and dissolved gas data[J].Insulating Materials,2023,56(12):43-53.
[2]
张慧娟,齐金平,李鸿伟,等.基于隐马尔科夫模型的牵引变压器油纸绝缘剩余寿命预测[J].绝缘材料,2024,57(3):96-101.
ZHANGHuijuan, QIJinping, LIHongwei, et al. Residual life prediction of oil-paper insulation for traction transformers based on hidden Markov model[J]. Insulating Materials,2024,57(3):96-101.
[3]
LIUJiaxin, ZHAOZijian, ZHONGYuanchen, et al. Prediction of the dissolved gas concentration in power transformer oil based on sarima model[J]. Energy Reports,2022,8:1360-1367.
[4]
LUShixiang, LINGuoying, QUEHuakun, et al. Grey relational analysis using gaussian process regression method for dissolved gas concentration prediction[J]. International Journal of Machine Learning and Cybernetics,2019,10(6):1313-1322.
[5]
肖怀硕,李清泉,施亚林,等.灰色理论-变分模态分解和NSGA-Ⅱ优化的支持向量机在变压器油中气体预测中的应用[J].中国电机工程学报,2017,37(12):3643-3653,3694.
XIAOHuaishuo, LIQingquan. SHI Yalin,et al. Prediction of dissolved gases in oil for transformer based on grey theory-variational mode decomposition and support vector machine improved by NSGA-II[J]. Proceedings of the CSEE,2017,37(12):3643-3653,3694.
[6]
黄新波,蒋卫涛,朱永灿,等.基于时间序列和支持向量机的变压器故障预测[J].高电压技术,2020,46(7):2530-2538.
HUANGXinbo, JIANGWeitao, ZHUYongcan, et al. Transfor-mer fault prediction based on time series and support vector machine[J]. High Voltage Engineering,2020,46(7):2530-2538.
[7]
SPEISERJ L,MILLERM E, TOOZEJ, et al. A comparison of random forest variable selection methods for classification prediction modeling[J]. Expert Systems with Applications,2019,134:93-101.
[8]
李恩文,王力农,宋斌,等.基于改进模糊聚类算法的变压器油色谱分析[J].电工技术学报,2018,33(19):4594-4602.
LIEnwen, WANGLinong, SONGBin, et al. Analysis of transformer oil chromatography based on improved fuzzy clustering algorithm[J]. Transactions of China Electrotechnical Society, 2018,33(19):4594-4602.
[9]
ZENGWeiliang, CAOYunfei, FENGLutao, et al. Hybrid CEEMDAN-DBN-ELM for online DGA serials and transformer status forecasting[J]. Electric Power Systems Research,2023,217:109176.
[10]
刘云鹏,许自强,董王英,等.基于经验模态分解和长短期记忆神经网络的变压器油中溶解气体浓度预测方法[J].中国电机工程学报,2019,39(13):3998-4008.
LIUYunpeng, XUZiqiang, DONGWangying, et al. Concentration prediction of dissolved gases in transformer oil based on empirical mode decomposition and long short-term memory neural networks[J]. Proceedings of the CSEE,2019,39(13):3998-4008.
[11]
杨童亮,胡东,唐超,等.基于SMA-VMD-GRU模型的变压器油中溶解气体含量预测[J].电工技术学报,2023,38(1):117-130.
YANGTongliang, HUDong, TANGChao, et al. Prediction of dissolved gas content in transformer oil based on SMA-VMD-GRU model[J]. Transactions of China Electrotechnical Society, 2023,38(1):117-130.
[12]
吴汶倢,何怡刚,段嘉珺,等.油中溶解气体的灰色模型残差修正融合算法[J].电子测量与仪器学报,2018,32(10):87-94.
WUWenjie, HEYigang, DUANJiajun, et al. Residual-modified fusion algorithm of dissolved gasin transformer oil based on grey model[J]. Journal of Electronic Measurement and Instrumentation,2018,32(10):87-94.
[13]
马宏忠,肖雨松,孙永腾,等.基于ICEEMDAN和时变权重集成预测模型的变压器油中溶解气体含量预测[J].高电压技术,2024,50(1):210-220.
MAHongzhong, XIAOYusong, SUNYongteng, et al. Prediction of dissolved gas concentration in transformer oil based on ICEEMDAN and time-varying weight integrated prediction model[J]. High Voltage Engineering,2024,50(1):210-220.
[14]
MAXin, HUHao, SHANGYizi. A new method for transformer fault prediction based on multifeature enhancement and refined long short-term memory[J]. IEEE Transactions on Instrumentation and Measurement,2021,70:1-11.
[15]
刘展程,王爽,唐波.基于SSA-BiGRU-Attention模型的变压器油中溶解气体含量预测[J].高电压技术,2022,48(8):2972-2981.
LIUZhancheng, WANGShuang, TANGBo. Prediction of dissolved gas content in transformer oil based on SSA-BiGRU-Attention model [J]. High Voltage Engineering,2022,48(8):2972-2981.
[16]
赵洋,王瀚墨,康丽,等.基于时间卷积网络的短期电力负荷预测[J].电工技术学报,2022,37(5):1242-1251.
ZHAOYang, WANGHanmo, KANGLi, et al. Temporal convolution network-based short-term electrical load forecasting[J]. Transactions of China Electrotechnical Society,2022,37(5):1242-1251.
[17]
吴博,梁循,张树森,等.图神经网络前沿进展与应用[J].计算机学报,2022,45(1):35-68.
WUBo, LIANGXun, ZHANGShusen, et al. Advances and applications in graph neural network[J]. Chinese Journal of Computer,2022,45(1):35-68.
[18]
CHENLing, CHENDonghui, SHANGZongjiang, et al. Multi-scale adaptive graph neural network for multivariate time series forecasting[J]. IEEE Transactions on Knowledge and Data Engineering,2023,35(10):10748-10761.
[19]
LIAOWenlong, WANGShouxiang, BAK-JENSENB, et al. Ultra-short-term interval prediction of wind power based on graph neural network and improved bootstrap technique[J]. Journal of Modern Power Systems and Clean Energy,2023,11(4):1100-1114.
[20]
欧阳福莲,王俊,周杭霞.基于改进迁移学习和多尺度CNN-BiLSTM-Attention的短期电力负荷预测方法[J].电力系统保护与控制,2023,51(2):132-140.
OUYANGFulian, WANGJun, ZHOUHangxia. Short-term power load forecasting method based on improved hierarchical transfer learning and multi-scale CNN-BiLSTM-Attention[J].Power System Protection and Control,2023,51(2):132-140.
[21]
GAOShanghua, CHENGMingming, ZHAOKai, et al. Res2Net: a new multi-scale backbone architecture[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2021,43(2):652-662.
[22]
孙亦皓,刘浩,胡天宇,等.基于时空关联特征与GCN-FEDformer的风速短期预测方法[J].中国电机工程报,2024,44(21):8496-8507.
SUNYihao, LIUHao, HUTianyu, et al. Short-term wind speed forecasting based on GCN and FEDformer[J]. Proceedings of the CSEE,2024,44(21):8496-8507.
[23]
陈光宇,袁文辉,徐晓春,等.基于残差图卷积深度网络的电网无功储备需求快速计算方法[J].电工技术学报,2023,38(17):4683-4700.
CHENGuangyu, YUANWenhui, XUXiaochun, et al. Fast calculation method for grid reactive power reserve demand based on residual graph convolutional deep network[J]. Transactions of China Electrotechnical Society,2023,38(17):4683-4700.
[24]
ZHANJun, WUChengkun, YANGCanqun, et al. Condition monitoring of wind turbines based on spatial-temporal feature aggregation networks[J]. Renewable Energy,2022,200:751-766.
[25]
王辉东,姚海燕,郭强,等.基于多尺度卷积神经网络的变压器故障诊断方法[J].电力科学与技术学报,2023,38(4):104-112.
WANGHuidong, YAOHaiyan, GUOQiang, et al. A transformer fault diagnosis method based on multiscale 1DCNN[J]. Journal of Electric Power Science and Technology,2023,38(4):104-112.
2025年第58卷第6期
PDF下载
140
61
引用本文
BibTeX
文章信息
doi: 10.16790/j.cnki.1009-9239.im.2025.06.015
  • 接收时间:2024-05-21
  • 首发时间:2025-12-04
  • 出版时间:2025-06-20
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-05-21
  • 修回日期:2024-06-19
基金
国家重点研发计划项目(2023YFB2408203)
作者信息
    1.华北电力大学 电力工程系,河北 保定 071003
    2.国网保定供电公司,河北 保定 071066
    3.新能源电力系统全国重点实验室(华北电力大学),北京 102206

通讯作者:

谢军(1988-),男(汉族),江苏江都人,副教授,博士,主要从事高电压试验技术、电力设备状态监测的研究工作。
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/jycl/CN/10.16790/j.cnki.1009-9239.im.2025.06.015
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