Article(id=1152988935132275039, tenantId=1146029695717560320, journalId=1146119893612605453, issueId=1152988930656948403, articleNumber=null, orderNo=null, doi=null, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1715616000000, receivedDateStr=2024-05-14, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1752823583634, onlineDateStr=2025-07-18, pubDate=1739980800000, pubDateStr=2025-02-20, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1752823583634, onlineIssueDateStr=2025-07-18, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1752823583634, creator=13701087609, updateTime=1752823583634, updator=13701087609, issue=Issue{id=1152988930656948403, tenantId=1146029695717560320, journalId=1146119893612605453, year='2025', volume='43', issue='2', pageStart='143', pageEnd='284', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1752823582567, creator=13701087609, updateTime=1753694496025, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1156641806499570521, tenantId=1146029695717560320, journalId=1146119893612605453, issueId=1152988930656948403, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1156641806499570522, tenantId=1146029695717560320, journalId=1146119893612605453, issueId=1152988930656948403, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=217, endPage=224, ext={EN=ArticleExt(id=1152988935430070624, articleId=1152988935132275039, tenantId=1146029695717560320, journalId=1146119893612605453, language=EN, title=Considering the health status of wind turbines and the dual attention mechanism CNN-BiLSTM ultra-short-term power prediction, columnId=null, journalTitle=Renewable Energy Resources, columnName=null, runingTitle=null, highlight=null, articleAbstract=

In order to improve the accuracy of ultrashortterm power prediction of wind turbines, this paper proposes a CNNBiLSTM ultrashortterm power prediction method considering the health status of wind turbines and dual attention mechanism. Firstly, considering the influence of the interaction between the environmental factors and the components of the wind turbine on the output power of the wind turbine, he relative error of the normal operation of each component of the wind turbine is used as the deterioration degree of the monitoring index. Secondly, the fuzzy comprehensive evaluation method assesses the health of wind turbines, and the historical data set is categorized based on the evaluation results. Finally, the dual attention mechanism CNN BiLSTM model is used to construct an ultrashortterm power prediction model for the classified data set. The experimental results show that the RMSE and MAE considering the health status of wind turbines are reduced by 17.3% and 20.5% respectively compared with the RSME and MSE without considering the health status of wind turbines.

, 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=Kaiwei Zhang, Zhong Wen, Shengpeng Yang, Zihan Hu, Jian Ding), CN=ArticleExt(id=1152988959559901853, articleId=1152988935132275039, tenantId=1146029695717560320, journalId=1146119893612605453, language=CN, title=考虑风电机组健康状况与双重注意力机制 CNN – BILSTM 的超短期功率预测, columnId=null, journalTitle=可再生能源, columnName=null, runingTitle=null, highlight=null, articleAbstract=

为提升风电机组超短期功率预测的准确性,文章提出了一种考虑风电机组健康状况与双重注意力机制 CNNBiLSTM 的超短期功率预测模型。首先,综合考虑环境因素与风电机组各子部件的相互作用对风电机组输出功率的影响,将风电机组各个子部件正常运行时的相对误差作为监测指标的劣化度;然后,采用模糊综合评价法对风电机组健康状况进行评估,根据评估结果对其历史数据集进行健康状况划分;最后,采用双重注意力机制 CNNBiLSTM 模型对分类后的数据集构建超短期功率预测模型。实验结果表明,在风电机组功率预测过程中,相较于未考虑机组健康状况,考虑机组健康状况的均方根误差(RMSE)和平均绝对误差(MAE)分别降低了17.3%和20.5%。

, correspAuthors=null, authorNote=null, correspAuthorsNote=
文中(1968-),男,硕士,副教授,研究方向为输电线路工程技术、电力系统运行与控制。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=jugGXKMFdE/+YVpGhuAV1w==, magXml=wx0uuDKuGshw1zlkY2bnTA==, pdfUrl=null, pdf=R+XyC+XoQcR2GW4QgOVvLg==, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=2o3LSecTdHBKowscS1LFAQ==, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=张开伟, 文中, 杨生鹏, 胡梓涵, 丁剑)}, authors=[Author(id=1159145545848967442, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, 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=1159145545903493396, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145545848967442, language=EN, stringName=Kaiwei Zhang, firstName=Kaiwei, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1159145545953825045, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145545848967442, language=CN, stringName=张开伟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 三峡大学 电气与新能源学院 湖北 宜昌 443002, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1159145545697972491, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=1, ext=[AuthorCompanyExt(id=1159145545702166796, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China), AuthorCompanyExt(id=1159145545710555405, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 三峡大学 电气与新能源学院 湖北 宜昌 443002)])]), Author(id=1159145545991573783, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=2791415163@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1159145546050294041, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145545991573783, language=EN, stringName=Zhong Wen, firstName=Zhong, middleName=null, lastName=Wen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1159145546096431387, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145545991573783, language=CN, stringName=文中, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 三峡大学 电气与新能源学院 湖北 宜昌 443002, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1159145545697972491, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=1, ext=[AuthorCompanyExt(id=1159145545702166796, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China), AuthorCompanyExt(id=1159145545710555405, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 三峡大学 电气与新能源学院 湖北 宜昌 443002)])]), Author(id=1159145546146763038, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, 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=1159145546201288993, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145546146763038, language=EN, stringName=Shengpeng Yang, firstName=Shengpeng, middleName=null, lastName=Yang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1159145546251620643, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145546146763038, language=CN, stringName=杨生鹏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 三峡大学 电气与新能源学院 湖北 宜昌 443002, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1159145545697972491, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=1, ext=[AuthorCompanyExt(id=1159145545702166796, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China), AuthorCompanyExt(id=1159145545710555405, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 三峡大学 电气与新能源学院 湖北 宜昌 443002)])]), Author(id=1159145546301952294, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, 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=1159145546360672553, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145546301952294, language=EN, stringName=Zihan Hu, firstName=Zihan, middleName=null, lastName=Hu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1159145546427781420, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145546301952294, language=CN, stringName=胡梓涵, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 三峡大学 电气与新能源学院 湖北 宜昌 443002, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1159145545697972491, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=1, ext=[AuthorCompanyExt(id=1159145545702166796, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China), AuthorCompanyExt(id=1159145545710555405, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 三峡大学 电气与新能源学院 湖北 宜昌 443002)])]), Author(id=1159145546499084591, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, 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=1159145546557804849, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145546499084591, language=EN, stringName=Jian Ding, firstName=Jian, middleName=null, lastName=Ding, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 Shanghai Survey and Design Institute Co., Ltd. Shanghai 200434 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1159145546599747890, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, authorId=1159145546499084591, language=CN, stringName=丁剑, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 上海勘测设计研究院有限公司 上海 200434, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1159145545781858574, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=2, ext=[AuthorCompanyExt(id=1159145545798635791, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545781858574, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Shanghai Survey and Design Institute Co., Ltd. Shanghai 200434 China), AuthorCompanyExt(id=1159145545802830096, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545781858574, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 上海勘测设计研究院有限公司 上海 200434)])])], keywords=[Keyword(id=1159145546725577011, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, orderNo=1, keyword=ultra-short-term), Keyword(id=1159145546788491572, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, orderNo=2, keyword=power prediction), Keyword(id=1159145546838823221, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, orderNo=3, keyword=health status), Keyword(id=1159145546884960566, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, orderNo=4, keyword=dual attention mechanism), Keyword(id=1159145546935292215, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, orderNo=5, keyword=CNN-BILSTM model), Keyword(id=1159145546998206776, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, orderNo=1, keyword=超短期), Keyword(id=1159145547040149817, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, orderNo=2, keyword=功率预测), Keyword(id=1159145547082092858, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, orderNo=3, keyword=健康状况), Keyword(id=1159145547119841595, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, orderNo=4, keyword=双重注意力机制), Keyword(id=1159145547161784636, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, orderNo=5, keyword=CNN-BiLSTM 模型)], refs=[Reference(id=1159145548860477788, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2021, volume=36, issue=22, pageStart=4627, pageEnd=4636, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=马伟明, journalName=电工技术学报, refType=null, unstructuredReference=马伟明. 关于电工学科前沿技术发展的若干思考[J]. 电工技术学报, 2021, 36(22): 4627-4636., articleTitle=关于电工学科前沿技术发展的若干思考, refAbstract=null), Reference(id=1159145548915003741, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=7, pageStart=8, pageEnd=9, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=舒印彪, journalName=中国电力企业管理, refType=null, unstructuredReference=舒印彪. 发展新型电力系统助力实现“双碳”目标[J]. 中国电力企业管理, 2021(7): 8-9., articleTitle=发展新型电力系统助力实现“双碳”目标, refAbstract=null), Reference(id=1159145548956946782, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2023, volume=214, issue=null, pageStart=108817, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=Wang S, Sun Y H, Zhang S M, journalName=Electric Power Systems Research, refType=null, unstructuredReference=Wang S, Sun Y H, Zhang S M, et al. Very short-term probabilistic prediction of PV based on multi -period error dis-tribution[J]. Electric Power Systems Research, 2023, 214: 108817., articleTitle=Very short-term probabilistic prediction of PV based on multi -period error dis-tribution, refAbstract=null), Reference(id=1159145549040832863, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2021, volume=36, issue=19, pageStart=4148, pageEnd=4160, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=李军徽, 张嘉辉, 李翠萍, journalName=电工技术学报, refType=null, unstructuredReference=李军徽, 张嘉辉, 李翠萍, 等. 参与调峰的储能系统配置方案及经济性分析[J]. 电工技术学报, 2021, 36(19): 4148-4160., articleTitle=参与调峰的储能系统配置方案及经济性分析, refAbstract=null), Reference(id=1159145549103747424, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2020, volume=52, issue=6, pageStart=234, pageEnd=241, url=null, language=null, rfNumber=[5], rfOrder=4, authorNames=张红涛, 韩婧, 谭联, journalName=工程科学与技术, refType=null, unstructuredReference=张红涛, 韩婧, 谭联, 等. 基于 FCM 和 SSA-ELM 的超短期风功率预测[J]. 工程科学与技术, 2020, 52(6): 234-241., articleTitle=基于 FCM 和 SSA-ELM 的超短期风功率预测, refAbstract=null), Reference(id=1159145549149884769, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2024, volume=36, issue=7, pageStart=59, pageEnd=66, url=null, language=null, rfNumber=[6], rfOrder=5, authorNames=潘鹏程, 刘晖, 王仁明, journalName=电力系统及其自动化学报, refType=null, unstructuredReference=潘鹏程, 刘晖, 王仁明. 自适应密度聚类组合数据清洗的LSTM 风电功率预测[J]. 电力系统及其自动化学报, 2024, 36(7): 59-66., articleTitle=自适应密度聚类组合数据清洗的LSTM 风电功率预测, refAbstract=null), Reference(id=1159145549196022114, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2023, volume=36, issue=7, pageStart=22, pageEnd=30, url=null, language=null, rfNumber=[7], rfOrder=6, authorNames=邓韦斯, 卢斯煜, 刘显茁, journalName=广东电力, refType=null, unstructuredReference=邓韦斯, 卢斯煜, 刘显茁, 等. 基于相空间重构和 BiLSTM 的风电功率短期预测[J]. 广东电力, 2023, 36(7): 22-30., articleTitle=基于相空间重构和 BiLSTM 的风电功率短期预测, refAbstract=null), Reference(id=1159145549246353763, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2019, volume=39, issue=5, pageStart=1259, pageEnd=1268, url=null, language=null, rfNumber=[8], rfOrder=7, authorNames=杨茂, 周宜, journalName=中国电机工程学报, refType=null, unstructuredReference=杨茂, 周宜. 计及风电场状态的风电功率超短期预测[J]. 中国电机工程学报, 2019, 39(5): 1259-1268., articleTitle=计及风电场状态的风电功率超短期预测, refAbstract=null), Reference(id=1159145549292491108, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2018, volume=47, issue=6, pageStart=694, pageEnd=713, url=null, language=null, rfNumber=[9], rfOrder=8, authorNames=张静, 李柠, 李少远, journalName=信息与控制, refType=null, unstructuredReference=张静, 李柠, 李少远, 等. 基于数据的风电机组发电机健康状况评估[J]. 信息与控制, 2018, 47(6): 694-713., articleTitle=基于数据的风电机组发电机健康状况评估, refAbstract=null), Reference(id=1159145549342822757, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2020, volume=36, issue=8, pageStart=57, pageEnd=62, url=null, language=null, rfNumber=[10], rfOrder=9, authorNames=孙海蓉, 王皓茹, 王瑞珈, journalName=电力科学与工程, refType=null, unstructuredReference=孙海蓉, 王皓茹, 王瑞珈, 等. 基于云模型的风电机组运行状态评价[J]. 电力科学与工程, 2020, 36(8): 57-62., articleTitle=基于云模型的风电机组运行状态评价, refAbstract=null), Reference(id=1159145549388960102, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=9, pageStart=7439, pageEnd=7446, url=null, language=null, rfNumber=[11], rfOrder=10, authorNames=S Chen, X Guo, S Han, journalName=IEEE Access, refType=null, unstructuredReference=S Chen, X Guo, S Han, et al. A novel stochastic cloud model for statistical characterization of wind turbine output[J]. IEEE Access, 2021(9): 7439-7446., articleTitle=A novel stochastic cloud model for statistical characterization of wind turbine output, refAbstract=null), Reference(id=1159145549439291751, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2022, volume=42, issue=12, pageStart=4292, pageEnd=4303, url=null, language=null, rfNumber=[12], rfOrder=11, authorNames=符杨, 任子旭, 魏书荣, journalName=中国电机工程学报, refType=null, unstructuredReference=符杨, 任子旭, 魏书荣, 等. 基于改进 LSTM-TCN 模型的海上风电超短期功率预测[J]. 中国电机工程学报, 2022, 42(12): 4292-4303., articleTitle=基于改进 LSTM-TCN 模型的海上风电超短期功率预测, refAbstract=null), Reference(id=1159145549502206312, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=5, pageStart=1884, pageEnd=1892, url=null, language=null, rfNumber=[13], rfOrder=12, authorNames=王渝红, 史云翔, 周旭, journalName=高电压技术, refType=null, unstructuredReference=王渝红, 史云翔, 周旭, 等. 基于时间模式注意力机制的BiLSTM 多风电机组超短期功率预测[J]. 高电压技术, 2022, 48(5): 1884-1892., articleTitle=基于时间模式注意力机制的BiLSTM 多风电机组超短期功率预测, refAbstract=null), Reference(id=1159145549548343657, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, doi=null, pmid=null, pmcid=null, year=2021, volume=45, issue=1, pageStart=28, pageEnd=35, url=null, language=null, rfNumber=[14], rfOrder=13, authorNames=凡航, 张雪敏, 梅生伟, journalName=电力系统自动化, refType=null, unstructuredReference=凡航, 张雪敏, 梅生伟, 等. 基于时空神经网络的风电场超短期风速预测模型[J]. 电力系统自动化, 2021, 45(1): 28-35., articleTitle=基于时空神经网络的风电场超短期风速预测模型, refAbstract=null)], funds=[Fund(id=1159145548768203099, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, awardId=52107108, language=CN, fundingSource=国家自然科学基金项目(52107108), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1159145545697972491, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=1, ext=[AuthorCompanyExt(id=1159145545702166796, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Electrical and New Energy Three Gorges University Yichang 443002 China), AuthorCompanyExt(id=1159145545710555405, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545697972491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 三峡大学 电气与新能源学院 湖北 宜昌 443002)]), AuthorCompany(id=1159145545781858574, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, xref=2, ext=[AuthorCompanyExt(id=1159145545798635791, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545781858574, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Shanghai Survey and Design Institute Co., Ltd. Shanghai 200434 China), AuthorCompanyExt(id=1159145545802830096, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, companyId=1159145545781858574, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 上海勘测设计研究院有限公司 上海 200434)])], figs=[ArticleFig(id=1159145547291808061, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 1, caption=Standard power characteristic curve, figureFileSmall=fUhgcmOMfr2oQUxzgOhFDQ==, figureFileBig=5ItWLtt+UJ0h2qp3yzfm2A==, tableContent=null), ArticleFig(id=1159145547337945406, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 1, caption=标准功率特性曲线, figureFileSmall=fUhgcmOMfr2oQUxzgOhFDQ==, figureFileBig=5ItWLtt+UJ0h2qp3yzfm2A==, tableContent=null), ArticleFig(id=1159145547384082751, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 2, caption=Wind speed power relationship under various conditions, figureFileSmall=wF0Poz3XGMghWUQuck6uzQ==, figureFileBig=prbgnQp7odjMb2ypbxp8IQ==, tableContent=null), ArticleFig(id=1159145547421831488, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 2, caption=不同状态下风速-功率关系, figureFileSmall=wF0Poz3XGMghWUQuck6uzQ==, figureFileBig=prbgnQp7odjMb2ypbxp8IQ==, tableContent=null), ArticleFig(id=1159145547467968833, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 3, caption=Health assessment process flowchart, figureFileSmall=n7yXIQ8M7uFXy/vOpI8/Ng==, figureFileBig=RlJajYn2cNhhcPfpRs9Ztw==, tableContent=null), ArticleFig(id=1159145547518300482, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 3, caption=健康状况评估流程, figureFileSmall=n7yXIQ8M7uFXy/vOpI8/Ng==, figureFileBig=RlJajYn2cNhhcPfpRs9Ztw==, tableContent=null), ArticleFig(id=1159145547551854915, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 4, caption=Membership functions of each grade, figureFileSmall=WZmhvpfsyrzg09GTI/lb4g==, figureFileBig=QpyZwgasMipRkVM7vtUd/g==, tableContent=null), ArticleFig(id=1159145547593797956, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 4, caption=各等级下的隶属度函数, figureFileSmall=WZmhvpfsyrzg09GTI/lb4g==, figureFileBig=QpyZwgasMipRkVM7vtUd/g==, tableContent=null), ArticleFig(id=1159145547635740997, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 5, caption=Schematic diagram of the convolutional layer operation, figureFileSmall=xBunN6eISocfsGYwvGJ91g==, figureFileBig=GmkSSN15R5H4j8e6X1VEXw==, tableContent=null), ArticleFig(id=1159145547677684038, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 5, caption=卷积层工作原理图, figureFileSmall=xBunN6eISocfsGYwvGJ91g==, figureFileBig=GmkSSN15R5H4j8e6X1VEXw==, tableContent=null), ArticleFig(id=1159145547719627079, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 6, caption=Technology roadmap, figureFileSmall=Negk3D/ZXoM1KX30DXr2fw==, figureFileBig=PsKIDjUTqoog/JalCbnZFg==, tableContent=null), ArticleFig(id=1159145547769958728, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 6, caption=技术路线, figureFileSmall=Negk3D/ZXoM1KX30DXr2fw==, figureFileBig=PsKIDjUTqoog/JalCbnZFg==, tableContent=null), ArticleFig(id=1159145547816096073, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 7, caption=Temperature of main bearing of gearbox, figureFileSmall=fTYjOZWrrKCQnMcq12iSDA==, figureFileBig=R/AegYFor/zWl0Nskfi/qg==, tableContent=null), ArticleFig(id=1159145547858039114, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 7, caption=齿轮箱主轴承温度, figureFileSmall=fTYjOZWrrKCQnMcq12iSDA==, figureFileBig=R/AegYFor/zWl0Nskfi/qg==, tableContent=null), ArticleFig(id=1159145547904176459, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 8, caption=Comparison of evaluation results, figureFileSmall=VxorqXrjA+2C1qSU2KgW7A==, figureFileBig=RpNR0GJZwkyZRf+j/VUWfA==, tableContent=null), ArticleFig(id=1159145547950313804, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 8, caption=评估结果对比, figureFileSmall=VxorqXrjA+2C1qSU2KgW7A==, figureFileBig=RpNR0GJZwkyZRf+j/VUWfA==, tableContent=null), ArticleFig(id=1159145547996451149, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Fig. 9, caption=Comparison of prediction results, figureFileSmall=WZNufuFR79QizHNeJp9Vsg==, figureFileBig=1zwJmWFd+fgAToFsF8fvVQ==, tableContent=null), ArticleFig(id=1159145548050977102, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=图 9, caption=预测结果, figureFileSmall=WZNufuFR79QizHNeJp9Vsg==, figureFileBig=1zwJmWFd+fgAToFsF8fvVQ==, tableContent=null), ArticleFig(id=1159145548097114447, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Table 1, caption=Wind turbine condition evaluation index system, figureFileSmall=null, figureFileBig=null, tableContent=
监测类别 监测属性
环境变量 风速、风向、温度、气压等
发电机系统 发电机转子转速、绕组温度、发电机轴承温度等
机舱系统 机舱温度、对风角度、叶片角度等
变桨系统 变桨电机温度、桨叶角度、变桨柜电容温度、变桨 电机温度等
齿轮箱系统 齿轮箱油温、主轴承温度、齿轮箱输入轴承温度等
), ArticleFig(id=1159145548147446096, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=表 1, caption=风电机组状况评价指标体系, figureFileSmall=null, figureFileBig=null, tableContent=
监测类别 监测属性
环境变量 风速、风向、温度、气压等
发电机系统 发电机转子转速、绕组温度、发电机轴承温度等
机舱系统 机舱温度、对风角度、叶片角度等
变桨系统 变桨电机温度、桨叶角度、变桨柜电容温度、变桨 电机温度等
齿轮箱系统 齿轮箱油温、主轴承温度、齿轮箱输入轴承温度等
), ArticleFig(id=1159145548189389137, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Table 2, caption=Classification of wind turbine operation status, figureFileSmall=null, figureFileBig=null, tableContent=
状态等级 状态描述
健康 指标处于正常范围且靠近最佳
良好 指标基本合格
一般 部分指标接近警告阈值, 有逐渐劣化的趋势
预警 指标超出阈值, 有明显的劣化
), ArticleFig(id=1159145548252303698, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=表 2, caption=风电机组运行状态划分, figureFileSmall=null, figureFileBig=null, tableContent=
状态等级 状态描述
健康 指标处于正常范围且靠近最佳
良好 指标基本合格
一般 部分指标接近警告阈值, 有逐渐劣化的趋势
预警 指标超出阈值, 有明显的劣化
), ArticleFig(id=1159145548290052435, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Table 3, caption=Gearbox main bearing temperature and monitoring variable correlation coefficient, figureFileSmall=null, figureFileBig=null, tableContent=
监测变量 相关系数 监测变量 相关系数
风速 0.78 齿轮箱油温 0.62
相对湿度 0.25 发电机绕组温度 0.73
温度 0.40 发电机转速 0.82
气压 0.15 齿轮箱输出轴承温度 0.91
发电机轴承温度 0.76 齿轮箱输入轴承温度 0.89
有功功率 0.64
), ArticleFig(id=1159145548340384084, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=表 3, caption=齿轮箱主轴承温度与监测变量的相关系数, figureFileSmall=null, figureFileBig=null, tableContent=
监测变量 相关系数 监测变量 相关系数
风速 0.78 齿轮箱油温 0.62
相对湿度 0.25 发电机绕组温度 0.73
温度 0.40 发电机转速 0.82
气压 0.15 齿轮箱输出轴承温度 0.91
发电机轴承温度 0.76 齿轮箱输入轴承温度 0.89
有功功率 0.64
), ArticleFig(id=1159145548407492949, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Table 4, caption=Health assessment of wind turbine, figureFileSmall=null, figureFileBig=null, tableContent=
采样 时刻 本文模型 评估 结果 对比模型 评估 结果
T1 $\left\lbrack {1,0,0,0}\right\rbrack$ 健康 $\left\lbrack {{0.95},{0.05},0,0}\right\rbrack$ 健康
T2 $\left\lbrack {1,0,0,0}\right\rbrack$ 健康 $\left\lbrack {{0.95},{0.05},0,0}\right\rbrack$ 健康
T3 $\left\lbrack {{0.97},{0.03},0,0}\right\rbrack$ 健康 $\left\lbrack {{0.9},{0.1},0,0}\right\rbrack$ 健康
T4 $\left\lbrack {{0.8},{0.2},0,0}\right\rbrack$ 健康 $\left\lbrack {{0.8},{0.1},{0.1},0}\right\rbrack$ 健康
T5 $\left\lbrack {{0.48},{0.52},0,0}\right\rbrack$ 良好 $\left\lbrack {{0.6},{0.32},{0.08},0}\right\rbrack$ 健康
T6 $\left\lbrack {0,{0.12},{0.88},0}\right\rbrack$ 一般 $\left\lbrack {{0.48},{0.4},{0.12},0}\right\rbrack$ 健康
T7 $\left\lbrack {0,0,0,1}\right\rbrack$ 预警 $\left\lbrack {0,0,0,1}\right\rbrack$ 预警
), ArticleFig(id=1159145548457824598, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=表 4, caption=风电机组健康状况评估, figureFileSmall=null, figureFileBig=null, tableContent=
采样 时刻 本文模型 评估 结果 对比模型 评估 结果
T1 $\left\lbrack {1,0,0,0}\right\rbrack$ 健康 $\left\lbrack {{0.95},{0.05},0,0}\right\rbrack$ 健康
T2 $\left\lbrack {1,0,0,0}\right\rbrack$ 健康 $\left\lbrack {{0.95},{0.05},0,0}\right\rbrack$ 健康
T3 $\left\lbrack {{0.97},{0.03},0,0}\right\rbrack$ 健康 $\left\lbrack {{0.9},{0.1},0,0}\right\rbrack$ 健康
T4 $\left\lbrack {{0.8},{0.2},0,0}\right\rbrack$ 健康 $\left\lbrack {{0.8},{0.1},{0.1},0}\right\rbrack$ 健康
T5 $\left\lbrack {{0.48},{0.52},0,0}\right\rbrack$ 良好 $\left\lbrack {{0.6},{0.32},{0.08},0}\right\rbrack$ 健康
T6 $\left\lbrack {0,{0.12},{0.88},0}\right\rbrack$ 一般 $\left\lbrack {{0.48},{0.4},{0.12},0}\right\rbrack$ 健康
T7 $\left\lbrack {0,0,0,1}\right\rbrack$ 预警 $\left\lbrack {0,0,0,1}\right\rbrack$ 预警
), ArticleFig(id=1159145548520739159, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Table 5, caption=Parameter of forecast models, figureFileSmall=null, figureFileBig=null, tableContent=
模型 参数
BiLSTM num_layers: 2, cell_size: 32
CNN-BiLSTM filter: $2 \times 2$ , pooling, $2 \times 2$ num_layers: 2, cell_size: 32
机组状况+BiLSTM num_layers: 2, cell_size: 32
机组状况+CNN-BiLSTM filter: $2 \times 2$ , pooling, $2 \times 2$ , num_layer: 2, cell_size: 32
本文模型 filter: $2 \times 2$ , pooling: $2 \times 2$ num_layer: 2, cell_size: 32
), ArticleFig(id=1159145548566876504, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=表 5, caption=预测模型参数, figureFileSmall=null, figureFileBig=null, tableContent=
模型 参数
BiLSTM num_layers: 2, cell_size: 32
CNN-BiLSTM filter: $2 \times 2$ , pooling, $2 \times 2$ num_layers: 2, cell_size: 32
机组状况+BiLSTM num_layers: 2, cell_size: 32
机组状况+CNN-BiLSTM filter: $2 \times 2$ , pooling, $2 \times 2$ , num_layer: 2, cell_size: 32
本文模型 filter: $2 \times 2$ , pooling: $2 \times 2$ num_layer: 2, cell_size: 32
), ArticleFig(id=1159145548617208153, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=EN, label=Table 6, caption=Assessment of wind turbine generator set health status, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE MAE
BiLSTM 8.63 6.37
CNN-BiLSTM 8.38 5.91
机组状况+BiLSTM 8.14 5.82
机组状况+CNN-BiLSTM 7.89 5.76
本文模型 7.13 5.06
), ArticleFig(id=1159145548663345498, tenantId=1146029695717560320, journalId=1146119893612605453, articleId=1152988935132275039, language=CN, label=表 6, caption=风电机组健康状态评估, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE MAE
BiLSTM 8.63 6.37
CNN-BiLSTM 8.38 5.91
机组状况+BiLSTM 8.14 5.82
机组状况+CNN-BiLSTM 7.89 5.76
本文模型 7.13 5.06
)], attaches=null, journal=Journal(id=1146118781765529606, delFlag=0, nameCn=可再生能源, nameEn=Renewable Energy Resources, nameHistory1=null, nameHistory2=null, issn=1671-5292, eissn=, cn=21-1469/TK, coden=null, periodic=0, language=CN, oaType=null, 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=JwZjzvOuLF6gg7XdYa5wzg==, journalPrice=null, startedYear=null, abbrevIsoEn=Renew Energy Res, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1755588886026, createdBy=null, updatedBy=15831073675, firstLetterCn=R, firstLetterEn=R, subjectCode=Engineering, subjectName=工程, subjectCodeEn=Engineering, subjectNameEn=null, picCn=JwZjzvOuLF6gg7XdYa5wzg==, picEn=JbD+CPZ3qeyTMjY1NLqExg==, jcr=null, cjcr=null, exts=[JournalExt(id=1164587454112678277, language=CN, name=可再生能源, 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://kzsny.com/, createdTime=1755588886049, updatedTime=1755588886049, createdBy=15831073675, updatedBy=15831073675, submissionGuidelinesUrl=https://ncny.cbpt.cnki.net/EditorEN/PromptPageInfo.aspx?t=v&c=1, submissionAuthorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=1, submissionEditorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=3, submissionReviewUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=2, submissionCeEditorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=3, submissionAeEditorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=3, option={"copyright":""}), JournalExt(id=1164587454167204230, language=EN, name=Renewable Energy Resources, 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=, createdTime=1755588886062, updatedTime=1755588886062, createdBy=15831073675, updatedBy=15831073675, submissionGuidelinesUrl=, submissionAuthorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=1, submissionEditorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=3, submissionReviewUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=2, submissionCeEditorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=3, submissionAeEditorUrl=https://ncny.cbpt.cnki.net/EditorEN/index.aspx?t=3, option={"copyright":""})], databaseList=null, tenantJournalId=1146119893612605453, websiteList=[Website(id=1148243202370429360, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146119893612605453, 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/kzsny/CN, language=CN, createTime=1751692112772, createBy=18614031015, updateTime=1753519535151, updateBy=18614031015, name=《可再生能源》-中文站点, tplId=1146099689490845704, title=可再生能源, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148619523667816493, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202370429360, code=articleTextType, value=kx, createTime=1751781834761, updateTime=1751781834761, creator=18614031015, updator=18614031015), WebsiteProps(id=1148619523642650666, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202370429360, code=banner, value=null, createTime=1751781834755, updateTime=1751781834755, creator=18614031015, updator=18614031015), WebsiteProps(id=1148619523634262057, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202370429360, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=y1rtJt0tFAqUB4W3g4tlSw==, createTime=1751781834753, updateTime=1751781834753, creator=18614031015, updator=18614031015), WebsiteProps(id=1148619523659427884, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202370429360, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751781834759, updateTime=1751781834759, creator=18614031015, updator=18614031015), WebsiteProps(id=1148619523651039275, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202370429360, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751781834757, updateTime=1751781834757, creator=18614031015, updator=18614031015)]), Website(id=1155908164001353748, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146119893612605453, 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/kzsny/EN, language=EN, createTime=1753519582037, createBy=18614031015, updateTime=1753519582037, updateBy=18614031015, name=《可再生能源》-英文站点, tplId=1146101810881728533, title=Renewable Energy Resources, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155908456914739359, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155908164001353748, code=articleTextType, value=kx, createTime=1753519651869, updateTime=1753519651869, creator=18614031015, updator=18614031015), WebsiteProps(id=1155908456893767836, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155908164001353748, code=banner, value=null, createTime=1753519651864, updateTime=1753519651864, creator=18614031015, updator=18614031015), WebsiteProps(id=1155908456885379227, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155908164001353748, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=y1rtJt0tFAqUB4W3g4tlSw==, createTime=1753519651862, updateTime=1753519651862, creator=18614031015, updator=18614031015), WebsiteProps(id=1155908456906350750, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155908164001353748, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753519651867, updateTime=1753519651867, creator=18614031015, updator=18614031015), WebsiteProps(id=1155908456902156445, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155908164001353748, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753519651866, updateTime=1753519651866, creator=18614031015, updator=18614031015)])], journalTitle=可再生能源, weixinUrl=null, journalUrl=null, iacademicId=null, status=0, seqNo=null, journalTitleEn=Renewable Energy Resources, journalPhotoCn=JwZjzvOuLF6gg7XdYa5wzg==, journalPhotoEn=JbD+CPZ3qeyTMjY1NLqExg==, journalFirstLetter=R, 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/kzsny/CN/Y2025/V43/I2/217, detailUrlEn=https://castjournals.cast.org.cn/joweb/kzsny/EN/Y2025/V43/I2/217, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kzsny/CN/PDF/Y2025/V43/I2/217, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kzsny/EN/PDF/Y2025/V43/I2/217, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
考虑风电机组健康状况与双重注意力机制 CNN – BILSTM 的超短期功率预测
收藏切换
PDF下载
张开伟 1 , 文中 1 , 杨生鹏 1 , 胡梓涵 1 , 丁剑 2
可再生能源 | 2025,43(2): 217-224
收起
收藏切换
可再生能源 | 2025, 43(2): 217-224
考虑风电机组健康状况与双重注意力机制 CNN – BILSTM 的超短期功率预测
全屏
张开伟1, 文中1 , 杨生鹏1, 胡梓涵1, 丁剑2
作者信息
  • 1 三峡大学 电气与新能源学院 湖北 宜昌 443002
  • 2 上海勘测设计研究院有限公司 上海 200434

通讯作者:

文中(1968-),男,硕士,副教授,研究方向为输电线路工程技术、电力系统运行与控制。E-mail:
Considering the health status of wind turbines and the dual attention mechanism CNN-BiLSTM ultra-short-term power prediction
Kaiwei Zhang1, Zhong Wen1 , Shengpeng Yang1, Zihan Hu1, Jian Ding2
Affiliations
  • 1 School of Electrical and New Energy Three Gorges University Yichang 443002 China
  • 2 Shanghai Survey and Design Institute Co., Ltd. Shanghai 200434 China
出版时间: 2025-02-20
文章导航
收藏切换

为提升风电机组超短期功率预测的准确性,文章提出了一种考虑风电机组健康状况与双重注意力机制 CNNBiLSTM 的超短期功率预测模型。首先,综合考虑环境因素与风电机组各子部件的相互作用对风电机组输出功率的影响,将风电机组各个子部件正常运行时的相对误差作为监测指标的劣化度;然后,采用模糊综合评价法对风电机组健康状况进行评估,根据评估结果对其历史数据集进行健康状况划分;最后,采用双重注意力机制 CNNBiLSTM 模型对分类后的数据集构建超短期功率预测模型。实验结果表明,在风电机组功率预测过程中,相较于未考虑机组健康状况,考虑机组健康状况的均方根误差(RMSE)和平均绝对误差(MAE)分别降低了17.3%和20.5%。

超短期  /  功率预测  /  健康状况  /  双重注意力机制  /  CNN-BiLSTM 模型

In order to improve the accuracy of ultrashortterm power prediction of wind turbines, this paper proposes a CNNBiLSTM ultrashortterm power prediction method considering the health status of wind turbines and dual attention mechanism. Firstly, considering the influence of the interaction between the environmental factors and the components of the wind turbine on the output power of the wind turbine, he relative error of the normal operation of each component of the wind turbine is used as the deterioration degree of the monitoring index. Secondly, the fuzzy comprehensive evaluation method assesses the health of wind turbines, and the historical data set is categorized based on the evaluation results. Finally, the dual attention mechanism CNN BiLSTM model is used to construct an ultrashortterm power prediction model for the classified data set. The experimental results show that the RMSE and MAE considering the health status of wind turbines are reduced by 17.3% and 20.5% respectively compared with the RSME and MSE without considering the health status of wind turbines.

ultra-short-term  /  power prediction  /  health status  /  dual attention mechanism  /  CNN-BILSTM model
张开伟, 文中, 杨生鹏, 胡梓涵, 丁剑. 考虑风电机组健康状况与双重注意力机制 CNN – BILSTM 的超短期功率预测. 可再生能源, 2025 , 43 (2) : 217 -224 .
Kaiwei Zhang, Zhong Wen, Shengpeng Yang, Zihan Hu, Jian Ding. Considering the health status of wind turbines and the dual attention mechanism CNN-BiLSTM ultra-short-term power prediction[J]. Renewable Energy Resources, 2025 , 43 (2) : 217 -224 .
经过数十年的发展, 风力发电技术日趋成熟, 在新型电力系统建设背景下, 风电的开发利用仍将保持高速增长。预计到 2060 年,全社会用电量将超过 16 万亿 $\mathrm{{kW}} \cdot \mathrm{h}$ ,新能源发电装机容量将达到 50 亿 kW,新能源发电量占比将由目前的 8% 提高到 ${60}\%$ 以上 [ 1 , 2 ] 。然而,由于风力发电的随机性和不确定性, 风电并网后会给电力系统调频、调峰以及电网调度等带来诸多挑战 [ 3 ] 。因此,对风电功率进行准确地超短期预测, 对于提高风电的消纳、保障电网安全稳定运行至关重要 [ 4 ] 。国内外针对超短期风电功率预测进行了大量的研究, 相关预测方法已从最初的物理模型法、统计推理法演变至以深度学习为代表的预测方法 [ 5 ] 。作为深度学习模型的代表, 长短期记忆网络(LSTM)因其独特的记忆功能和出色的非线性表征能力, 在时间序列预测领域备受关注。文献 [ 6 ] 提出的基于 LSTM 的多变量超短期预测方法较传统机器学习算法的预测精度更高。文献[ 7 ]基于嵌入定理从风电功率序列中重构出相空间, 并对组合重构后的风电功率数据采用双向 LSTM 进行预测。上述模型虽然能够较为全面地考虑各种因素对风电功率的影响, 但模型结构相对复杂, 对数据具有较强的依赖性。此外,这些预测方法主要对正常运行条件下的风电机组进行研究。
自投产之时,风电机组的叶片、齿轮箱、发电机等部件就处于逐渐劣化阶段。由于风电机组长期运行在恶劣环境中, 这些部件的劣化速度可能远超预期。现有研究普遍认为, 风电机组的监测数据中包含了与风电机组健康状况密切相关的信息, 在构建预测模型时, 往往未能将风电机组健康状况作为独立因素来考量, 这也是造成对新投产的风电机组功率预测准确性往往高于已运行多年的老机组的主要原因。因此, 在进行风电机组的功率预测时,考虑健康状况是非常重要的。也有少数学者对此进行了研究, 文献[ 8 ]通过随机矩阵计算出的平均谱半径来衡量单台风电机组的健康状况。文献[ 9 ]通过高斯混合模型和多状态特征融合技术建立健康状况评估模型, 并采用健康衰退指数作为风电机组健康状况的评价指标。文献[ 10 , 11 ] 通过劣化度算法、组合赋权确定权重,以及基于准则的云模型对风电机组健康状况进行评价。这些研究均表明, 风电机组健康状况对功率预测有着重要影响, 然而风电机组所处环境的风速波动较大, 导致高斯混合模型以及隶属云模型在评估风电机组的劣化程度时具有一定的滞后性。 已有评估模型仅关注当前时刻的数值, 没有考虑到风电机组的劣化是随时间变化的。
风电机组从健康状态到出现故障是一个持续变化的过程, 这种动态变化在现有的评估模型中并未得到充分的体现。因此, 本文提出了一种考虑风电机组健康状况与双重注意力机制 CNN-BiL-STM 的超短期功率预测模型。引入动态劣化度计算方法, 综合考虑当前时刻及附近时刻各项指标的影响, 从而动态地反映出风电机组的健康状况。 采用模糊综合评估法将风电机组的健康状况评估问题转化为模糊决策问题,针对评估结果构建双重注意力机制 CNN-BiLSTM 预测模型。最后通过算例对比分析, 验证本文所提模型的优势。
风速-功率曲线是评估风电机组运行性能的重要依据。在风电机组出厂之前,厂家会提供标准功率曲线图(图 1)。
图 2 为健康状态和故障状态下风电机组的风速-功率曲线。
图 2 可知: 尽管型号相同的两台风电机组处于同一地理位置, 由于运行状态不同, 使得相同风速下的输出功率有明显差异;当风速为 $8\mathrm{\;m}/\mathrm{s}$ 时,两台机组的功率差约为 ${110}\mathrm{\;{kW}}$ ; 当风速达到 ${12}\mathrm{m}/\mathrm{s}$ 时,功率差增至约 ${240}\mathrm{{kW}}$ 。因此,在进行风电机组超短期功率预测时, 充分考虑健康状态非常重要。
本文综合考虑环境因素和风电机组各子系统运行状态, 对风电机组健康状况进行评估。选择发电机系统、机舱系统、变桨系统、齿轮箱系统作为风电机组的子系统层。从每个子系统中挑选出能够有效反映其健康状况的变量共 23 个,环境变量 8 个, 根据子系统层变量与环境变量建立风电机组的层次结构(表 1)。
在进行风电机组健康状况评估时, 首先, 将 BiLSTM 模型预测误差作为子系统变量劣化度。将每个子系统变量劣化度代入岭形分布隶属度函数,逐层计算各子系统隶属度矩阵;然后,采用组合赋权法确定各子系统层和各子系统变量权重; 最后,根据隶属度矩阵与组合权重向量乘积评估出风电机组健康状况,评估流程如图 3 所示。
根据评估结果将风电机组健康状况划分为 4 个健康等级 (表 2), 不同健康状况下风电机组对风能的转化效率不同,预警时转化效率最差。
在积累的大量风电机组数据下, 采用子系统变量的相对误差建立风电机组子系统变量与正常运行的映射关系。本文定义子系统变量的预测值与实际值的相对误差${\varepsilon }_{i}$来反映子系统变量的劣化度。
${\varepsilon }_{i} = \left| \frac{{y}^{\prime } - {y}_{i}}{{y}_{i}}\right| , i \in N$
式中:${y}^{\prime }$为子系统变量的预测值;${y}_{i}$为子系统变量的实际值;$i$为计算样本数,$N = 1,2,3\cdots$
为了方便后续的计算与分析, 本文将各个子系统变量劣化度值归一到$\left\lbrack {0,1}\right\rbrack$,归一化后的劣化度为
${g}_{i} = \left\{ \begin{array}{ll} 1 & {\varepsilon }_{i} > {\varepsilon }_{\max } \\ \frac{{\varepsilon }_{i}}{{\varepsilon }_{\max }} & {\varepsilon }_{i} < {\varepsilon }_{\max } \end{array}\right.$
式中:${g}_{i}$为子系统变量劣化度;${\varepsilon }_{\max }$为子系统变量允许劣化最大值。
对风电机组各个子系统变量进行劣化度分析,并归一化得到介于$\left\lbrack {0,1}\right\rbrack$的数字量,但数字量不能直观地反映风电机组的实际运行状态。因此, 采用岭形隶属度函数将数字量转化为更具描述性的状态量。为充分覆盖子系统变量劣化度, 本文采用了升岭型分布函数、中间型分布函数以及降岭型分布函数, 细化了劣化度的评估标准。具体的分布函数为
${r}_{s1}\left( {g}_{i}\right) = \left\{ \begin{array}{ll} 1, & {0.1} \geq {g}_{i} \\ \frac{1}{2} - \frac{1}{2}\sin \frac{\pi }{0.2}\left( {{g}_{i} - {0.2}}\right) , & {0.1} < {g}_{i} \leq {0.3} \\ 0, & {0.3} < {g}_{i} \end{array}\right.$
${r}_{s2}\left( {g}_{i}\right) = \left\{ \begin{array}{ll} 0, & {0.1} \geq {g}_{i} \\ \frac{1}{2} + \frac{1}{2}\sin \frac{\pi }{0.2}\left( {{g}_{i} - {0.2}}\right) , & {0.1} < {g}_{i} \leq {0.3} \\ 1, & {0.3} < {g}_{i} \leq {0.4} \\ \frac{1}{2} - \frac{1}{2}\sin \frac{\pi }{0.2}\left( {{g}_{i} - {0.2}}\right) , & {0.4} < {g}_{i} \leq {0.6} \\ 0, & {0.6} < {g}_{i} \end{array}\right.$
${r}_{s3}\left( {g}_{i}\right) = \left\{ \begin{array}{ll} 0, & {0.4} \geq {g}_{i} \\ \frac{1}{2} + \frac{1}{2}\sin \frac{\pi }{0.2}\left( {{g}_{i} - {0.5}}\right) , & {0.4} < {g}_{i} \leq {0.6} \\ 1, & {0.6} < {g}_{i} \leq {0.7} \\ \frac{1}{2} - \frac{1}{2}\sin \frac{\pi }{0.2}\left( {{g}_{i} - {0.8}}\right) , & {0.7} < {g}_{i} \leq {0.9} \\ 0, & {0.9} < {g}_{i} \end{array}\right.$
${r}_{s4}\left( {g}_{i}\right) = \left\{ \begin{array}{ll} 0, & {0.7} \geq {g}_{i} \\ \frac{1}{2} + \frac{1}{2}\sin \frac{\pi }{0.2}\left( {{g}_{i} - {0.8}}\right) , & {0.7} < {g}_{i} \leq {0.9} \\ 1, & {0.9} < {g}_{i} \end{array}\right.$
式中:${r}_{sl}$为第$i$个指标在评价等级$l$的隶属度,$l =$1,2,3,4。
结合式 (3)$\sim \left( 6\right)$,根据子系统$s$中的变量计算出子系统隶属度矩阵。
${\mathbf{R}}_{s} = \left\lbrack \begin{matrix} {r}_{s1}\left( {g}_{i1}\right) & {r}_{s2}\left( {g}_{i1}\right) & {r}_{s3}\left( {g}_{i1}\right) & {r}_{s4}\left( {g}_{i1}\right) \\ {r}_{s1}\left( {g}_{i2}\right) & {r}_{s2}\left( {g}_{i2}\right) & {r}_{s3}\left( {g}_{i2}\right) & {r}_{s4}\left( {g}_{i2}\right) \\ \vdots & \vdots & \vdots & \vdots \\ {r}_{s1}\left( {g}_{in}\right) & {r}_{s2}\left( {g}_{in}\right) & {r}_{s3}\left( {g}_{in}\right) & {r}_{s4}\left( {g}_{in}\right) \end{matrix}\right\rbrack$
图 4 为各等级下的隶属度函数曲线。
由于不同子系统层和子系统变量在风电机组健康状态评估中的重要性不同, 本文采用层次分析法来确定各子系统层和子系统变量的主观权重,并使用熵权法确定各子系统层和子系统变量的客观权重。这种权重的综合运用可以更准确地考虑到各指标的重要性, 从而更全面地评估风电机组的运行状况。组合权重表达式为
${w}_{j} = \frac{{w}_{j}^{\prime }{w}_{j}^{ * }}{\mathop{\sum }\limits_{{j = 1}}^{m}{w}_{j}^{\prime }{w}_{j}^{ * }}$
式中:${w}_{j}^{\prime }$为层次分析法计算的各项权重;${w}_{j}^{ * }$为熵权法计算的各项权重;${w}_{j}$为组合赋权法所得权重$(j =$$1,2,\cdots , m)$
由式 (7),(8)得到子系统$s$的评估结果为
${B}_{s} = {\mathbf{W}}_{s} \times {\mathbf{R}}_{s}$
式中:${\mathbf{W}}_{s}$为风电机组各个指标的权重向量。
由子系统的评估结果, 进而评估出风电机组的健康状况。
$B = W \times \left\lbrack \begin{matrix} {B}_{1} \\ \vdots \\ {B}_{s} \end{matrix}\right\rbrack = \left\lbrack \begin{array}{llll} {c}_{1} & {c}_{2} & {c}_{3} & {c}_{4} \end{array}\right\rbrack$
式中:${c}_{1},{c}_{2},{c}_{3},{c}_{4}$分别为健康、良好、一般、预警 4 个隶属度值。
根据最大隶属度原则,选取式 (10) 中对应的最大值作为风电机组的健康状况,即:
$V = \{ v \mid v = \max \left( B\right) \}$
CNN 是一种具有卷积计算的深度结构前馈神经网络, 有着强大的特征提取能力。在 CNN 中, 卷积层是其核心组成部分, 卷积层工作原理如图5 所示。通过使用不同的卷积核对输入数据进行计算, 从而得到包含不同特征的映射矩阵。
BiLSTM 由一层正向 LSTM 和一层反向 LSTM 组成, 能够更好地挖掘时间序列正向与反向依赖关系, 解决了 LSTM 递归神经网络在时序反向传播训练过程中梯度消失和梯度爆炸的问题, 对时间序列具有良好预测性能 [ 12 ] 。BiLSTM 各变量之间的计算式为
${f}_{t} = \sigma \left( {{W}_{f} \times \left\lbrack {{h}_{t - 1},{X}_{t}}\right\rbrack + {b}_{f}}\right)$
${i}_{t} = \sigma \left( {{W}_{i} \times \left\lbrack {{h}_{t - 1},{X}_{t}}\right\rbrack + {b}_{i}}\right)$
${\widetilde{C}}_{t} = \tanh \left( {{W}_{c} \times \left\lbrack {{h}_{t - 1},{X}_{t}}\right\rbrack + {b}_{c}}\right)$
${C}_{t} = {f}_{t} \times {C}_{t - 1} + {i}_{t} \times {\widetilde{C}}_{t}$
${o}_{t} = \sigma \left( {{W}_{o} \times \left\lbrack {{h}_{t - 1},{X}_{t}}\right\rbrack + {b}_{o}}\right)$
${h}_{t} = {o}_{t} \times \tanh \left( {C}_{t}\right)$
式中:${f}_{l}$为遗忘门;${i}_{l}$为输入门;${\widetilde{C}}_{l}$为临时记忆单元;${C}_{t}$为记忆单元;${C}_{t - 1}$为上一时刻记忆单元;${o}_{t}$为输出单元;${W}_{f},{W}_{i},{W}_{c}$均为权重阈值;${b}_{f},{b}_{i},{b}_{c},{b}_{o}$均为偏置;$\sigma$, tanh 均为激活函数;${X}_{t}$为预测模型的输入值;${h}_{t}$为预测模型的输出值。
注意力机制是一种通过算法来模拟人脑注意力的模型, 通过对不同输入特征分配不同的权重, 能够减少信息冗余 [ 13 ] 。注意力机制有利于提高模型特征提取的预测精度, 在一定程度上提高了模型的可解释性。
${\tau }_{t} = \mathbf{u}\tanh \left( {\mathbf{w} \cdot {\mathbf{h}}_{t} + \mathbf{b}}\right)$
${\alpha }_{t} = \frac{\exp \left( {\tau }_{t}\right) }{\mathop{\sum }\limits_{{j = 1}}^{t}{\tau }_{j}}$
式中:${\tau }_{t}$$t$时刻的注意力权重;${\mathbf{h}}_{t}$$t$时刻的特征向量;$\mathbf{u},\mathbf{w}$均为权重向量;${\alpha }_{t}$为归一化注意力权重;$\mathbf{b}$为偏置向量。
本文提出的考虑风电机组健康状况与双重注意力机制 CNN-BiLSTM 的超短期功率预测技术路线的总体思路如下:
①基于风电场收集的大量风电机组运行数据, 对风电机组进行健康状况评估, 根据评估结果对风电机组进行健康状况划分;
②基于 encoder-decoder 框架的双阶段注意力机制对 CNN-BiLSTM 模型进行改进。将特征注意力机制融入 CNN 模块中, 作为特征注意力模块编码器。通过 CNN 卷积层、池化层和注意力机制对输入特征权重进行提取, 能够达到自适应提取序列特征信息的目的;
③ 将时间注意力机制融入到 BiLSTM 模型中作为解码器, 通过 BiLSTM 对时间序列隐含信息的全面学习, 并结合时间注意力机制能够自适应地处理历史序列信息, 增强关键时刻信息的提取能力。
本文提出的技术路线如图 6 所示。
特征注意力模块由两层一维卷积结构和注意力机制组成。首先,输入数据经过卷积层进行特征提取与映射增加数据维度; 然后, 通过池化层实现数据的稀疏处理, 降低数据复杂度与运算量; 最后, 在全连接层进行特征融合与决策, 减少输入特征信息在传递过程中丢失。其组成结构为
${\alpha }_{f} = \operatorname{Attention}\left( {H}_{a}\right)$
${z}_{f} = {\alpha }_{f}{H}_{a}$
式中:${\alpha }_{f}$为特征权重;Attention$\left( \cdot \right)$为注意力机制;${H}_{a}$$\mathrm{{CNN}}$全连接层输出;${z}_{f}$为注意力模块输出。
时间注意力模块由 BiLSTM 与注意力机制共同组成, 可对特征注意力模块输出量进行解码。首先,利用 BiLSTM 对特征注意力模块输出${\alpha }_{f}$进行双向学习; 然后,采用时间注意力模块根据$t$个时间步的历史节点对当前时间步的影响程度, 自适应地为 BiLSTM 输出的隐含特征赋予不同权重。
时间注意力模块组成结构为
${h}_{t}^{ + } = {\mathrm{{LSTM}}}^{ + }\left( {{h}_{t - 1},{z}_{f, t},{c}_{t - 1}}\right)$
${h}_{t} = {\operatorname{LSTM}}^{ - }\left( {{h}_{t + 1},{z}_{f, t},{c}_{t + 1}}\right)$
${h}_{b} = {\mathbf{W}}_{h} \cdot {h}^{ + } + {\mathbf{W}}_{h}^{\prime }{h}^{ - } + {\mathbf{b}}_{h}$
${\alpha }_{s} = \operatorname{Attention}\left( {H}_{b}\right)$
$r = {\alpha }_{s}{H}_{b}$
$P = \sigma \left( {{W}_{r}r + {b}_{r}}\right)$
式中:${h}_{t}^{ + },{h}_{t}^{ - }$分别为$t$时刻BiLSTM 的正、反向隐藏状态;${\mathbf{W}}_{h},{\mathbf{W}}_{h}^{\prime }$分别为 BiLSTM 正、反向权重矩阵;${\mathbf{b}}_{h}$为偏置向量;$P$为预测结果;${W}_{r}$为全连接层权重;${b}_{r}$为全连接层偏置。
对泌阳某风电场 2018 年的监测数据集进行预测分析。数据集中除气象数据外, 还包括评估风电机组健康状况的多种状态变量表征。选取 1~5 月风电机组发生故障前的数据作为训练数据集, 06 月 01 日 00:00(T1)-06 月 02 日 00:00(T7) 风电机组运行数据作为待评价参数, 对评估模型进行验证。评估时间间隔为$4\mathrm{\;h}$,已知该风电机组 06 月 02 日 00:00 后停机检修。数据处理与模型构建过程均在 PyCharm Community Edition 2022.2.3 软件中实现。
以齿轮箱主轴承温度异常为例, 对风电机组健康状况进行评估与分析。鉴于 1~5 月风电机组整体运行状况稳定, 故将这一时段的运行数据作为风电机组正常运行数据。从而分析得到各个子系统变量对应的阈值和权重,并通过 Pearson 相关系数法计算齿轮箱主轴承温度与各类监测变量之间的相关性(表 3)。
选取相关系数大于 0.6 的监测变量作为预测模型的输入, 得到齿轮箱主轴承温度变化曲线(图 7)。
按照传统判别方法, 齿轮箱主轴承温度属于越小越优型,所以在采样点 78 附近状态最优。由本文所采用的方法可知,当采样点在 0~60 时,预测值与实际值误差最小, 说明在这一阶段齿轮箱主轴承性能最好,尚未出现劣化迹象。当采样点在 60~100 时,预测值开始偏离实际值,表明齿轮箱主轴承开始发生劣化。在采样点 100 以后,劣化趋势愈发明显。因此,采样点 78 附近被视为齿轮箱主轴承劣化的初始阶段,而非最佳状态。
表 4 为风电机组健康状况评估结果。
图 7表 4 可知:在$\mathrm{T}1,\mathrm{\;T}2,\mathrm{\;T}3$时刻,模型预测值与实际值误差较小, 齿轮箱主轴承处于最优状态, 风电机组评估为健康状态; 在 T4 时刻, 齿轮箱主轴承开始劣化, 风电机组隶属度矩阵由 0.97 逐渐降低至 0.80 ,但评估结果仍为健康状态; 在 T5, T6 时刻, 齿轮箱主轴承劣化程度明显加剧,导致风电机组健康状况由良好转变为一般,表明齿轮箱主轴承劣化已经开始对风电机组整体运行性能产生负面影响;在 T7 时刻后,齿轮箱主轴承温度超过限定值,导致风电机组进入预警状态。 对比单一权重的传统劣化度计算方法, 本文模型的结果具有一致性。
图 8 为评估结果对比曲线。
图 8 可知, 单一权重评估模型对风电机组的劣化评估具有迟滞性,不能及时有效地反映出风电机组的劣化程度, 本文模型能够有效地反映风电机组在不同时间点的性能状况变化。
为验证本文模型的有效性,构建多个预测模型进行对比分析, 包括 BiLSTM 预测模型、结合卷积神经网络的 CNN-BiLSTM 预测模型, 以及在这两种模型基础上考虑风电机组健康状况的变种模型。各预测模型的参数见表 5,预测结果如图 9 所示。
图 9 可知:相较于其他模型,本文模型具有更高的预测精度, 特别是在风电机组健康状况发生变化的时刻; 在 T1-T4 时刻, 风电机组处于健康状况, 各预测模型的预测值均能较为准确地贴合实际功率曲线; 在 T5 时刻, 风电机组处于良好状况, 未考虑风电机组健康状况的预测模型的预测值相较于实际功率曲线出现了明显的偏差; 在 T6, T7 时刻以后,风电机组性能进一步恶化,导致未考虑风电机组健康状况的预测模型的误差逐渐累积并扩大, 本文模型由于融入了健康状况信息, 模型的预测值仍能与实际功率曲线的变化趋势保持高度吻合。
为了准确地评估各预测模型的性能, 采用均方根误差(RMSE)、平均绝对误差(MAE)作为预测模型的评价指标 [ 14 ]
${E}_{\mathrm{{RMSE}}} = \sqrt{\frac{1}{n}\mathop{\sum }\limits_{{t = 1}}^{n}{\left( {y}_{i} - {\widehat{y}}_{i}\right) }^{2}}$
${R}_{\mathrm{{MAE}}} = \frac{1}{n}\mathop{\sum }\limits_{{t = 1}}^{n}\left| {{y}_{i} - {\widehat{y}}_{i}}\right|$
式中:${\widehat{y}}_{i}$为第$i$个预测值;${y}_{i}$为第$i$个真实值,$i =$$1,2,\cdots {n}_{ \circ }$
表 6 为风电机组健康状态评估表。
表 6 可知: 在考虑风电机组健康状态的情况下, 相较于 CNN-BiLSTM 模型, 本文模型的 RMSE 和 MAE 分别降低了 9.6%和 12%,主要是由于注意力机制突出了重要特征和重要时间点的影响;与仅考虑风电机组健康状况的 BiLSTM 模型相比,本文模型的 RMSE 和 MAE 分别降低了 12.4%和 13%;与 BiLSTM 模型相比,本文模型的 RMSE 和 MAE 分别降低了 17.3% 和 20.5%;与 CNN-BiLSTM 模型相比, 本文模型的 RMSE 和 MAE 分别降低了 14.9%和 14.3%。
本文提出了一种考虑风电机组健康状况与双重注意力机制 CNN-BiLSTM 的超短期功率预测模型。深入分析了风电机组的健康状况对输出功率的影响, 得到以下结论。
①在风电机组健康状况评估中,采用劣化度计算方法, 可以有效地削弱各部件之间的相互干扰以及环境因素对评估结果产生的潜在影响。同时, 采用模糊综合评估法能够更深入地了解风电机组的状态转化过程。
②考虑风电机组健康状况,运用 CNN 有效地提取数据特征,并融入 BiLSTM 模型中,能够显著提升功率预测精度。与仅考虑机组健康状态的 BiLSTM 模型相比, 本文模型的 RMSE 和 MAE 分别降低了 12.4%和 13%。
③在 CNN-BiLSTM 模型的基础上, 考虑风电机组健康状况, 增加特征注意力模块和时间注意力模块, 可突出重要特征和重要时间点的影响, 对模型预测精度有显著提升。
  • 国家自然科学基金项目(52107108)
参考文献 引证文献
排序方式:
[1]
马伟明. 关于电工学科前沿技术发展的若干思考[J]. 电工技术学报, 2021, 36(22): 4627-4636.
[2]
舒印彪. 发展新型电力系统助力实现“双碳”目标[J]. 中国电力企业管理, 2021(7): 8-9.
[3]
Wang S, Sun Y H, Zhang S M, et al. Very short-term probabilistic prediction of PV based on multi -period error dis-tribution[J]. Electric Power Systems Research, 2023, 214: 108817.
[4]
李军徽, 张嘉辉, 李翠萍, 等. 参与调峰的储能系统配置方案及经济性分析[J]. 电工技术学报, 2021, 36(19): 4148-4160.
[5]
张红涛, 韩婧, 谭联, 等. 基于 FCM 和 SSA-ELM 的超短期风功率预测[J]. 工程科学与技术, 2020, 52(6): 234-241.
[6]
潘鹏程, 刘晖, 王仁明. 自适应密度聚类组合数据清洗的LSTM 风电功率预测[J]. 电力系统及其自动化学报, 2024, 36(7): 59-66.
[7]
邓韦斯, 卢斯煜, 刘显茁, 等. 基于相空间重构和 BiLSTM 的风电功率短期预测[J]. 广东电力, 2023, 36(7): 22-30.
[8]
杨茂, 周宜. 计及风电场状态的风电功率超短期预测[J]. 中国电机工程学报, 2019, 39(5): 1259-1268.
[9]
张静, 李柠, 李少远, 等. 基于数据的风电机组发电机健康状况评估[J]. 信息与控制, 2018, 47(6): 694-713.
[10]
孙海蓉, 王皓茹, 王瑞珈, 等. 基于云模型的风电机组运行状态评价[J]. 电力科学与工程, 2020, 36(8): 57-62.
[11]
S Chen, X Guo, S Han, et al. A novel stochastic cloud model for statistical characterization of wind turbine output[J]. IEEE Access, 2021(9): 7439-7446.
[12]
符杨, 任子旭, 魏书荣, 等. 基于改进 LSTM-TCN 模型的海上风电超短期功率预测[J]. 中国电机工程学报, 2022, 42(12): 4292-4303.
[13]
王渝红, 史云翔, 周旭, 等. 基于时间模式注意力机制的BiLSTM 多风电机组超短期功率预测[J]. 高电压技术, 2022, 48(5): 1884-1892.
[14]
凡航, 张雪敏, 梅生伟, 等. 基于时空神经网络的风电场超短期风速预测模型[J]. 电力系统自动化, 2021, 45(1): 28-35.
2025年第43卷第2期
PDF下载
254
132
引用本文
BibTeX
文章信息
  • 接收时间:2024-05-14
  • 首发时间:2025-07-18
  • 出版时间:2025-02-20
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-05-14
基金
国家自然科学基金项目(52107108)
作者信息
    1 三峡大学 电气与新能源学院 湖北 宜昌 443002
    2 上海勘测设计研究院有限公司 上海 200434

通讯作者:

文中(1968-),男,硕士,副教授,研究方向为输电线路工程技术、电力系统运行与控制。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kzsny/CN/1152988935132275039
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