Article(id=1217789891237298326, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1217789884081820362, articleNumber=null, orderNo=null, doi=10.12404/j.issn.1671-1815.2405170, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1720540800000, receivedDateStr=2024-07-10, revisedDate=1744300800000, revisedDateStr=2025-04-11, acceptedDate=null, acceptedDateStr=null, onlineDate=1768273335513, onlineDateStr=2026-01-13, pubDate=1753632000000, pubDateStr=2025-07-28, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1768273335513, onlineIssueDateStr=2026-01-13, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1768273335513, creator=13701087609, updateTime=1768273335513, updator=13701087609, issue=Issue{id=1217789884081820362, tenantId=1146029695717560320, journalId=1146123166801305609, year='2025', volume='25', issue='21', pageStart='8761', pageEnd='9209', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1768273333807, creator=13701087609, updateTime=1768273602927, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1217791012932604619, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1217789884081820362, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1217791012932604620, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1217789884081820362, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=8936, endPage=8944, ext={EN=ArticleExt(id=1217789891832889554, articleId=1217789891237298326, tenantId=1146029695717560320, journalId=1146123166801305609, language=EN, title=Fault Diagnosis in Distribution Networks with Distributed Generation Based on Improved Graph Neural Network, columnId=1156262733675876713, journalTitle=Science Technology and Engineering, columnName=Papers·Electrical Technology, runingTitle=null, highlight=null, articleAbstract=

The weak fault characteristics and the presence of numerous harmonic signals in distribution networks with renewable energy sources reduce the effectiveness of traditional fault diagnosis methods. A fault diagnosis method based on an improved graph neural network was proposed. Wavelet transform was applied to extract the detail coefficients of current and voltage before and after faults. Weighted projection correlation analysis was performed to calculate the correlation between electrical quantities. Highly correlated quantities were selected as inputs to construct the fault diagnosis model using a graph neural network. Fault simulation models for different voltage levels were developed in MATLAB/Simulink. The results indicate that fault signals are effectively enhanced, and faults are accurately located and classified in distribution networks with renewable energy sources at different voltage levels. Good diagnostic performance is maintained in the presence of data loss and noise, demonstrating strong robustness and generalization.

, correspAuthors=Bao-hua WANG, 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=Deng-yu HU, Bao-hua WANG, Jin-hong LIU), CN=ArticleExt(id=1217789895662289556, articleId=1217789891237298326, tenantId=1146029695717560320, journalId=1146123166801305609, language=CN, title=基于改进图神经网络的含源配电网故障诊断方法及效果, columnId=1156262734506353627, journalTitle=科学技术与工程, columnName=论文·电工技术, runingTitle=null, highlight=null, articleAbstract=

分布式电源大量接入,导致含源配网故障弱特征化以及故障时刻产生大量谐波信号,传统故障诊断方法应用效果不佳。提出一种基于改进图神经网络的含源配网故障诊断方法。首先,利用小波变换提取故障前后电流电压细节系数;其次,通过加权投影关联分析法计算各电气量之间的关联度;再次,选择关联度较高的电气量作为输入搭建基于图神经网络的含源配网故障诊断模型;最后,在MATLAB/Simulink中搭建了不同电压等级的含源配网故障仿真模型。结果表明,该故障诊断方法能有效强化故障信号并在不同电压等级的含源配网下对故障准确定位与分类,在数据缺失与噪声环境下也能保持良好的诊断性能,具有良好的鲁棒性与泛化性。

, correspAuthors=王宝华, authorNote=null, correspAuthorsNote=
* 王宝华(1968—),男,汉族,江苏南通人,博士,副教授。研究方向:电力系统继电保护及安全自动装置。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=QDm0km86M5pDOkmIYlXKNg==, magXml=9OeB779IHxEWUbqzSkc9gw==, pdfUrl=null, pdf=VqRf+A//0o1BlpBjRzmS7g==, pdfFileSize=6867652, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=9whxCll7IlD5jHSnkGn58w==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=98wsWvrF3BWgb1Y49far6Q==, mapNumber=null, authorCompany=null, fund=null, authors=

胡登宇(2000—),男,汉族,江苏扬州人,硕士研究生。研究方向:配电自动化。E-mail:

, authorsList=胡登宇, 王宝华, 刘晋宏)}, authors=[Author(id=1217860113721901568, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=equalhu@njust.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1217860113902256664, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, authorId=1217860113721901568, language=EN, stringName=Deng-yu HU, firstName=Deng-yu, middleName=null, lastName=HU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1217860114019697188, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, authorId=1217860113721901568, language=CN, stringName=胡登宇, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=南京理工大学自动化学院, 南京 210094, bio={"content":"

胡登宇(2000—),男,汉族,江苏扬州人,硕士研究生。研究方向:配电自动化。E-mail:

"}, bioImg=null, bioContent=

胡登宇(2000—),男,汉族,江苏扬州人,硕士研究生。研究方向:配电自动化。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1217860113587683826, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, xref=null, ext=[AuthorCompanyExt(id=1217860113596072437, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China), AuthorCompanyExt(id=1217860113604461045, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=南京理工大学自动化学院, 南京 210094)])]), Author(id=1217860114162303546, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wangbaohuaa@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1217860114338464338, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, authorId=1217860114162303546, language=EN, stringName=Bao-hua WANG, firstName=Bao-hua, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1217860114472682087, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, authorId=1217860114162303546, language=CN, stringName=王宝华, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=南京理工大学自动化学院, 南京 210094, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1217860113587683826, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, xref=null, ext=[AuthorCompanyExt(id=1217860113596072437, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China), AuthorCompanyExt(id=1217860113604461045, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=南京理工大学自动化学院, 南京 210094)])]), Author(id=1217860114581734005, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, 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=1217860114854363780, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, authorId=1217860114581734005, language=EN, stringName=Jin-hong LIU, firstName=Jin-hong, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1217860115995214479, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, authorId=1217860114581734005, language=CN, stringName=刘晋宏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=南京理工大学自动化学院, 南京 210094, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1217860113587683826, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, xref=null, ext=[AuthorCompanyExt(id=1217860113596072437, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China), AuthorCompanyExt(id=1217860113604461045, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=南京理工大学自动化学院, 南京 210094)])])], keywords=[Keyword(id=1217860116301398697, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, orderNo=1, keyword=fault diagnosis in distribution networks), Keyword(id=1217860116481753784, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, orderNo=2, keyword=maximal overlap discrete wavelet transform), Keyword(id=1217860116620165824, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, orderNo=3, keyword=grey correlation degree), Keyword(id=1217860116754383572, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, orderNo=4, keyword=weighted gray relational projection method), Keyword(id=1217860116871824098, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, orderNo=5, keyword=graph neural network), Keyword(id=1217860117022819056, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, orderNo=1, keyword=故障诊断), Keyword(id=1217860117169619706, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, orderNo=2, keyword=极大重叠离散小波变换), Keyword(id=1217860117320614662, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, orderNo=3, keyword=灰色关联度), Keyword(id=1217860117438055185, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, orderNo=4, keyword=加权灰色关联投影法), Keyword(id=1217860117572272925, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, orderNo=5, keyword=图神经网络)], refs=[Reference(id=1217860123725316250, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=43, issue=34, pageStart=152, pageEnd=154, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=史丹, journalName=价值工程, refType=null, unstructuredReference=史丹. 电气自动化技术在电力工程中的应用[J]. 价值工程, 2024, 43(34): 152-154., articleTitle=电气自动化技术在电力工程中的应用, refAbstract=null), Reference(id=1217860123842756771, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=43, issue=34, pageStart=152, pageEnd=154, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=Shi Dan, journalName=Value Engineering, refType=null, unstructuredReference=Shi Dan. Application of electrical automation technology in power engineering[J]. Value Engineering, 2024, 43(34): 152-154., articleTitle=Application of electrical automation technology in power engineering, refAbstract=null), Reference(id=1217860123964391590, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2015, volume=30, issue=2, pageStart=569, pageEnd=578, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=Sheng W, Liu K, Liu Y, journalName=IEEE Transactions on Power Delivery, refType=null, unstructuredReference=Sheng W, Liu K, Liu Y, et al. Optimal placement and sizing of distributed generation via an improved nondominated sorting genetic algorithm II[J]. IEEE Transactions on Power Delivery, 2015, 30(2): 569-578., articleTitle=Optimal placement and sizing of distributed generation via an improved nondominated sorting genetic algorithm II, refAbstract=null), Reference(id=1217860125319151791, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=16, pageStart=6367, pageEnd=6384, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=王新迎, 闫冬, 施展, journalName=中国电机工程学报, refType=null, unstructuredReference=王新迎, 闫冬, 施展, 等. 机器学习赋能的优化算法及其在新型电力系统中的应用与展望[J]. 中国电机工程学报, 2024, 44(16): 6367-6384., articleTitle=机器学习赋能的优化算法及其在新型电力系统中的应用与展望, refAbstract=null), Reference(id=1217860125453369526, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=16, pageStart=6367, pageEnd=6384, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=Wang Xinying, Yan Dong, Shi Zhan, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=Wang Xinying, Yan Dong, Shi Zhan, et al. Machine learning empowered optimization algorithms and their applications and prospects in new type power system[J]. Proceedings of the CSEE, 2024, 44(16): 6367-6384., articleTitle=Machine learning empowered optimization algorithms and their applications and prospects in new type power system, refAbstract=null), Reference(id=1217860125541449916, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2017, volume=37, issue=9, pageStart=2525, pageEnd=2538, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=牛耕, 周龙, 裴玮, journalName=中国电机工程学报, refType=null, unstructuredReference=牛耕, 周龙, 裴玮, 等. 低压有源配电网在线故障区间定位与识别方法[J]. 中国电机工程学报, 2017, 37(9): 2525-2538., articleTitle=低压有源配电网在线故障区间定位与识别方法, refAbstract=null), Reference(id=1217860125667279042, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2017, volume=37, issue=9, pageStart=2525, pageEnd=2538, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=Niu Geng, Zhou Long, Pei Wei, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=Niu Geng, Zhou Long, Pei Wei, et al. On-line fault section location and classification technique for low voltage active distribution network[J]. Proceedings of the CSEE, 2017, 37(9): 2525-2538., articleTitle=On-line fault section location and classification technique for low voltage active distribution network, refAbstract=null), Reference(id=1217860125772136651, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=18, pageStart=7710, pageEnd=7718, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=黎观锋, 梁志坚, 杨武, journalName=科学技术与工程, refType=null, unstructuredReference=黎观锋, 梁志坚, 杨武. 基于改进二进制蛇优化算法的配电网故障定位[J]. 科学技术与工程, 2024, 24(18): 7710-7718., articleTitle=基于改进二进制蛇优化算法的配电网故障定位, refAbstract=null), Reference(id=1217860125889577170, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=18, pageStart=7710, pageEnd=7718, url=null, language=null, rfNumber=[5], rfOrder=8, authorNames=Li Guanfeng, Liang Zhijian, Yang Wu, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Li Guanfeng, Liang Zhijian, Yang Wu. Method of distribution network fault location based on improved binary snake optimization algorithm[J]. Science Technology and Engineering, 2024, 24(18): 7710-7718., articleTitle=Method of distribution network fault location based on improved binary snake optimization algorithm, refAbstract=null), Reference(id=1217860126036377818, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2019, volume=47, issue=2, pageStart=80, pageEnd=86, url=null, language=null, rfNumber=[6], rfOrder=9, authorNames=谢善益, 肖斐, 艾芊, journalName=电力系统保护与控制, refType=null, unstructuredReference=谢善益, 肖斐, 艾芊, 等. 基于并行隐马尔科夫模型的电能质量扰动事件分类[J]. 电力系统保护与控制, 2019, 47(2): 80-86., articleTitle=基于并行隐马尔科夫模型的电能质量扰动事件分类, refAbstract=null), Reference(id=1217860126195761379, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2019, volume=47, issue=2, pageStart=80, pageEnd=86, url=null, language=null, rfNumber=[6], rfOrder=10, authorNames=Xie Shanyi, Xiao Fei, Ai Qian, journalName=Power System Protection and Control, refType=null, unstructuredReference=Xie Shanyi, Xiao Fei, Ai Qian, et al. Parallel hidden Markov model based classification of power quality disturbance events[J]. Power System Protection and Control, 2019, 47(2): 80-86., articleTitle=Parallel hidden Markov model based classification of power quality disturbance events, refAbstract=null), Reference(id=1217860126321590507, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2022, volume=22, issue=33, pageStart=14752, pageEnd=14758, url=null, language=null, rfNumber=[7], rfOrder=11, authorNames=齐振兴, 张倩, 丁津津, journalName=科学技术与工程, refType=null, unstructuredReference=齐振兴, 张倩, 丁津津, 等. 基于特征迁移和深度学习的配电网故障定位[J]. 科学技术与工程, 2022, 22(33): 14752-14758., articleTitle=基于特征迁移和深度学习的配电网故障定位, refAbstract=null), Reference(id=1217860126472585458, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2022, volume=22, issue=33, pageStart=14752, pageEnd=14758, url=null, language=null, rfNumber=[7], rfOrder=12, authorNames=Qi Zhenxing, Zhang Qian, Ding Jinjin, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Qi Zhenxing, Zhang Qian, Ding Jinjin, et al. Fault location in distribution networks based on feature transfer and deep learning[J]. Science Technology and Engineering, 2022, 22(33): 14752-14758., articleTitle=Fault location in distribution networks based on feature transfer and deep learning, refAbstract=null), Reference(id=1217860126594220286, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=16, pageStart=6726, pageEnd=6733, url=null, language=null, rfNumber=[8], rfOrder=13, authorNames=金星, 周凯翔, 于海洲, journalName=科学技术与工程, refType=null, unstructuredReference=金星, 周凯翔, 于海洲, 等. 基于卷积神经网络与Transformer的电能质量扰动分类方法[J]. 科学技术与工程, 2024, 24(16): 6726-6733., articleTitle=基于卷积神经网络与Transformer的电能质量扰动分类方法, refAbstract=null), Reference(id=1217860126720049409, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=16, pageStart=6726, pageEnd=6733, url=null, language=null, rfNumber=[8], rfOrder=14, authorNames=Jin Xing, Zhou Kaixiang, Yu Haizhou, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Jin Xing, Zhou Kaixiang, Yu Haizhou, et al. Power quality disturbances classification method based on convolutional neural network and Transformer[J]. Science Technology and Engineering, 2024, 24(16): 6726-6733., articleTitle=Power quality disturbances classification method based on convolutional neural network and Transformer, refAbstract=null), Reference(id=1217860126871044362, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2020, volume=29, issue=9, pageStart=131, pageEnd=135, url=null, language=null, rfNumber=[9], rfOrder=15, authorNames=杨华, 李喜旺, 司志坚, journalName=计算机系统应用, refType=null, unstructuredReference=杨华, 李喜旺, 司志坚, 等. 基于图神经网络的配电网故障预测[J]. 计算机系统应用, 2020, 29(9): 131-135., articleTitle=基于图神经网络的配电网故障预测, refAbstract=null), Reference(id=1217860127059788055, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2020, volume=29, issue=9, pageStart=131, pageEnd=135, url=null, language=null, rfNumber=[9], rfOrder=16, authorNames=Yang Hua, Li Xiwang, Si Zhijian, journalName=Computer Systems and Applications, refType=null, unstructuredReference=Yang Hua, Li Xiwang, Si Zhijian, et al. Accident prediction of power distribution network based on graph neural network[J]. Computer Systems and Applications, 2020, 29(9): 131-135., articleTitle=Accident prediction of power distribution network based on graph neural network, refAbstract=null), Reference(id=1217860127181422878, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2021, volume=45, issue=6, pageStart=2113, pageEnd=2121, url=null, language=null, rfNumber=[10], rfOrder=17, authorNames=李佳玮, 王小君, 和敬涵, journalName=电网技术, refType=null, unstructuredReference=李佳玮, 王小君, 和敬涵, 等. 基于图注意力网络的配电网故障定位方法[J]. 电网技术, 2021, 45(6): 2113-2121., articleTitle=基于图注意力网络的配电网故障定位方法, refAbstract=null), Reference(id=1217860127345000741, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2021, volume=45, issue=6, pageStart=2113, pageEnd=2121, url=null, language=null, rfNumber=[10], rfOrder=18, authorNames=Li Jiawei, Wang Xiaojun, He Jinghan, journalName=Power System Technology, refType=null, unstructuredReference=Li Jiawei, Wang Xiaojun, He Jinghan, et al. Distribution network fault location based on graph attention network[J]. Power System Technology, 2021, 45(6): 2113-2121., articleTitle=Distribution network fault location based on graph attention network, refAbstract=null), Reference(id=1217860127454052651, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2016, volume=36, issue=9, pageStart=2508, pageEnd=2515, url=null, language=null, rfNumber=[11], rfOrder=19, authorNames=李彦冬, 郝宗波, 雷航, journalName=计算机应用, refType=null, unstructuredReference=李彦冬, 郝宗波, 雷航. 卷积神经网络研究综述[J]. 计算机应用, 2016, 36(9): 2508-2515, 2565., articleTitle=卷积神经网络研究综述, refAbstract=null), Reference(id=1217860127630213428, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2016, volume=36, issue=9, pageStart=2508, pageEnd=2515, url=null, language=null, rfNumber=[11], rfOrder=20, authorNames=Li Yandong, Hao Zongbo, Lei Hang, journalName=Journal of Computer Applications, refType=null, unstructuredReference=Li Yandong, Hao Zongbo, Lei Hang. Survey of convolutional neural network[J]. Journal of Computer Applications, 2016, 36(9): 2508-2515, 2565., articleTitle=Survey of convolutional neural network, refAbstract=null), Reference(id=1217860127747653946, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=31, pageStart=13350, pageEnd=13360, url=null, language=null, rfNumber=[12], rfOrder=21, authorNames=韦波, 马君慧, 胡永, journalName=科学技术与工程, refType=null, unstructuredReference=韦波, 马君慧, 胡永, 等. 大倾角储层煤层气多层合采产量控制地质与工程因素[J]. 科学技术与工程, 2023, 23(31): 13350-13360., articleTitle=大倾角储层煤层气多层合采产量控制地质与工程因素, refAbstract=null), Reference(id=1217860127827345726, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=31, pageStart=13350, pageEnd=13360, url=null, language=null, rfNumber=[12], rfOrder=22, authorNames=Wei Bo, Ma Junhui, Hu Yong, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Wei Bo, Ma Junhui, Hu Yong, et al. Geological and engineering controlling factors of multi-layer combined production capacity of steeply inclined coal reservoir[J]. Science Technology and Engineering, 2023, 23(31): 13350-13360., articleTitle=Geological and engineering controlling factors of multi-layer combined production capacity of steeply inclined coal reservoir, refAbstract=null), Reference(id=1217860127932203334, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2015, volume=39, issue=12, pageStart=50, pageEnd=55, url=null, language=null, rfNumber=[13], rfOrder=23, authorNames=吴潇雨, 和敬涵, 张沛, journalName=电力系统自动化, refType=null, unstructuredReference=吴潇雨, 和敬涵, 张沛, 等. 基于灰色投影改进随机森林算法的电力系统短期负荷预测[J]. 电力系统自动化, 2015, 39(12): 50-55., articleTitle=基于灰色投影改进随机森林算法的电力系统短期负荷预测, refAbstract=null), Reference(id=1217860128032866636, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2015, volume=39, issue=12, pageStart=50, pageEnd=55, url=null, language=null, rfNumber=[13], rfOrder=24, authorNames=Wu Xiaoyu, He Jinghan, Zhang Pei, journalName=Automation of Electric Power Systems, refType=null, unstructuredReference=Wu Xiaoyu, He Jinghan, Zhang Pei, et al. Power system short-term load forecasting based on improved random forest with grey relation projection[J]. Automation of Electric Power Systems, 2015, 39(12): 50-55., articleTitle=Power system short-term load forecasting based on improved random forest with grey relation projection, refAbstract=null), Reference(id=1217860128133529939, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2016, volume=29, issue=null, pageStart=3844, pageEnd=3852, url=null, language=null, rfNumber=[14], rfOrder=25, authorNames=Defferrard M, Bresson X, Vandergheynst P, journalName=Advances in Neural Information Processing Systems, refType=null, unstructuredReference=Defferrard M, Bresson X, Vandergheynst P. Convolutional neural networks on graphs with fastlocalized spectral filtering[J]. Advances in Neural Information Processing Systems, 2016, 29: 3844-3852., articleTitle=Convolutional neural networks on graphs with fastlocalized spectral filtering, refAbstract=null), Reference(id=1217860128229998937, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=1609, pageEnd=null, url=null, language=null, rfNumber=[15], rfOrder=26, authorNames=Kipf T N, Welling M, journalName=ArXiv, refType=null, unstructuredReference=Kipf T N, Welling M. Semi-supervised classification with graph convolutional networks[J]. ArXiv, 2017: 1609.02907., articleTitle=Semi-supervised classification with graph convolutional networks, refAbstract=null)], funds=[Fund(id=1217860123553349764, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, awardId=BK20230934, language=CN, fundingSource=江苏省自然科学基金(BK20230934), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1217860113587683826, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, xref=null, ext=[AuthorCompanyExt(id=1217860113596072437, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China), AuthorCompanyExt(id=1217860113604461045, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, companyId=1217860113587683826, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=南京理工大学自动化学院, 南京 210094)])], figs=[ArticleFig(id=1217860117886845757, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.1, caption=Power system modeling, figureFileSmall=BNRm7cgpDAS26opDl+kYEQ==, figureFileBig=vU5K6it2un32TyU9D6URog==, tableContent=null), ArticleFig(id=1217860118138504011, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图1, caption=电力系统建模

点1~14为系统电气节点;G为分布式电源

, figureFileSmall=BNRm7cgpDAS26opDl+kYEQ==, figureFileBig=vU5K6it2un32TyU9D6URog==, tableContent=null), ArticleFig(id=1217860118318859096, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.2, caption=Schematic diagram of fault location structure based on graph neural network, figureFileSmall=ucSdi0tvN9qrh8dKw9r4Fg==, figureFileBig=bWuGd1s4HYVo3rdb9JU5hw==, tableContent=null), ArticleFig(id=1217860118562128737, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图2, caption=图神经网络故障定位结构示意图, figureFileSmall=ucSdi0tvN9qrh8dKw9r4Fg==, figureFileBig=bWuGd1s4HYVo3rdb9JU5hw==, tableContent=null), ArticleFig(id=1217860118679569265, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.3, caption=Flowchart of the improved GCN, figureFileSmall=Mw+PnJ1mbbYJ5/2izqECFw==, figureFileBig=I9/mjvgFOXs1fGZCgAx3rg==, tableContent=null), ArticleFig(id=1217860118792815480, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图3, caption=改进图神经网络流程图, figureFileSmall=Mw+PnJ1mbbYJ5/2izqECFw==, figureFileBig=I9/mjvgFOXs1fGZCgAx3rg==, tableContent=null), ArticleFig(id=1217860118927033216, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.4, caption=IEEE 33-bus distribution network model with DG, figureFileSmall=7FPRU+dlvoKNdKd1+5G4Ag==, figureFileBig=imn05bimF+wCuZXCxfjvng==, tableContent=null), ArticleFig(id=1217860119111582605, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图4, caption=IEEE33节点含源配电网模型

DG1、DG2分别表示不同的分布式电源

, figureFileSmall=7FPRU+dlvoKNdKd1+5G4Ag==, figureFileBig=imn05bimF+wCuZXCxfjvng==, tableContent=null), ArticleFig(id=1217860120055301016, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.5, caption=First-level detail coefficients of voltage, figureFileSmall=ngJlRIUIebKx6N/fMYYHYQ==, figureFileBig=ZOkj3C7mkC6JgmAfnhz60Q==, tableContent=null), ArticleFig(id=1217860120227267489, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图5, caption=电压MODWT第一级细节系数值, figureFileSmall=ngJlRIUIebKx6N/fMYYHYQ==, figureFileBig=ZOkj3C7mkC6JgmAfnhz60Q==, tableContent=null), ArticleFig(id=1217860120428594093, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.6, caption=First-level detail coefficients of current, figureFileSmall=6GqtSzjAuQ98r0/8xloX3w==, figureFileBig=BYdnY66yTEkfiVYjxG8i9w==, tableContent=null), ArticleFig(id=1217860120646697916, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图6, caption=电流MODWT第一级细节系数, figureFileSmall=6GqtSzjAuQ98r0/8xloX3w==, figureFileBig=BYdnY66yTEkfiVYjxG8i9w==, tableContent=null), ArticleFig(id=1217860120827052996, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.7, caption=Traditional distribution network detail coefficients, figureFileSmall=dmn0SZven0LI42c3FHaqhw==, figureFileBig=z4dDgK4Wl71/qnKNRVo6vw==, tableContent=null), ArticleFig(id=1217860120969659340, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图7, caption=传统配网细节系数, figureFileSmall=dmn0SZven0LI42c3FHaqhw==, figureFileBig=z4dDgK4Wl71/qnKNRVo6vw==, tableContent=null), ArticleFig(id=1217860121099682773, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.8, caption=Correlation plot, figureFileSmall=McmchzxOczRWe978zGJAIA==, figureFileBig=16XdaUkqlLDIxnRWFIH4Mw==, tableContent=null), ArticleFig(id=1217860121242289119, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图8, caption=关联系数图, figureFileSmall=McmchzxOczRWe978zGJAIA==, figureFileBig=16XdaUkqlLDIxnRWFIH4Mw==, tableContent=null), ArticleFig(id=1217860121376506855, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.9, caption=Loss and accuracy of the improved graph neural network model, figureFileSmall=IaLeE4nQu0INrn1JBhAodA==, figureFileBig=hQK4W0ZVbuVSplgoynRDqg==, tableContent=null), ArticleFig(id=1217860121577833460, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图9, caption=改进图神经网络模型损失与准确率, figureFileSmall=IaLeE4nQu0INrn1JBhAodA==, figureFileBig=hQK4W0ZVbuVSplgoynRDqg==, tableContent=null), ArticleFig(id=1217860121703662592, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.10, caption=Performance of the model with different electrical inputs, figureFileSmall=ratwn9b92IYCXJpFKhv8nA==, figureFileBig=9Ss/L3SDdPJVvWvQtdNPQg==, tableContent=null), ArticleFig(id=1217860121888210959, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图10, caption=不同电气量输入的模型性能, figureFileSmall=ratwn9b92IYCXJpFKhv8nA==, figureFileBig=9Ss/L3SDdPJVvWvQtdNPQg==, tableContent=null), ArticleFig(id=1217860122081148955, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.11, caption=t-SNE dimensionality reduction visualization of fault classification, figureFileSmall=tjj1Tzh+EiPppTojLCuJtA==, figureFileBig=Whs0BpHnxERQzV6BRJxpWQ==, tableContent=null), ArticleFig(id=1217860122194395171, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图11, caption=故障分类t-SNE降维可视化效果图, figureFileSmall=tjj1Tzh+EiPppTojLCuJtA==, figureFileBig=Whs0BpHnxERQzV6BRJxpWQ==, tableContent=null), ArticleFig(id=1217860122274086953, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.12, caption=Classification accuracy of different diagnostic models under various noise levels, figureFileSmall=tMItK+qjYlP2bwR/lEnU9g==, figureFileBig=BfM4fGefzor7JeR13VEh9A==, tableContent=null), ArticleFig(id=1217860122387333173, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图12, caption=不同噪声下诊断模型分类准确率, figureFileSmall=tMItK+qjYlP2bwR/lEnU9g==, figureFileBig=BfM4fGefzor7JeR13VEh9A==, tableContent=null), ArticleFig(id=1217860122546716735, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Fig.13, caption=Classification accuracy of diagnostic model under different data loss conditions, figureFileSmall=aK07YpuczViqTF5QC1hCZQ==, figureFileBig=b85X1mRiAMJpaABBLeyAqA==, tableContent=null), ArticleFig(id=1217860122668351559, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=图13, caption=不同数据缺失下诊断模型分类准确率, figureFileSmall=aK07YpuczViqTF5QC1hCZQ==, figureFileBig=b85X1mRiAMJpaABBLeyAqA==, tableContent=null), ArticleFig(id=1217860122781597777, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Table 1, caption=

GCN network structure and parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层类型 网络层参数 卷积核 输出层
输入层 图数据 nfeat×512
特征提取器 GCN-conv1 512×512 nfeat×512
GCN-conv2 512×512 nfeat×512
特征分类器 FC-layer 512×nclass nfeat×nclass
softmax nfeat×nclass
), ArticleFig(id=1217860122915815511, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=表1, caption=

GCN网络结构及参数

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层类型 网络层参数 卷积核 输出层
输入层 图数据 nfeat×512
特征提取器 GCN-conv1 512×512 nfeat×512
GCN-conv2 512×512 nfeat×512
特征分类器 FC-layer 512×nclass nfeat×nclass
softmax nfeat×nclass
), ArticleFig(id=1217860123033256034, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=EN, label=Table 2, caption=

Confusion matrix for fault classification

, figureFileSmall=null, figureFileBig=null, tableContent=
预测标签
0 1 2 3
真实标签 0 1.00 0.00 0.00 0.00
1 0.00 0.99 0.00 0.01
2 0.00 0.00 1.00 0.00
3 0.01 0.00 0.01 0.98
), ArticleFig(id=1217860123188445294, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1217789891237298326, language=CN, label=表2, caption=

故障分类混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
预测标签
0 1 2 3
真实标签 0 1.00 0.00 0.00 0.00
1 0.00 0.99 0.00 0.01
2 0.00 0.00 1.00 0.00
3 0.01 0.00 0.01 0.98
)], 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.1671-1815.2405170, detailUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/10.12404/j.issn.1671-1815.2405170, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/PDF/10.12404/j.issn.1671-1815.2405170, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/PDF/10.12404/j.issn.1671-1815.2405170, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于改进图神经网络的含源配电网故障诊断方法及效果
收藏切换
PDF下载
胡登宇 , 王宝华 * , 刘晋宏
科学技术与工程 | 论文·电工技术 2025,25(21): 8936-8944
收起
收藏切换
科学技术与工程 | 论文·电工技术 2025, 25(21): 8936-8944
基于改进图神经网络的含源配电网故障诊断方法及效果
全屏
胡登宇 , 王宝华* , 刘晋宏
作者信息
  • 南京理工大学自动化学院, 南京 210094
  • 胡登宇(2000—),男,汉族,江苏扬州人,硕士研究生。研究方向:配电自动化。E-mail:

通讯作者:

* 王宝华(1968—),男,汉族,江苏南通人,博士,副教授。研究方向:电力系统继电保护及安全自动装置。E-mail:
Fault Diagnosis in Distribution Networks with Distributed Generation Based on Improved Graph Neural Network
Deng-yu HU , Bao-hua WANG* , Jin-hong LIU
Affiliations
  • School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China
出版时间: 2025-07-28 doi: 10.12404/j.issn.1671-1815.2405170
文章导航
收藏切换

分布式电源大量接入,导致含源配网故障弱特征化以及故障时刻产生大量谐波信号,传统故障诊断方法应用效果不佳。提出一种基于改进图神经网络的含源配网故障诊断方法。首先,利用小波变换提取故障前后电流电压细节系数;其次,通过加权投影关联分析法计算各电气量之间的关联度;再次,选择关联度较高的电气量作为输入搭建基于图神经网络的含源配网故障诊断模型;最后,在MATLAB/Simulink中搭建了不同电压等级的含源配网故障仿真模型。结果表明,该故障诊断方法能有效强化故障信号并在不同电压等级的含源配网下对故障准确定位与分类,在数据缺失与噪声环境下也能保持良好的诊断性能,具有良好的鲁棒性与泛化性。

故障诊断  /  极大重叠离散小波变换  /  灰色关联度  /  加权灰色关联投影法  /  图神经网络

The weak fault characteristics and the presence of numerous harmonic signals in distribution networks with renewable energy sources reduce the effectiveness of traditional fault diagnosis methods. A fault diagnosis method based on an improved graph neural network was proposed. Wavelet transform was applied to extract the detail coefficients of current and voltage before and after faults. Weighted projection correlation analysis was performed to calculate the correlation between electrical quantities. Highly correlated quantities were selected as inputs to construct the fault diagnosis model using a graph neural network. Fault simulation models for different voltage levels were developed in MATLAB/Simulink. The results indicate that fault signals are effectively enhanced, and faults are accurately located and classified in distribution networks with renewable energy sources at different voltage levels. Good diagnostic performance is maintained in the presence of data loss and noise, demonstrating strong robustness and generalization.

fault diagnosis in distribution networks  /  maximal overlap discrete wavelet transform  /  grey correlation degree  /  weighted gray relational projection method  /  graph neural network
胡登宇, 王宝华, 刘晋宏. 基于改进图神经网络的含源配电网故障诊断方法及效果. 科学技术与工程, 2025 , 25 (21) : 8936 -8944 . DOI: 10.12404/j.issn.1671-1815.2405170
Deng-yu HU, Bao-hua WANG, Jin-hong LIU. Fault Diagnosis in Distribution Networks with Distributed Generation Based on Improved Graph Neural Network[J]. Science Technology and Engineering, 2025 , 25 (21) : 8936 -8944 . DOI: 10.12404/j.issn.1671-1815.2405170
新能源的接入使传统配电网转变为多端含源供电网络[1-2],其拓扑结构复杂且新能源出力不稳定。电力电子装置的限流控制导致故障特征弱化,并产生大量谐波[3],传统的故障定位方法(如矩阵法和行波法)在这种含源配电网中适应性较差。此外,基于阈值设定或逻辑判断的故障判别方法在含源配电网中的应用效果也不佳。含源配电网由于其复杂性和故障特征弱化,传统的故障定位和判别方法难以有效应用。
含源配电网故障诊断的方法主要分为两类:一类是利用自动化设备信息改进三段式电流保护[4],以适应新能源接入后的变化;另一类是利用广域监测数据进行定位[5],通过分析监测数据,总结新的判据或利用算法进行故障诊断。两类方法对数据变化的灵敏度都有一定的要求,但目前的含源配网中故障弱特征化趋势愈加明显,依赖人工经验直接使用监测数据提取故障特征会导致诊断效果变差。小波变换能够准确提取特征信息,尤其在对突变信号的处理方面效果更显著。对于配电网故障这一类暂态切换速度快、时间长度不固定的信号,可以选择极大重叠离散小波变换(maximum overlap discrete wavelet transform,MODWT)[6],该方法能够检测并分解出快速切换的瞬态信号,可以任意地选择信号起始点,且对信号长度没有要求,适用于分析故障检测等问题。
人工智能算法由于能够高效处理海量数据,在电力系统故障诊断方面逐渐广泛应用。目前比较常用的人工智能算法有Petri网、遗传算法、神经网络等。文献[7]提出了一种基于特征迁移和深度学习的配电网故障定位方法,在不同线路结构的网络中都能精确地定位故障。文献[8]利用Transformer模型和卷积神经网络建立了电能质量扰动分类模型,克服了因感受野限制带来的识别分类问题。
随着图神经网络算法的发展,可以将已有的神经网络拓展到更加适合配电网的图域上,将配电网的节点和线路映射为图结构中的顶点和边,实现图数据与人工智能算法的结合。目前应用广泛的算法有图卷积神经网络(graph convolutional network,GCN)和图注意力网络(graph attention network,GAT)等。文献[9]利用GCN对配电网进行故障预测,充分考虑节点之间的联系和影响,在融合网络拓扑结构的基础上提高了故障预测模型的精度和鲁棒性。文献[10]提出了基于GAT的配电网故障定位方法,结合配电网拓扑与图数据拓扑的相似性,提高了故障定位模型对拓扑变化的适应能力。但应用场景均为传统配网,缺乏对含源配电网的适应能力。
考虑到作为算法模型输入数据集的含源配网电气量中,部分数据之间存在特征相似性过低的情况,同时作为输入会引起特征聚合过程复杂化以及结果陷入局部最优解,算法结果出现欠拟合或过拟合现象[11],影响最后的诊断准确率。关联度分析[12]能够准确判断出电气量数据之间的相关性,从而剔除其中关联度过低的电气量数据,有效减少输入维度,减少计算复杂度避免算法出现过拟合。灰色加权关联投影算法[13]是在灰色关联度分析法的基础上改进而来,在其基础上增加了加权和投影的环节。加权可以对故障诊断的评价因素进行排序,通过将现有样本加在待分析样本上进行投影来综合评价两者之间的关联度。
现利用基于灰色加权关联投影法改进的图神经网络算法建立适用于含源配网的故障诊断模型,考虑含源配网故障弱特征化,应用MODWT对电气量数据进行波形分解,提取信号的细节系数,强化故障特征。通过关联度算法剔除相似度低、信息冗余的数据,减少计算复杂度和过拟合的风险,并通过在Simulink中搭建不同电压等级下的含源配网故障仿真模型进行验证,通过与传统故障诊断方法的对比,验证所提方法在含源配网故障情景的适用性,以期提高故障诊断模型的特征提取能力和应用泛化效果。
MODWT是在离散小波变换的基础上改进而来,其功能是一种高度冗余且对样本没有要求的非正交小波变换,克服了离散小波变换收数据起点时间影响的缺陷,且通过MODWT变换得到的频域序列与原始数据序列时间步长一致,避免了传统小波变换中需要增加频域序列长度以配合深度学习算法输入序列长度的问题。其基本定义如下。
X={Xt}(t=0,1,…,N-1)为长度N的原始时间序列,其中第j层的小波滤波器系数和尺度滤波器系数分别为hj,tgj,t,t为滤波器对应的时间序列数。在此基础上定义MODWT第j层细节系数与尺度系数分别为$ \widetilde{W}_{j, t}$$ \widetilde{V}_{j, t}$,可表示为
$\widetilde{W}_{j, t}=\sum_{l=0}^{T-1} \widetilde{h}_{j, 1}^{o} X_{t-l \bmod T}$
$ \widetilde{V}_{j, t}=\sum_{l=0}^{T-1} \widetilde{g}_{j, 1}^{o} X_{t-l \bmod T}$
式中:j为小波的层数,且j=1,2,…,J0;$ \widetilde{h_{j, t}^{o}}$$\widetilde{g_{j, t}^{o}}$分别为以T为周期进行周期化后得到的小波滤波器系数;mod为两数相除取余数;由式(1)式(2)可得J0个细节系数与尺度系数,分别为
$\widetilde{W}_{j}=\left\{\widetilde{W}_{j, 0}, \widetilde{W}_{j, 1}, \cdots, \widetilde{W}_{j, T-1}\right\}, j=1,2, \cdots, J_{0}$
$\widetilde{V}_{j}=\left\{\widetilde{V}_{j, 0}, \widetilde{V}_{j, 1}, \cdots, \widetilde{V}_{j, T-1}\right\}, j=1,2, \cdots, J_{0}$
参考谱理论,细节系数$\widetilde{W}_{1}$与尺度系数$\widetilde{V}_{1}$与变换采样频带有关,$ \widetilde{W}_{1}$采样频带为[fs/4,fs/2], $\widetilde{V}_{1}$采样频带为[0,fs/4],其中fs该频带对应的频率。对于含源配网故障诊断的故障数据,适合采用细节系数$\widetilde{W}_{1}$对其中的谐波分量进行分析。
图神经网络(graph neural network,GNN)作为一种图深度学习算法,首要步骤是构建输入拓扑。从图论的角度来看,电力网络具有天然的图结构,多数电力系统拓扑都呈树形辐射状。假设不考虑系统内部元件的情况就可以将电力系统的拓扑抽象成由节点和边构成的图。电力系统建模如图1所示。在任务场景中,将线路节点抽象成图中的节点,架空线路抽象成图中连接节点的边。
将由节点和边构成的图(Graph)记为G=(V,E),其中V表示节点集合,E表示边集合。在图谱理论中,定义拉普拉斯矩阵L
L=D-A
式(5)中:A为邻接矩阵,表示各个顶点之间的邻接关系,若相邻则元素为1,否则为0;D为对角度矩阵,对角元素为各个顶点的度,即与该顶点关联的顶点数量。
在此基础上定义归一化拉普拉斯矩阵L
$ \boldsymbol{L}=\boldsymbol{I}_{n}-\boldsymbol{D}^{-\frac{1}{2}} \boldsymbol{A} \boldsymbol{D}^{-\frac{1}{2}} $
式(6)中:In为单位阵。对L进行特征分解及对角化,有
$ \boldsymbol{L} u_{i}=\lambda_{i} u_{i} $
$ \boldsymbol{L}=\boldsymbol{U} \boldsymbol{\Lambda} \boldsymbol{U}^{\mathrm{T}} $
式中:U=[u1,u2,…,un];Λ=diag([λ1,λ2,…,λn])。
在此基础上定义图傅里叶变换为
x ^=UTx
式(9)中:x为顶点域信号; x ^为谱域信号。图傅里叶变换就是将特征向量作为基函数,将图信号从顶点域变换到谱域上。
图卷积与传统卷积神经网络的思想类似,即在图的谱域信号变换上定义一个滤波器,由卷积定理得
x*Gy=U[(UTx)☉(UTy)]
式(10)中:xy均为顶点域信号;☉为Hadamard积;*G为图卷积;UTy为谱域图滤波器。令UTy= [ θ 0 , θ 1 , , θ n - 1 ] T,并定义图卷积核gθ=diag(θ0,θ1,…,θn-1),则
x*Gy=UgθUTx
式(11)将图傅里叶变换与图卷积结合起来,但还存在一些缺陷。Defferrard等[14]利用切比雪夫多项式对gθ进行近似处理,Kipf 等[15]在此基础上提出完善得图卷积神经网络,即
$ x *{ }_{G} y=\left(I_{n}+D^{-1 / 2} A D^{-1 / 2}\right) x $
In+D-1/2AD-1/2归一化为 D ˜ - 1 / 2 A ˜ D ˜ - 1 / 2,其中 A ˜=A+In,可得在图神经网络结构中,图卷积层的表达式为
Z=Relu[ A ^XW(0)]
式(13)中: A ^= D ˜ - 1 / 2 A ˜ D ˜ - 1 / 2;X为输入特征矩阵;W为滤波器参数矩阵;Z为经过卷积操作后的输出矩阵;Relu为卷积层激活函数。
图2所示,基于图神经网络的配电网故障诊断模型,是将配电网的图拓扑结构对应的邻接矩阵作为输入,各个节点的输入特征是节点自身相应的电气量。图卷积的本质就是邻居节点之间的特征聚合从而得到新的节点特征,对节点相互之间的关系特征进行学习,再经过全连接层进行节点分类,最后实现故障诊断,即二分类的故障定位以及多分类的故障分类任务。
在含源配网中,部分冗余、相似度较低的数据会提高算法计算复杂度,可能还会导致输出结果过拟合,降低诊断精度。因此需要选择合适的输入数据集样本来提高训练效果,加权灰色关联投影法通过加权处理和投影分析的方式可得到更加合适的电气量输入数据集。
加权灰色关联投影法是在灰色关联度分析法的基础上改进而来,基础的灰色关联度用于衡量参考数据集与比较数据集的相似性,在实际应用中,不同的电气量对故障诊断结果的影响效果不同,因此需要对各电气量进行加权处理。加权处理后,将高维数据投影到低维空间并保留数据的主要特征,通过投影分析确定变量之间的相似性或关联度。其主要流程如下。
步骤1 提取输入信号。
在含源配网中,选择三相电流电压、有功无功功率、电流电压谐波、频率、相角、电压不平衡度、新能源出力等作为输入特征数据。由于不同物理量之间的单位量纲不一致,因此在输入数据前需要对各物理量进行归一化处理。
步骤2 构建灰色关联判断矩阵。
利用灰色关联度分析法计算选取的电气量之间的关联系数,得到灰色关联判断矩阵为
$ \boldsymbol{F}=\left[\begin{array}{cccc} F_{11} & F_{12} & \cdots & F_{1 i} \\ F_{21} & F_{22} & \cdots & F_{2 i} \\ \vdots & \vdots & & \vdots \\ F_{i 1} & F_{i 2} & \cdots & F_{i i} \end{array}\right] $
式(14)中:Fij为第i项与第j项电气量数据之间的关联系数,且有Fii=1,Fij=Fji
步骤3 计算影响权重。
采用熵权法确立各影响因素的权重,得到对应的权向量为
W=diag(W1,W2,…,Wi)
式(15)中:Wm为第m项数据的权重值。
步骤4 计算加权灰色关联矩阵。
将步骤2与步骤3中的判断矩阵与权重矩阵按照式(16)进行计算,即
F'=WFW
步骤5 计算灰色关联投影值。
对加权矩阵F'进行特征值分解,并选取前k个特征值较大的特征向量构成投影矩阵P,再将原始电气量数据X投影到低维空间得到投影后的数据X'
X'=XP
步骤6 构建模型输入数据集。
对数据集X'进行灰色关联度分析得到其关联矩阵F'X,分析其与原始关联矩阵F之间的差异,综合评估电气量之间关联系数的变化,并选取变化前后都保持高关联度的电气量作为输入数据集。
在得到合适的输入数据后,还需要对其进行处理。故障模型的诊断精度需要着重关注故障数据的判别结果,因此需要平衡故障与非故障数据,为此采用随机欠采样的方式随机剔除一部分非故障数据,再与故障数据一起形成新的数据集。同时再用式(18)的方式对数据进行归一化处理,将特征输入限定在[0,1]内,即
$ x^{\prime}=\frac{x-\min (x)}{\max (x)-\min (x)} $
式(18)中:max(x)、min(x)分别为数据集中x的最大值、最小值。
结合3.1节得到改进图神经网络流程图如图3所示。
考虑到不同电压等级的含源配电网中,其故障类型有所区别,相对应的故障诊断模型也不一样。以图4所示的IEEE33节点有源配电网模型为基础,在MATLAB/Simulink上搭建了含源配网的故障模型,同时设计了不同电压等级,不同新能源种类的仿真场景,以此来验证本文方法的泛化性与适应性。
线路中每一段线路首段以及DG1与DG2的接入点处均设置了测量点,共计34个。
实验基于Pytorch 与PyTorch Geometric框架实现,CPU为Intel Core i7-9750H,GPU为NVIDIA GeForce GTX 1660Ti,运行内存为32 G。GCN模型的网络层及卷积核结构如表1所示。
GCN模型包含两个卷积层GCN-conv和一个全连接层FC-layer。超参数选取如下:批处理大小batchsize为32,epoch 为100,优化器选择Adam,学习率为1×10-5,损失函数为交叉熵损失。
1)故障定位
系统设置为DG1、DG2处接入风电的35 kV接地系统,分别在含源配网的32段线路之间设置故障信号,仿真时长为0.5 s,故障起始时间设置为0.3 s,到0.34 s时故障切除,采集故障前0.02 s至故障切除共0.06 s内的各测量点的各类电气量数据。通过MODWT对数据信号进行加强后,提取其一级小波细节系数作为原始数据集。数据集中每段线路包含故障数据200个,非故障数据100个。按照前文所述欠采样的方式对数据进行欠采样,得到新数据集中故障数据与非故障数据比例为1∶1,共计6 400组数据。将数据集按照9∶1的比例划分为训练集与测试集。其中训练集有5 760个样本,测试集有640个样本。
2)故障分类
在故障分类时,考虑到新能源类型与电压等级基于IEEE33节点的含源配网设置了两种不同的场景。
场景一:DG1、DG2处接入光伏的10 kV中性点不接地系统,选取配网的7、14、21、28号测量点对应的线路并分别仿真AB、AC、BC、ABC共计4类相间故障,按照故障定位中的仿真设计采集故障开始至切除时刻的数据共4×4×200=3 200组,再将其按9∶1的比例划分为训练集与测试集。
场景二:DG1、DG2处接入逆变型风机组的35 kV中性点经接地变接地系统,选取同样的线路段并分别仿真AB、AC、BC、ABC、AG、BG、CG、ABG、ACG、BCG、ABCG共计10类故障,得到故障数据8 000组,并划分为验证集与测试集。
首先观察小波变换的结果,在距离DG1接入点最近的13号节点处设置AG故障,并选择相邻的12号测量点数据作为参考,故障时电压波形经MODWT变换效果如图5所示。
图5可知,故障发生时刻,系统产生大量谐波信号,经过MODWT后,故障信号被放大而正常信号值被适当削弱,达到了强化故障信号的效果。由于分级后的二级及以后的细节系数信号微弱,此处不做考虑,只选用第一级细节系数。
对于谐波信号丰富的含源配网电压与电流,都可以利用这种变换方法来增强故障信号,图6为该处监测点的电流信号细节系数值。
图6可知,电流经过变换后的故障信号也得到了放大,因此可选取第一级细节系数作为其输入特征量。但是对比电压信号的强化效果,电流故障信号的强化效果并不显著,在对其余电气量变换之后对故障特征的强化效果也不佳,其原因是谐波信号在这些电气量中的特征不明显,因此只对电压电流进行了MODWT变换,保留了其余电气量的原始值。
而在传统配网中,由于其中的谐波成分较少,因此利用MODWT来强化故障信号的效果反而不明显。变换效果如图7所示。
图7可知,变换后的故障特征并未得到明显的强化,故障暂态的细节系数相较于正常状态没有较大的差异,仅在故障时刻开始时有部分变化,因此该变换不适用于传统配电网故障诊断。
得到输入数据集后,再对各个电气量数据进行关联度分析,选择高关联度的数据作为输入信号。每一组数据中除了经过MODWT变换的三相电压U三相电流I第一级细节系数外,还包含有功功率P、无功功率Q、谐波成分H、电压相角Uθ、电流相角Iθ、频率f、电压不平衡度VU以及新能源出力Pr。将这些数据利用加权灰色关联投影法处理后得到关联系数图如图8所示。
对比原始数据集与经过加权投影法的电气量关联系数可知,谐波H、相角θ、频率f等电气量关联系数在经过加权投影处理后进一步降低,而UIPQPrVU电气量的关联度在变化前后都比较高,说明在故障情况下,它们变化趋势相似,因此作为算法的输入信号可以提高模型的精度,得到更好的诊断效果。
选择经MODWT变换提取的三相电压U以及三相电流I的第一级细节系数值、有功功率P、无功功率Q、电压不平衡度VU以及新能源出力Pr作为图神经网络模型的输入,输入维度nfeat为10,故障定位为二分类问题,输出维度nclass设置为2。模型的训练集与测试集的损失loss与准确率曲线如图9所示。
图9(a)可知,经过500个epoch的训练,训练集损失loss接近于0,准确率几乎达到100%,说明模型未出现过拟合与欠拟合。训练完成后,将测试集数据输入到故障诊断模型里,得到故障定位结果。由图9(b)可知,测试集的定位准确率达到了99.8%。
为了验证不同电气量输入特征对定位结果的影响,选择不同的电气量组合再次进行训练与测试。组合一是传统配电网故障诊断的三相电流电压;组合二是分析配网稳定性的有功无功功率、电压以及电压相角;组合三是所采集的所有电气量;组合四是经过本文方法选取的电气量。图10是上述4种组合的测试集效果。
图10可知,本文方法(组合四)选择了合适的电气量输入信号,模型的诊断性能也最佳。组合三中由于输入过多电气量,导致模型特征集合效果不佳,进而影响故障诊断的性能。
场景一故障分类的配电网故障仿真模型与定位模型一致,因此选择同样的电气量作为输入数据集,训练完成后输入测试集数据,得到其故障分类混淆矩阵如表2所示。
从混淆矩阵中可知,诊断模型对于场景一中故障四分类问题的标签预测效果很好,测试集的故障分类准确率达到99.25%,表明本文方法具有良好的故障分类能力。
为了更加直观地表现本文方法的诊断效果,利用t-分布式随机邻域嵌入(t-distributed stochastic neighbor embedding,t-SNE)将测试集数据经过算法每一层卷积核后的输出特征向量降维可视化,通过分布图的方式对结果进行分析,如图11所示。
图11中的4个故障标签分别对应AB、AC、BC、ABC四类相间故障,每一层特征向量数量与测试集数据样本一致。输入层特征图中明显存在特征重叠现象,ABC相间故障对应的红点与其他三类故障的重叠部分最多,自身的故障特征也比较分散。经过一层图卷积后各类故障特征开始聚合,经过二层卷积后聚合完毕。不同故障类别特征类间距离增大,且相同故障类型特征类内间距减小,验证了本文方法具有良好的特征提取、聚合以及分类能力。
场景二含源配网模型为中高压风电并网的中性点接地系统,相对于场景一而言,其故障类型增加了接地故障,故障类别达到了10种。为了验证本文方法的优越性,分别在数据缺失和噪声环境下对比分析不同模型分类结果。
1)不同噪声环境下模型分类精度的对比
为了验证故障诊断模型的抗噪性能,分别设计了在无噪声环境以及信噪比(signal-noise ratio,SNR)分别为-5 dB和-10 dB这3种环境下,传统卷积神经网络、图神经网络以及所提的改进图神经网络的故障分类模型的对比试验。其中SNR的定义如式(19)所示。
$ \mathrm{SNR}=10 \lg \frac{P_{\text {signal }}}{P_{\text {noise }}} $
式(19)中:SNR为信噪比,dB;Psignal为信号功率;Pnoise为噪声功率。
经过MODWT变换以及关联度分析后,选取的输入特征电气量为三相电压U、三相电流I、有功功率P、无功功率Q以及新能源出力Pr。对比模型在不同噪声环境下的故障分类结果如图12所示。
图12可以明显看出,噪声越大,故障分类精度越低。其中,本文方法在3种噪声环境下都表现最佳,在-5 dB、-10 dB的噪声环境下都大幅领先其他模型。虽然在无噪环境下,CNN与GCN都能表现出较好的分类效果,但在噪声环境下准确率大幅下降,相反改进GCN一致保持较高的准确率,验证了所提的改进GCN模型具有优异的鲁棒性。
2)数据缺失对分类精度的影响
为了验证数据缺失对结果的影响,对无噪声数据进行随机数据缺失处理,即对节点的电气量测量值按一定比例进行随机丢弃,且丢弃数据的输入特征值直接用0代替。并设计了无数据缺失、数据缺失比例为10%和20%共3种情景下的对比实验,结果如图13所示。一般数据缺失超过30%会严重影响模型性能,本文方法在缺失比例20%的情景下仍能保持较高精度并优于其他模型,验证了所提的改进GCN模型具有优异的鲁棒性与泛化性。
提出了基于改进图神经网络的含源配网故障诊断方法,首先通过MODWT变换放大信号,强化故障特征,然后利用关联度分析结合图神经网络算法,在不同场景中自适应地选择合适的输入信号,有效快速地提取故障特征从而实现故障诊断,通过实验与分析得到以下结论。
(1)所提算法适用于建立含源配网的故障诊断模型,能增强含源配网中的故障信号并通过自适应地选取输入信号,避免了使用传统三相电压电流直接作为输入量带来的局限性。
(2)所提方法具有良好的泛化性,可以分别应用于配电网故障定位与故障诊断,并在不同电压等级不同新能源类型下都能达到较高的准确率。
(3)方法具有良好的鲁棒性,能极大程度地降低环境噪声与数据缺失的影响,在强噪环境以及数据缺失的情境下仍具有良好的诊断能力。
  • 江苏省自然科学基金(BK20230934)
参考文献 引证文献
排序方式:
[1]
史丹. 电气自动化技术在电力工程中的应用[J]. 价值工程, 2024, 43(34): 152-154.
Shi Dan. Application of electrical automation technology in power engineering[J]. Value Engineering, 2024, 43(34): 152-154.
[2]
Sheng W, Liu K, Liu Y, et al. Optimal placement and sizing of distributed generation via an improved nondominated sorting genetic algorithm II[J]. IEEE Transactions on Power Delivery, 2015, 30(2): 569-578.
[3]
王新迎, 闫冬, 施展, 等. 机器学习赋能的优化算法及其在新型电力系统中的应用与展望[J]. 中国电机工程学报, 2024, 44(16): 6367-6384.
Wang Xinying, Yan Dong, Shi Zhan, et al. Machine learning empowered optimization algorithms and their applications and prospects in new type power system[J]. Proceedings of the CSEE, 2024, 44(16): 6367-6384.
[4]
牛耕, 周龙, 裴玮, 等. 低压有源配电网在线故障区间定位与识别方法[J]. 中国电机工程学报, 2017, 37(9): 2525-2538.
Niu Geng, Zhou Long, Pei Wei, et al. On-line fault section location and classification technique for low voltage active distribution network[J]. Proceedings of the CSEE, 2017, 37(9): 2525-2538.
[5]
黎观锋, 梁志坚, 杨武. 基于改进二进制蛇优化算法的配电网故障定位[J]. 科学技术与工程, 2024, 24(18): 7710-7718.
Li Guanfeng, Liang Zhijian, Yang Wu. Method of distribution network fault location based on improved binary snake optimization algorithm[J]. Science Technology and Engineering, 2024, 24(18): 7710-7718.
[6]
谢善益, 肖斐, 艾芊, 等. 基于并行隐马尔科夫模型的电能质量扰动事件分类[J]. 电力系统保护与控制, 2019, 47(2): 80-86.
Xie Shanyi, Xiao Fei, Ai Qian, et al. Parallel hidden Markov model based classification of power quality disturbance events[J]. Power System Protection and Control, 2019, 47(2): 80-86.
[7]
齐振兴, 张倩, 丁津津, 等. 基于特征迁移和深度学习的配电网故障定位[J]. 科学技术与工程, 2022, 22(33): 14752-14758.
Qi Zhenxing, Zhang Qian, Ding Jinjin, et al. Fault location in distribution networks based on feature transfer and deep learning[J]. Science Technology and Engineering, 2022, 22(33): 14752-14758.
[8]
金星, 周凯翔, 于海洲, 等. 基于卷积神经网络与Transformer的电能质量扰动分类方法[J]. 科学技术与工程, 2024, 24(16): 6726-6733.
Jin Xing, Zhou Kaixiang, Yu Haizhou, et al. Power quality disturbances classification method based on convolutional neural network and Transformer[J]. Science Technology and Engineering, 2024, 24(16): 6726-6733.
[9]
杨华, 李喜旺, 司志坚, 等. 基于图神经网络的配电网故障预测[J]. 计算机系统应用, 2020, 29(9): 131-135.
Yang Hua, Li Xiwang, Si Zhijian, et al. Accident prediction of power distribution network based on graph neural network[J]. Computer Systems and Applications, 2020, 29(9): 131-135.
[10]
李佳玮, 王小君, 和敬涵, 等. 基于图注意力网络的配电网故障定位方法[J]. 电网技术, 2021, 45(6): 2113-2121.
Li Jiawei, Wang Xiaojun, He Jinghan, et al. Distribution network fault location based on graph attention network[J]. Power System Technology, 2021, 45(6): 2113-2121.
[11]
李彦冬, 郝宗波, 雷航. 卷积神经网络研究综述[J]. 计算机应用, 2016, 36(9): 2508-2515, 2565.
Li Yandong, Hao Zongbo, Lei Hang. Survey of convolutional neural network[J]. Journal of Computer Applications, 2016, 36(9): 2508-2515, 2565.
[12]
韦波, 马君慧, 胡永, 等. 大倾角储层煤层气多层合采产量控制地质与工程因素[J]. 科学技术与工程, 2023, 23(31): 13350-13360.
Wei Bo, Ma Junhui, Hu Yong, et al. Geological and engineering controlling factors of multi-layer combined production capacity of steeply inclined coal reservoir[J]. Science Technology and Engineering, 2023, 23(31): 13350-13360.
[13]
吴潇雨, 和敬涵, 张沛, 等. 基于灰色投影改进随机森林算法的电力系统短期负荷预测[J]. 电力系统自动化, 2015, 39(12): 50-55.
Wu Xiaoyu, He Jinghan, Zhang Pei, et al. Power system short-term load forecasting based on improved random forest with grey relation projection[J]. Automation of Electric Power Systems, 2015, 39(12): 50-55.
[14]
Defferrard M, Bresson X, Vandergheynst P. Convolutional neural networks on graphs with fastlocalized spectral filtering[J]. Advances in Neural Information Processing Systems, 2016, 29: 3844-3852.
[15]
Kipf T N, Welling M. Semi-supervised classification with graph convolutional networks[J]. ArXiv, 2017: 1609.02907.
2025年第25卷第21期
PDF下载
258
117
引用本文
BibTeX
文章信息
doi: 10.12404/j.issn.1671-1815.2405170
  • 接收时间:2024-07-10
  • 首发时间:2026-01-13
  • 出版时间:2025-07-28
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-07-10
  • 修回日期:2025-04-11
基金
江苏省自然科学基金(BK20230934)
作者信息
    南京理工大学自动化学院, 南京 210094

通讯作者:

* 王宝华(1968—),男,汉族,江苏南通人,博士,副教授。研究方向:电力系统继电保护及安全自动装置。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2405170
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