Article(id=1190337957757292996, tenantId=1146029695717560320, journalId=1189987059142926344, issueId=1190337956201202212, articleNumber=null, orderNo=null, doi=10.19457/j.1001-2095.dqcd25932, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1715529600000, receivedDateStr=2024-05-13, revisedDate=1719763200000, revisedDateStr=2024-07-01, acceptedDate=null, acceptedDateStr=null, onlineDate=1761728284539, onlineDateStr=2025-10-29, pubDate=1755619200000, pubDateStr=2025-08-20, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1761728284539, onlineIssueDateStr=2025-10-29, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1761728284539, creator=13701087609, updateTime=1761728284539, updator=13701087609, issue=Issue{id=1190337956201202212, tenantId=1146029695717560320, journalId=1189987059142926344, year='2025', volume='55', issue='8', pageStart='3', pageEnd='96', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=0, createTime=1761728284168, creator=13701087609, updateTime=1761728464442, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1190338712388079738, tenantId=1146029695717560320, journalId=1189987059142926344, issueId=1190337956201202212, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1190338712388079739, tenantId=1146029695717560320, journalId=1189987059142926344, issueId=1190337956201202212, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=58, endPage=69, ext={EN=ArticleExt(id=1190337958004756934, articleId=1190337957757292996, tenantId=1146029695717560320, journalId=1189987059142926344, language=EN, title=Extreme Weather Photovoltaic Power Ultra-short-term Forecasting Based on CGAN and CNN-SE-BiLSTM, columnId=null, journalTitle=Electric Drive, columnName=null, runingTitle=null, highlight=null, articleAbstract=

A K-means clustering algorithm was proposed and a conditional Wasserstein generative adversarial network with gradient penalty(CWGAN-GP)to address the problem of imbalanced photovoltaic generation data caused by the low occurrence probability of extreme weather. A prediction approach combining bidirectional long short-term memory(BFLSTM)with convolutional neural network was introduced and incorporating channel attention mechanism to enhance the PV power prediction performance by integrating spatio-temporal features and dynamically adjusting the importance of feature channels. Firstly,correlation analysis and K-means algorithm were utilized to select and label various environmental factors. Then,extreme weather labels with fewer samples after clustering were selected,and CWGAN-GP was used for data augmentation.Finally,the augmented dataset was used to train the CNN-SE-BiLSTM prediction model for PV power prediction under extreme weather conditions.Simulation modeling was conducted using data from a certain PV power station,and the results demonstrate that augmenting the original extreme weather training set with CGAN-GP helps improve the prediction accuracy of the model. Moreover,CNN-SE-BiLSTM shows higher prediction accuracy among five weather categories compared to other traditional models,indicating that the proposed method is suitable for ultra-short-term photovoltaic power prediction.

, 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=Lan TANG, Liwen HUANG, Chenglei WANG), CN=ArticleExt(id=1190338346741240664, articleId=1190337957757292996, tenantId=1146029695717560320, journalId=1189987059142926344, language=CN, title=基于CGAN和CNN-SE-BiLSTM的极端天气光伏功率超短期预测, columnId=1190325328129131131, journalTitle=电气传动, columnName=综合能源与现代电网, runingTitle=null, highlight=null, articleAbstract=

针对因极端天气出现概率较低导致的光伏发电数据不平衡的问题,提出一种K-means聚类算法和基于Wasserstein距离含梯度惩罚项的条件生成对抗网络实现极端天气数据的分类扩充,并提出了一种结合双向长短期记忆网络与卷积神经网络并融入通道注意力机制的预测方法,旨在通过整合时空特征和动态调节特征通道重要性来提升光伏功率预测性能。首先,使用相关性分析和K-means算法对多种环境因素进行筛选,并对其进行划分以及添加标签。其次,选择聚类后数量较少的极端天气标签,使用CWGAN-GP对其进行样本扩充。最后,将扩充后的数据集作为训练集训练CNN-SE-BiLSTM预测模型,实现极端天气的光伏功率预测。以某光伏电站数据进行仿真建模,结果表明:使用CGAN-GP对原始极端天气训练集进行扩充有助于提高模型的预测精度。同时,CNN-SE-BiLSTM在五类天气中的预测误差较其他传统模型有更高的预测进度,说明所提方法适用于光伏功率超短期预测。

, correspAuthors=null, authorNote=null, correspAuthorsNote=
王成磊(1997—),男,硕士,主要研究方向为人工智能在电力系统中的运用,Email:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=OlhNCwoW1DUTtwmy04Hhfg==, magXml=HPMeuSAOFbkyEq3uMqr/Aw==, pdfUrl=null, pdf=sXd91ryNOoRtGd5zFkAO+A==, pdfFileSize=6644425, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=g5/cbwe7giQoIPS/JhUsbQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=QDxNz/3PprHxLX/jtuln6A==, mapNumber=null, authorCompany=null, fund=null, authors=

唐岚(1977—),男,博士,副教授,主要研究方向为电力系统优化与运行,Email:

, authorsList=唐岚, 黄力文, 王成磊)}, authors=[Author(id=1190338347261334375, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=tanglan@kust.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1190338347341026154, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, authorId=1190338347261334375, language=EN, stringName=Lan TANG, firstName=Lan, middleName=null, lastName=TANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650051,Yunnan,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1190338347412329324, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, authorId=1190338347261334375, language=CN, stringName=唐岚, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.昆明理工大学 电力工程学院,云南 昆明 650051, bio={"content":"

唐岚(1977—),男,博士,副教授,主要研究方向为电力系统优化与运行,Email:

"}, bioImg=null, bioContent=

唐岚(1977—),男,博士,副教授,主要研究方向为电力系统优化与运行,Email:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1190338346976121693, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, xref=null, ext=[AuthorCompanyExt(id=1190338346980315998, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338346976121693, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650051,Yunnan,China), AuthorCompanyExt(id=1190338346988704607, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338346976121693, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.昆明理工大学 电力工程学院,云南 昆明 650051)])]), Author(id=1190338347479438191, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1190338347622044529, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, authorId=1190338347479438191, language=EN, stringName=Liwen HUANG, firstName=Liwen, middleName=null, lastName=HUANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650051,Yunnan,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1190338347743679348, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, authorId=1190338347479438191, language=CN, stringName=黄力文, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.昆明理工大学 电力工程学院,云南 昆明 650051, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1190338346976121693, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, xref=null, ext=[AuthorCompanyExt(id=1190338346980315998, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338346976121693, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650051,Yunnan,China), AuthorCompanyExt(id=1190338346988704607, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338346976121693, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.昆明理工大学 电力工程学院,云南 昆明 650051)])]), Author(id=1190338347848536950, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=2832961093@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1190338347978560377, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, authorId=1190338347848536950, language=EN, stringName=Chenglei WANG, firstName=Chenglei, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2. Yunnan Electric Power Research Institute,Yunnan Power Grid Co.,Ltd.,Kunming 650217,Yunnan,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1190338348116972411, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, authorId=1190338347848536950, language=CN, stringName=王成磊, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.云南电网有限责任公司 电力科学研究院,云南 昆明 650217, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1190338347173253985, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, xref=null, ext=[AuthorCompanyExt(id=1190338347181642594, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338347173253985, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. Yunnan Electric Power Research Institute,Yunnan Power Grid Co.,Ltd.,Kunming 650217,Yunnan,China), AuthorCompanyExt(id=1190338347190031204, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338347173253985, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.云南电网有限责任公司 电力科学研究院,云南 昆明 650217)])])], keywords=[Keyword(id=1190338348238607230, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, orderNo=1, keyword=photovoltaic power prediction), Keyword(id=1190338348330881920, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, orderNo=2, keyword=extreme weather generation), Keyword(id=1190338348397990786, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, orderNo=3, keyword=bidirectional long short-term memory(BiLSTM)), Keyword(id=1190338348481876868, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, orderNo=4, keyword=conditional Wasserstein generative adversarial network with gradient penalty(CWGAN-GP)), Keyword(id=1190338349534647174, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, orderNo=5, keyword=K-means clustering algorithm), Keyword(id=1190338349610144647, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, orderNo=1, keyword=光伏功率预测), Keyword(id=1190338349668864905, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, orderNo=2, keyword=极端天气生成), Keyword(id=1190338349735973771, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, orderNo=3, keyword=双向长短期记忆神经网络), Keyword(id=1190338349807276941, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, orderNo=4, keyword=Wasserstein距离含梯度惩罚项的条件生成对抗网络), Keyword(id=1190338349891163023, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, orderNo=5, keyword=K-means聚类算法)], refs=[Reference(id=1190338352906867672, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=1, pageStart=1, pageEnd=10, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=江秀臣, 许永鹏, 李曜丞, journalName=高电压技术, refType=null, unstructuredReference=江秀臣, 许永鹏, 李曜丞, 等. 新型电力系统背景下的输变电数字化转型[J]. 高电压技术, 2022, 48(1):1-10., articleTitle=新型电力系统背景下的输变电数字化转型, refAbstract=null), Reference(id=1190338352994948058, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=1, pageStart=1, pageEnd=10, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=JIANG Xiuchen, XU Yongpeng, LI Yaocheng, journalName=High Voltage Engineering, refType=null, unstructuredReference=JIANG Xiuchen, XU Yongpeng, LI Yaocheng, et al. Digitalization transformation of power transmission and transformation under the background of new power system[J]. High Voltage Engineering, 2022, 48(1):1-10., articleTitle=Digitalization transformation of power transmission and transformation under the background of new power system, refAbstract=null), Reference(id=1190338354009969628, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=43, issue=3, pageStart=392, pageEnd=404, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=肖瑶, 钮文泽, 魏高升, journalName=发电技术, refType=null, unstructuredReference=肖瑶, 钮文泽, 魏高升, 等. 太阳能光伏/光热技术研究现状与发展趋势综述[J]. 发电技术, 2022, 43(3):392-404., articleTitle=太阳能光伏/光热技术研究现状与发展趋势综述, refAbstract=null), Reference(id=1190338354081272798, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=43, issue=3, pageStart=392, pageEnd=404, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=XIAO Yao, NIU Wenze, WEI Gaosheng, journalName=Power Generation Technology, refType=null, unstructuredReference=XIAO Yao, NIU Wenze, WEI Gaosheng, et al. Review on research status and developing tendency of solar photovoltaic/thermal technology[J]. Power Generation Technology, 2022, 43(3):392-404., articleTitle=Review on research status and developing tendency of solar photovoltaic/thermal technology, refAbstract=null), Reference(id=1190338354173547488, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=5, pageStart=1943, pageEnd=1951, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=卢俊杰, 蔡涛, 郎建勋, journalName=高电压技术, refType=null, unstructuredReference=卢俊杰, 蔡涛, 郎建勋, 等. 基于集群划分的光伏电站集群发电功率短期预测方法[J]. 高电压技术, 2022, 48(5):1943-1951., articleTitle=基于集群划分的光伏电站集群发电功率短期预测方法, refAbstract=null), Reference(id=1190338354236462050, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=5, pageStart=1943, pageEnd=1951, url=null, language=null, rfNumber=[3], rfOrder=5, authorNames=LU Junjie, CAI Tao, LANG Jianxun, journalName=High Voltage Engineering, refType=null, unstructuredReference=LU Junjie, CAI Tao, LANG Jianxun, et al. Short-term power output forecasting of clustered photovoltaic solar plants based on cluster partition[J]. High Voltage Engineering, 2022, 48(5):1943-1951., articleTitle=Short-term power output forecasting of clustered photovoltaic solar plants based on cluster partition, refAbstract=null), Reference(id=1190338354320348131, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=24, issue=6, pageStart=52, pageEnd=63, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=石文辉, 屈姬贤, 罗魁, journalName=中国工程科学, refType=null, unstructuredReference=石文辉, 屈姬贤, 罗魁, 等. 高比例新能源并网与运行发展研究[J]. 中国工程科学, 2022, 24(6):52-63., articleTitle=高比例新能源并网与运行发展研究, refAbstract=null), Reference(id=1190338354383262691, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=24, issue=6, pageStart=52, pageEnd=63, url=null, language=null, rfNumber=[4], rfOrder=7, authorNames=SHI Wenhui, QU Jixian, LUO Kui, journalName=Strategic Study of CAE, refType=null, unstructuredReference=SHI Wenhui, QU Jixian, LUO Kui, et al. Grid-integration and operation of high proportioned new energy[J]. Strategic Study of CAE, 2022, 24(6):52-63., articleTitle=Grid-integration and operation of high proportioned new energy, refAbstract=null), Reference(id=1190338354441982948, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2017, volume=41, issue=21, pageStart=33, pageEnd=39, url=null, language=null, rfNumber=[5], rfOrder=8, authorNames=谭津, 邓长虹, 杨威, journalName=电力系统自动化, refType=null, unstructuredReference=谭津, 邓长虹, 杨威, 等. 微电网光伏发电的Adaboost天气聚类超短期预测方法[J]. 电力系统自动化, 2017, 41(21):33-39., articleTitle=微电网光伏发电的Adaboost天气聚类超短期预测方法, refAbstract=null), Reference(id=1190338354504897509, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2017, volume=41, issue=21, pageStart=33, pageEnd=39, url=null, language=null, rfNumber=[5], rfOrder=9, authorNames=TAN Jin, DENG Changhong, YANG Wei, journalName=Automation of Electric Power Systems, refType=null, unstructuredReference=TAN Jin, DENG Changhong, YANG Wei, et al. Ultra-short-term photovoltaic power forecasting in microgrid based on Adaboost clustering[J]. Automation of Electric Power Systems, 2017, 41(21):33-39., articleTitle=Ultra-short-term photovoltaic power forecasting in microgrid based on Adaboost clustering, refAbstract=null), Reference(id=1190338354567812070, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2017, volume=38, issue=7, pageStart=1, pageEnd=9, url=null, language=null, rfNumber=[6], rfOrder=10, authorNames=王洪坤, 葛磊蛟, 李宏伟, journalName=电力建设, refType=null, unstructuredReference=王洪坤, 葛磊蛟, 李宏伟, 等. 分布式光伏发电的特性分析与预测方法综述[J]. 电力建设, 2017, 38(7):1-9., articleTitle=分布式光伏发电的特性分析与预测方法综述, refAbstract=null), Reference(id=1190338354634920935, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2017, volume=38, issue=7, pageStart=1, pageEnd=9, url=null, language=null, rfNumber=[6], rfOrder=11, authorNames=WANG Hongkun, GE Leijiao, LI Hongwei, journalName=Electric Power Construction, refType=null, unstructuredReference=WANG Hongkun, GE Leijiao, LI Hongwei, et al. A review on characteristic analysis and prediction method of distributed PV[J]. Electric Power Construction, 2017, 38(7):1-9., articleTitle=A review on characteristic analysis and prediction method of distributed PV, refAbstract=null), Reference(id=1190338354710418408, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2019, volume=56, issue=1, pageStart=76, pageEnd=81, url=null, language=null, rfNumber=[7], rfOrder=12, authorNames=王继拓, 王万成, 陈宏伟, journalName=电测与仪表, refType=null, unstructuredReference=王继拓, 王万成, 陈宏伟. 基于回归——马尔科夫链的光伏发电功率预测[J]. 电测与仪表, 2019, 56(1):76-81., articleTitle=基于回归——马尔科夫链的光伏发电功率预测, refAbstract=null), Reference(id=1190338354794304489, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2019, volume=56, issue=1, pageStart=76, pageEnd=81, url=null, language=null, rfNumber=[7], rfOrder=13, authorNames=WANG Jituo, WANG Wancheng, CHEN Hongwei, journalName=Electrical Measurement & Instrumentation, refType=null, unstructuredReference=WANG Jituo, WANG Wancheng, CHEN Hongwei. Photovoltaic power generation forecasting based on regression——Markov chain[J]. Electrical Measurement & Instrumentation, 2019, 56(1):76-81., articleTitle=Photovoltaic power generation forecasting based on regression——Markov chain, refAbstract=null), Reference(id=1190338354945299434, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2016, volume=46, issue=6, pageStart=93, pageEnd=96, url=null, language=null, rfNumber=[8], rfOrder=14, authorNames=李安寿, 陈琦, 王子才, journalName=电气传动, refType=null, unstructuredReference=李安寿, 陈琦, 王子才, 等. 光伏发电系统功率预测方法综述[J]. 电气传动, 2016, 46(6):93-96., articleTitle=光伏发电系统功率预测方法综述, refAbstract=null), Reference(id=1190338355020796907, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2016, volume=46, issue=6, pageStart=93, pageEnd=96, url=null, language=null, rfNumber=[8], rfOrder=15, authorNames=LI Anshou, CHEN Qi, WANG Zicai, journalName=Electric Drive, refType=null, unstructuredReference=LI Anshou, CHEN Qi, WANG Zicai, et al. Review of power forecast methods for photovoltaic generating system[J]. Electric Drive, 2016, 46(6):93-96., articleTitle=Review of power forecast methods for photovoltaic generating system, refAbstract=null), Reference(id=1190338355092100076, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2024, volume=359, issue=2, pageStart=122709, pageEnd=null, url=null, language=null, rfNumber=[9], rfOrder=16, authorNames=HU Z, GAO Y, JI S, journalName=Applied Energy, refType=null, unstructuredReference=HU Z, GAO Y, JI S, et al. Improved multistep ahead photovol-taic power prediction model based on LSTM and self-attention with weather forecast data[J]. Applied Energy, 2024, 359(2):122709., articleTitle=Improved multistep ahead photovol-taic power prediction model based on LSTM and self-attention with weather forecast data, refAbstract=null), Reference(id=1190338355163403245, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2020, volume=162, issue=2, pageStart=1665, pageEnd=1683, url=null, language=null, rfNumber=[10], rfOrder=17, authorNames=GAO B, HUANG X, SHI J, journalName=Renewable Energy, refType=null, unstructuredReference=GAO B, HUANG X, SHI J, et al. Hourly forecasting of solar irradiance based on CEEMDAN and multi-strategy CNN-LSTM neural networks[J]. Renewable Energy, 2020, 162(2):1665-1683., articleTitle=Hourly forecasting of solar irradiance based on CEEMDAN and multi-strategy CNN-LSTM neural networks, refAbstract=null), Reference(id=1190338355226317806, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2024, volume=52, issue=2, pageStart=165, pageEnd=176, url=null, language=null, rfNumber=[11], rfOrder=18, authorNames=黄泽, 毕贵红, 谢旭, journalName=电力系统保护与控制, refType=null, unstructuredReference=黄泽, 毕贵红, 谢旭, 等. 基于MBI-PBI-ResNet的超短期光伏功率预测[J]. 电力系统保护与控制, 2024, 52(2):165-176., articleTitle=基于MBI-PBI-ResNet的超短期光伏功率预测, refAbstract=null), Reference(id=1190338355293426671, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2024, volume=52, issue=2, pageStart=165, pageEnd=176, url=null, language=null, rfNumber=[11], rfOrder=19, authorNames=HUANG Ze, BI Guihong, XIE Xu, journalName=Power System Protection and Control, refType=null, unstructuredReference=HUANG Ze, BI Guihong, XIE Xu, et al. Ultra-short-term PV power prediction based on MBI-PBI-ResNet[J]. Power System Protection and Control, 2024, 52(2):165-176., articleTitle=Ultra-short-term PV power prediction based on MBI-PBI-ResNet, refAbstract=null), Reference(id=1190338355368924144, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2023, volume=43, issue=3, pageStart=987, pageEnd=1004, url=null, language=null, rfNumber=[12], rfOrder=20, authorNames=邵振国, 张承圣, 陈飞雄, journalName=中国电机工程学报, refType=null, unstructuredReference=邵振国, 张承圣, 陈飞雄, 等. 生成对抗网络及其在电力系统中的应用综述[J]. 中国电机工程学报, 2023, 43(3):987-1004., articleTitle=生成对抗网络及其在电力系统中的应用综述, refAbstract=null), Reference(id=1190338357298303985, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2023, volume=43, issue=3, pageStart=987, pageEnd=1004, url=null, language=null, rfNumber=[12], rfOrder=21, authorNames=SHAO Zhenguo, ZHANG Chengsheng, CHEN Feixiong, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=SHAO Zhenguo, ZHANG Chengsheng, CHEN Feixiong, et al. A review on generative adversarial networks for power system applications[J]. Proceedings of the CSEE, 2023, 43(3):987-1004., articleTitle=A review on generative adversarial networks for power system applications, refAbstract=null), Reference(id=1190338357373801458, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2016, volume=85, issue=1, pageStart=83, pageEnd=95, url=null, language=null, rfNumber=[13], rfOrder=22, authorNames=HU Q H, ZHANG R J, ZHOU Y C, journalName=Renewable Energy, refType=null, unstructuredReference=HU Q H, ZHANG R J, ZHOU Y C. Transfer learning for short-term wind speed prediction with deep neural networks[J]. Renewable Energy, 2016, 85(1):83-95., articleTitle=Transfer learning for short-term wind speed prediction with deep neural networks, refAbstract=null), Reference(id=1190338357440910323, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2019, volume=181, issue=2, pageStart=443, pageEnd=462, url=null, language=null, rfNumber=[14], rfOrder=23, authorNames=WANG Fei, ZHANG Zhanyao, LIU Chun, journalName=Energy Conversion and Management, refType=null, unstructuredReference=WANG Fei, ZHANG Zhanyao, LIU Chun, et al. Generative adversarial networks and convolutional neural networks based weather classification model for day ahead short-term photovoltaic power forecasting[J]. Energy Conversion and Management, 2019, 181(2):443-462., articleTitle=Generative adversarial networks and convolutional neural networks based weather classification model for day ahead short-term photovoltaic power forecasting, refAbstract=null), Reference(id=1190338357503824884, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=11, pageStart=4342, pageEnd=4351, url=null, language=null, rfNumber=[15], rfOrder=24, authorNames=殷豪, 张铮, 丁伟锋, journalName=高电压技术, refType=null, unstructuredReference=殷豪, 张铮, 丁伟锋, 等. 基于生成对抗网络和LSTM-CSO的少样本光伏功率短期预测[J]. 高电压技术, 2022, 48(11):4342-4351., articleTitle=基于生成对抗网络和LSTM-CSO的少样本光伏功率短期预测, refAbstract=null), Reference(id=1190338358556595189, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=11, pageStart=4342, pageEnd=4351, url=null, language=null, rfNumber=[15], rfOrder=25, authorNames=YIN Hao, ZHANG Zheng, DING Weifeng, journalName=High Voltage Engineering, refType=null, unstructuredReference=YIN Hao, ZHANG Zheng, DING Weifeng, et al. Short-term prediction of small-sample photovoltaic power based on generative adversarial network and LSTM-CSO[J]. High Voltage Engineering, 2022, 48(11):4342-4351., articleTitle=Short-term prediction of small-sample photovoltaic power based on generative adversarial network and LSTM-CSO, refAbstract=null), Reference(id=1190338358627898358, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2020, volume=40, issue=17, pageStart=5527, pageEnd=5536, url=null, language=null, rfNumber=[16], rfOrder=26, authorNames=董骁翀, 孙英云, 蒲天骄, journalName=中国电机工程学报, refType=null, unstructuredReference=董骁翀, 孙英云, 蒲天骄. 基于条件生成对抗网络的可再生能源日前场景生成方法[J]. 中国电机工程学报, 2020, 40(17):5527-5536., articleTitle=基于条件生成对抗网络的可再生能源日前场景生成方法, refAbstract=null), Reference(id=1190338358690812919, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2020, volume=40, issue=17, pageStart=5527, pageEnd=5536, url=null, language=null, rfNumber=[16], rfOrder=27, authorNames=DONG Xiaochong, SUN Yingyun, PU Tianjiao, journalName=Proceedings of the CSEE, refType=null, unstructuredReference=DONG Xiaochong, SUN Yingyun, PU Tianjiao. Day-ahead scenario generation of renewable energy based on conditional GAN[J]. Proceedings of the CSEE, 2020, 40(17):5527-5536., articleTitle=Day-ahead scenario generation of renewable energy based on conditional GAN, refAbstract=null), Reference(id=1190338358804059128, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2023, volume=47, issue=23, pageStart=189, pageEnd=199, url=null, language=null, rfNumber=[17], rfOrder=28, authorNames=孙浩, 万灿, 曹照静, journalName=电力系统自动化, refType=null, unstructuredReference=孙浩, 万灿, 曹照静, 等. 基于条件生成对抗网络曲线生成的短期负荷概率预测[J]. 电力系统自动化, 2023, 47(23):189-199., articleTitle=基于条件生成对抗网络曲线生成的短期负荷概率预测, refAbstract=null), Reference(id=1190338358871167993, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=null, volume=47, issue=23, pageStart=189, pageEnd=199, url=null, language=null, rfNumber=[17], rfOrder=29, authorNames=SUN Hao, WAN Can, CAO Zhaojing, journalName=Automation of Electric Power Systems 2023, refType=null, unstructuredReference=SUN Hao, WAN Can, CAO Zhaojing, et al. Short-term load probabilistic forecasting based on conditional generative adversarial network curve generation[J]. Automation of Electric Power Systems 2023, 47(23):189-199., articleTitle=Short-term load probabilistic forecasting based on conditional generative adversarial network curve generation, refAbstract=null), Reference(id=1190338358980219898, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=257, issue=15, pageStart=1, pageEnd=17, url=null, language=null, rfNumber=[18], rfOrder=30, authorNames=LI Jianbin, CHEN Zhiqiang, CHENG Long, journalName=Energy, refType=null, unstructuredReference=LI Jianbin, CHEN Zhiqiang, CHENG Long, et al. Energy data generation with Wasserstein deep convolutional generative adversarial networks[J]. Energy, 2022, 257(15):1-17., articleTitle=Energy data generation with Wasserstein deep convolutional generative adversarial networks, refAbstract=null), Reference(id=1190338359055717371, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=31, authorNames=贾鹏云, journalName=基于深度学习算法的短期光伏发电功率预测研究, refType=null, unstructuredReference=贾鹏云. 基于深度学习算法的短期光伏发电功率预测研究[D]. 北京: 华北电力大学(北京), 2021., articleTitle=null, refAbstract=null), Reference(id=1190338359127020540, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=32, authorNames=JIA Pengyun, journalName=Research on short term photovoltaic power prediction based on deep learning, refType=null, unstructuredReference=JIA Pengyun. Research on short term photovoltaic power prediction based on deep learning[D]. Beijing: North China Electric Power University (Beijing), 2021., articleTitle=null, refAbstract=null), Reference(id=1190338359185740797, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[20], rfOrder=33, authorNames=陈嘉铭, journalName=基于深度学习和强化学习的光伏发电功率预测研究, refType=null, unstructuredReference=陈嘉铭. 基于深度学习和强化学习的光伏发电功率预测研究[D]. 广州: 广东工业大学, 2022., articleTitle=null, refAbstract=null), Reference(id=1190338359290598398, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[20], rfOrder=34, authorNames=CHEN Jiaming, journalName=Research on photovoltaic power forecasting based on deep learning and reinforcement learning, refType=null, unstructuredReference=CHEN Jiaming. Research on photovoltaic power forecasting based on deep learning and reinforcement learning[D]. Guangzhou: Guangdong University of Technology, 2022., articleTitle=null, refAbstract=null), Reference(id=1190338359353512959, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=47, issue=4, pageStart=1176, pageEnd=1184, url=null, language=null, rfNumber=[21], rfOrder=35, authorNames=孟安波, 陈嘉铭, 黎湛联, journalName=高电压技术, refType=null, unstructuredReference=孟安波, 陈嘉铭, 黎湛联, 等. 基于相似日理论和CSO-WGPR的短期光伏发电功率预测[J]. 高电压技术, 2021, 47(4):1176-1184., articleTitle=基于相似日理论和CSO-WGPR的短期光伏发电功率预测, refAbstract=null), Reference(id=1190338359420621824, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=47, issue=4, pageStart=1176, pageEnd=1184, url=null, language=null, rfNumber=[21], rfOrder=36, authorNames=MENG Anbo, CHEN Jiaming, LI Zhanlian, journalName=High Voltage Engineering, refType=null, unstructuredReference=MENG Anbo, CHEN Jiaming, LI Zhanlian, et al. Short-term photovoltaic power generation prediction based on similar day theory and CSO-WGPR[J]. High Voltage Engineering, 2021, 47(4):1176-1184., articleTitle=Short-term photovoltaic power generation prediction based on similar day theory and CSO-WGPR, refAbstract=null), Reference(id=1190338359496118272, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=7, issue=1, pageStart=66, pageEnd=77, url=null, language=null, rfNumber=[22], rfOrder=37, authorNames=QIAO J, PU T J, WANG X Y, journalName=CSEE Journal of Power and Energy Systems, refType=null, unstructuredReference=QIAO J, PU T J, WANG X Y. Renewable scenario generation using controllable generative adversarial networks with transparent latent space[J]. CSEE Journal of Power and Energy Systems, 2021, 7(1):66-77., articleTitle=Renewable scenario generation using controllable generative adversarial networks with transparent latent space, refAbstract=null), Reference(id=1190338359575810049, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2020, volume=44, issue=4, pageStart=1505, pageEnd=1513, url=null, language=null, rfNumber=[23], rfOrder=38, authorNames=刘云鹏, 许自强, 和家慧, journalName=电网技术, refType=null, unstructuredReference=刘云鹏, 许自强, 和家慧, 等. 基于条件式Wasserstein生成对抗网络的电力变压器故障样本增强技术[J]. 电网技术, 2020, 44(4):1505-1513., articleTitle=基于条件式Wasserstein生成对抗网络的电力变压器故障样本增强技术, refAbstract=null), Reference(id=1190338359642918914, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2020, volume=44, issue=4, pageStart=1505, pageEnd=1513, url=null, language=null, rfNumber=[23], rfOrder=39, authorNames=LIU Yunpeng, XU Ziqiang, HE Jiahui, journalName=Power System Technology, refType=null, unstructuredReference=LIU Yunpeng, XU Ziqiang, HE Jiahui, et al. Data augmentation method for power transformer fault diagnosis based on conditional Wasserstein generative adversarial network[J]. Power System Technology, 2020, 44(4):1505-1513., articleTitle=Data augmentation method for power transformer fault diagnosis based on conditional Wasserstein generative adversarial network, refAbstract=null), Reference(id=1190338359710027779, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2024, volume=50, issue=3, pageStart=1131, pageEnd=1144, url=null, language=null, rfNumber=[24], rfOrder=40, authorNames=王成磊, 匡熠, 濮永现, journalName=高电压技术, refType=null, unstructuredReference=王成磊, 匡熠, 濮永现, 等. 基于改进生成对抗网络场景生成的长-短期储能优化配置[J]. 高电压技术, 2024, 50(3):1131-1144., articleTitle=基于改进生成对抗网络场景生成的长-短期储能优化配置, refAbstract=null), Reference(id=1190338359772942340, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2024, volume=50, issue=3, pageStart=1131, pageEnd=1144, url=null, language=null, rfNumber=[24], rfOrder=41, authorNames=WANG Chenglei, KUANG Yi, PU Yongxian, journalName=High Voltage Engineering, refType=null, unstructuredReference=WANG Chenglei, KUANG Yi, PU Yongxian, et al. Optimal configuration of long-term and short-term energy storage based on improved generative adversarial network scenarios[J]. High Voltage Engineering, 2024, 50(3):1131-1144., articleTitle=Optimal configuration of long-term and short-term energy storage based on improved generative adversarial network scenarios, refAbstract=null), Reference(id=1190338359831662597, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=41, issue=S1, pageStart=1, pageEnd=6, url=null, language=null, rfNumber=[25], rfOrder=42, authorNames=任欢, 王旭光, journalName=计算机应用, refType=null, unstructuredReference=任欢, 王旭光. 注意力机制综述[J]. 计算机应用, 2021, 41(S1):1-6., articleTitle=注意力机制综述, refAbstract=null), Reference(id=1190338359894577158, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2021, volume=41, issue=S1, pageStart=1, pageEnd=6, url=null, language=null, rfNumber=[25], rfOrder=43, authorNames=REN Huan, WANG Xuguang, journalName=Journal of Computer Applications, refType=null, unstructuredReference=REN Huan, WANG Xuguang. Review of attention mechanism[J]. Journal of Computer Applications, 2021, 41(S1):1-6., articleTitle=Review of attention mechanism, refAbstract=null), Reference(id=1190338359970074631, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2020, volume=42, issue=8, pageStart=2011, pageEnd=2023, url=null, language=null, rfNumber=[26], rfOrder=44, authorNames=HU J, SHENG L, SUN G, journalName=IEEE Transactions on Pattern Analysis and Machine Intelligence, refType=null, unstructuredReference=HU J, SHENG L, SUN G. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8):2011-2023., articleTitle=Squeeze-and-excitation networks, refAbstract=null), Reference(id=1190338360037183496, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=46, issue=9, pageStart=3463, pageEnd=3476, url=null, language=null, rfNumber=[27], rfOrder=45, authorNames=毕贵红, 赵鑫, 陈臣鹏, journalName=电网技术, refType=null, unstructuredReference=毕贵红, 赵鑫, 陈臣鹏, 等. 基于多通道输入和PCNN-BiLSTM的光伏发电功率超短期预测[J]. 电网技术, 2022, 46(9):3463-3476., articleTitle=基于多通道输入和PCNN-BiLSTM的光伏发电功率超短期预测, refAbstract=null), Reference(id=1190338360100098057, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=46, issue=9, pageStart=3463, pageEnd=3476, url=null, language=null, rfNumber=[27], rfOrder=46, authorNames=BI Guihong, ZHAO Xin, CHEN Chenpeng, journalName=Power System Technology, refType=null, unstructuredReference=BI Guihong, ZHAO Xin, CHEN Chenpeng. Ultra-short-term prediction of photovoltaic power generation based on multi-channel input and PCNN-BiLSTM[J]. Power System Technology, 2022, 46(9):3463-3476., articleTitle=Ultra-short-term prediction of photovoltaic power generation based on multi-channel input and PCNN-BiLSTM, refAbstract=null), Reference(id=1190338360167206922, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=11, pageStart=4372, pageEnd=4388, url=null, language=null, rfNumber=[28], rfOrder=47, authorNames=王开艳, 杜浩东, 贾嵘, journalName=高电压技术, refType=null, unstructuredReference=王开艳, 杜浩东, 贾嵘, 等. 基于相似日聚类和QR-CNN-BiLSTM模型的光伏功率短期区间概率预测[J]. 高电压技术, 2022, 48(11):4372-4388., articleTitle=基于相似日聚类和QR-CNN-BiLSTM模型的光伏功率短期区间概率预测, refAbstract=null), Reference(id=1190338360238510091, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2022, volume=48, issue=11, pageStart=4372, pageEnd=4388, url=null, language=null, rfNumber=[28], rfOrder=48, authorNames=WANG Kaiyan, DU Haodong, JIA Rong, journalName=High Volta-ge Engineering, refType=null, unstructuredReference=WANG Kaiyan, DU Haodong, JIA Rong, et al. Short-term interval probability prediction of photovoltaic power based on similar daily clustering and QR-CNN-BiLSTM model[J]. High Volta-ge Engineering, 2022, 48(11):4372-4388., articleTitle=Short-term interval probability prediction of photovoltaic power based on similar daily clustering and QR-CNN-BiLSTM model, refAbstract=null), Reference(id=1190338360339173388, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[29], rfOrder=49, authorNames=乔依林, journalName=风电运行数据预处理技术及其应用研究, refType=null, unstructuredReference=乔依林. 风电运行数据预处理技术及其应用研究[D]. 北京: 华北电力大学(北京), 2019., articleTitle=null, refAbstract=null), Reference(id=1190338360418865165, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[29], rfOrder=50, authorNames=QIAO Yilin, journalName=Research on preprocessing of wind power operation data and its application, refType=null, unstructuredReference=QIAO Yilin. Research on preprocessing of wind power operation data and its application[J]. Beijing: North China Electric Power University (Beijing), 2019., articleTitle=null, refAbstract=null)], funds=[Fund(id=1190338352797815766, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, awardId=202002AF080001, language=CN, fundingSource=云南省重大科技专项计划(202002AF080001), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1190338346976121693, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, xref=null, ext=[AuthorCompanyExt(id=1190338346980315998, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338346976121693, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650051,Yunnan,China), AuthorCompanyExt(id=1190338346988704607, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338346976121693, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.昆明理工大学 电力工程学院,云南 昆明 650051)]), AuthorCompany(id=1190338347173253985, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, xref=null, ext=[AuthorCompanyExt(id=1190338347181642594, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338347173253985, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. Yunnan Electric Power Research Institute,Yunnan Power Grid Co.,Ltd.,Kunming 650217,Yunnan,China), AuthorCompanyExt(id=1190338347190031204, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, companyId=1190338347173253985, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.云南电网有限责任公司 电力科学研究院,云南 昆明 650217)])], figs=[ArticleFig(id=1190338350033769362, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.1, caption=CWGAN-GP training process diagram, figureFileSmall=csW0UUq0dQDJteUETdGw/g==, figureFileBig=g5/cbwe7giQoIPS/JhUsbQ==, tableContent=null), ArticleFig(id=1190338350100878228, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图1, caption=CWGAN-GP训练过程图, figureFileSmall=csW0UUq0dQDJteUETdGw/g==, figureFileBig=g5/cbwe7giQoIPS/JhUsbQ==, tableContent=null), ArticleFig(id=1190338350314787735, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.2, caption=SENet architecture diagram, figureFileSmall=k0qtW2kyIQGeiQInWctqyA==, figureFileBig=vZiH2zQm2348yNpAIGoEKQ==, tableContent=null), ArticleFig(id=1190338350373507993, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图2, caption=SENet结构图, figureFileSmall=k0qtW2kyIQGeiQInWctqyA==, figureFileBig=vZiH2zQm2348yNpAIGoEKQ==, tableContent=null), ArticleFig(id=1190338350465782683, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.3, caption=Convolutional unit structure diagram, figureFileSmall=dNsiKgWL3UNQU4eCIgVtHQ==, figureFileBig=Mids6slQZMDX5iou0j5jEg==, tableContent=null), ArticleFig(id=1190338350520308637, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图3, caption=卷积单元结构图, figureFileSmall=dNsiKgWL3UNQU4eCIgVtHQ==, figureFileBig=Mids6slQZMDX5iou0j5jEg==, tableContent=null), ArticleFig(id=1190338350595806111, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.4, caption=BiLSTM structure diagram, figureFileSmall=y7DrfF2rRWZYXxOem71naw==, figureFileBig=QePI9D0wKtdmZCic4KI8cw==, tableContent=null), ArticleFig(id=1190338350675497889, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图4, caption=BiLSTM结构图, figureFileSmall=y7DrfF2rRWZYXxOem71naw==, figureFileBig=QePI9D0wKtdmZCic4KI8cw==, tableContent=null), ArticleFig(id=1190338350767772579, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.5, caption=Structure diagram of CNN-SE-BiLSTM prediction model, figureFileSmall=rph4j4DlRX0x6thKWS8jsw==, figureFileBig=yuWj+9M5kszB58y0rVtXWg==, tableContent=null), ArticleFig(id=1190338350851658661, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图5, caption=CNN-SE-BiLSTM预测模型结构图, figureFileSmall=rph4j4DlRX0x6thKWS8jsw==, figureFileBig=yuWj+9M5kszB58y0rVtXWg==, tableContent=null), ArticleFig(id=1190338350922961831, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.6, caption=Workflow of the proposed approach, figureFileSmall=3FeyvyrVx8KEybzfSNVY5g==, figureFileBig=n98+xIQ10AVsM+907+tUUQ==, tableContent=null), ArticleFig(id=1190338350985876393, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图6, caption=本文工作流程, figureFileSmall=3FeyvyrVx8KEybzfSNVY5g==, figureFileBig=n98+xIQ10AVsM+907+tUUQ==, tableContent=null), ArticleFig(id=1190338351052985259, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.7, caption=Correlation analysis graph, figureFileSmall=3K8PABEjWWD1/ryYpaLjtQ==, figureFileBig=DDVQwhKSYq2vxbvMg60p2Q==, tableContent=null), ArticleFig(id=1190338351128482733, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图7, caption=相关性分析图, figureFileSmall=3K8PABEjWWD1/ryYpaLjtQ==, figureFileBig=DDVQwhKSYq2vxbvMg60p2Q==, tableContent=null), ArticleFig(id=1190338351203980207, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.8, caption=K-means clustering results plot, figureFileSmall=rlsxUf4Omz9Zqe5XtWNvag==, figureFileBig=M3nukE4plGSxDRz3u2YzpA==, tableContent=null), ArticleFig(id=1190338351262700465, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图8, caption=K-means聚类结果图, figureFileSmall=rlsxUf4Omz9Zqe5XtWNvag==, figureFileBig=M3nukE4plGSxDRz3u2YzpA==, tableContent=null), ArticleFig(id=1190338351334003635, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.9, caption=Extreme weather data generated by CWGAN-GP, figureFileSmall=YOV36fHx5KG8WueZEyEiAQ==, figureFileBig=FON9ekQUDicdYnE/Q/3JTg==, tableContent=null), ArticleFig(id=1190338351405306805, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图9, caption=CWGAN-GP生成极端天气数据, figureFileSmall=YOV36fHx5KG8WueZEyEiAQ==, figureFileBig=FON9ekQUDicdYnE/Q/3JTg==, tableContent=null), ArticleFig(id=1190338351501775799, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.10, caption=Extreme weather power prediction curves for different forecasting horizons, figureFileSmall=Gncrk5X49RRqgQVscHSerQ==, figureFileBig=al9LyVWIskmh2+oPHLDwIA==, tableContent=null), ArticleFig(id=1190338351585661881, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图10, caption=扩充不同天数下的极端天气功率预测曲线, figureFileSmall=Gncrk5X49RRqgQVscHSerQ==, figureFileBig=al9LyVWIskmh2+oPHLDwIA==, tableContent=null), ArticleFig(id=1190338351644382139, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Fig.11, caption=Power prediction curves of different forecasting models, figureFileSmall=e4kWTwGKl6LhDEc3NJD/Ug==, figureFileBig=5GiYh/QAtH+folYcNwTA8A==, tableContent=null), ArticleFig(id=1190338351719879613, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=图11, caption=不同预测模型的预测功率曲线, figureFileSmall=e4kWTwGKl6LhDEc3NJD/Ug==, figureFileBig=5GiYh/QAtH+folYcNwTA8A==, tableContent=null), ArticleFig(id=1190338351786988479, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Tab.1, caption=

Generator structure and parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
层数 名称 参数
Layer1 2D CONV 卷积核大小 3×3
滤波器数量 32
步长 1
正则化 数量 32
激活函数 LeakyRelu 0.2
Layer2 2D CONV 卷积核大小 3×3
滤波器数量 64
步长 1
正则化 数量 64
激活函数 LeakyRelu 0.2
Layer3 2D CONV 卷积核大小 3×3
滤波器数量 1
步长 1
激活函数 ReLU
), ArticleFig(id=1190338351875068865, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=表1, caption=

生成器的结构和参数

, figureFileSmall=null, figureFileBig=null, tableContent=
层数 名称 参数
Layer1 2D CONV 卷积核大小 3×3
滤波器数量 32
步长 1
正则化 数量 32
激活函数 LeakyRelu 0.2
Layer2 2D CONV 卷积核大小 3×3
滤波器数量 64
步长 1
正则化 数量 64
激活函数 LeakyRelu 0.2
Layer3 2D CONV 卷积核大小 3×3
滤波器数量 1
步长 1
激活函数 ReLU
), ArticleFig(id=1190338351950566339, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Tab.2, caption=

Discriminator structure and parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
层数 名称 参数
Layer1 2D CONV 卷积核大小 3×3
滤波器数量 32
步长 1
正则化 数量 32
激活函数 LeakyRelu 0.2
Layer2 2D CONV 卷积核大小 3×3
滤波器数量 64
步长 1
正则化 数量 64
激活函数 LeakyRelu 0.2
Layer3 2D CONV 卷积核大小 3×3
滤波器数量 1
步长 1
正则化 数量 64
激活函数 LeakyRelu 0.2
Layer4 全连接 神经元个数 1
), ArticleFig(id=1190338352055423941, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=表2, caption=

判别器的结构和参数

, figureFileSmall=null, figureFileBig=null, tableContent=
层数 名称 参数
Layer1 2D CONV 卷积核大小 3×3
滤波器数量 32
步长 1
正则化 数量 32
激活函数 LeakyRelu 0.2
Layer2 2D CONV 卷积核大小 3×3
滤波器数量 64
步长 1
正则化 数量 64
激活函数 LeakyRelu 0.2
Layer3 2D CONV 卷积核大小 3×3
滤波器数量 1
步长 1
正则化 数量 64
激活函数 LeakyRelu 0.2
Layer4 全连接 神经元个数 1
), ArticleFig(id=1190338352143504327, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Tab.3, caption=

CWGAN-GP training hyperparameters configuration

, figureFileSmall=null, figureFileBig=null, tableContent=
超参数名称 参数值
生成器、判别器学习率 0.000 01
生成器、判别器优化器 Adam
生成器、判别器训练迭代周期 1
总迭代次数 20 000
), ArticleFig(id=1190338352214807497, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=表3, caption=

CWGAN-GP训练超参数设置

, figureFileSmall=null, figureFileBig=null, tableContent=
超参数名称 参数值
生成器、判别器学习率 0.000 01
生成器、判别器优化器 Adam
生成器、判别器训练迭代周期 1
总迭代次数 20 000
), ArticleFig(id=1190338352298693579, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Tab.4, caption=

SENet parameter settings

, figureFileSmall=null, figureFileBig=null, tableContent=
层数 名称 参数
Layer1 全局平均池化层
Layer2 全连接层 神经元个数 16
激活函数 LeakyRelu
Layer3 全连接层 神经元个数 64
激活函数 Sigmoid
), ArticleFig(id=1190338352378385357, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=表4, caption=

SENet参数设置

, figureFileSmall=null, figureFileBig=null, tableContent=
层数 名称 参数
Layer1 全局平均池化层
Layer2 全连接层 神经元个数 16
激活函数 LeakyRelu
Layer3 全连接层 神经元个数 64
激活函数 Sigmoid
), ArticleFig(id=1190338352462271439, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Tab.5, caption=

Comparison of prediction errors for different data augmentation durations

, figureFileSmall=null, figureFileBig=null, tableContent=
扩充天数 MAE RMSE
未扩充 1.110 2.235
扩充5 d 0.580 1.174
扩充10 d 0.565 1.146
扩充15 d 0.560 1.110
扩充20 d 0.556 1.106
), ArticleFig(id=1190338352533574608, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=表5, caption=

不同数据扩充天数下的预测误差对比表

, figureFileSmall=null, figureFileBig=null, tableContent=
扩充天数 MAE RMSE
未扩充 1.110 2.235
扩充5 d 0.580 1.174
扩充10 d 0.565 1.146
扩充15 d 0.560 1.110
扩充20 d 0.556 1.106
), ArticleFig(id=1190338352642626513, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=EN, label=Tab.6, caption=

Comparison of prediction accuracy among different models

, figureFileSmall=null, figureFileBig=null, tableContent=
天气类型 模型 MAE RMSE
晴天 CNN-SE-BiLSTM 1.104 8 1.671 8
LSTM 1.675 4 3.599 1
BiLSTM 1.241 8 3.128 6
BP 1.754 9 3.824 7
多云 CNN-SE-BiLSTM 1.275 4 2.502 4
LSTM 2.377 2 3.254 4
BiLSTM 1.427 7 2.555 0
BP 3.027 3 3.940 5
少云 CNN-SE-BiLSTM 1.392 5 2.502 2
LSTM 2.283 1 3.975 2
BiLSTM 1.968 9 3.320 7
BP 2.153 7 3.767 8
阴雨 CNN-SE-BiLSTM 1.697 2 3.544 3
LSTM 2.474 7 7.359 0
BiLSTM 2.124 3 3.715 5
BP 2.848 0 3.977 9
极端天气 CNN-SE-BiLSTM 0.765 6 1.845 9
LSTM 1.446 0 2.460 1
BiLSTM 1.152 0 1.447 3
BP 1.428 6 3.208 3
), ArticleFig(id=1190338352697152467, tenantId=1146029695717560320, journalId=1189987059142926344, articleId=1190337957757292996, language=CN, label=表6, caption=

各模型预测精度对比

, figureFileSmall=null, figureFileBig=null, tableContent=
天气类型 模型 MAE RMSE
晴天 CNN-SE-BiLSTM 1.104 8 1.671 8
LSTM 1.675 4 3.599 1
BiLSTM 1.241 8 3.128 6
BP 1.754 9 3.824 7
多云 CNN-SE-BiLSTM 1.275 4 2.502 4
LSTM 2.377 2 3.254 4
BiLSTM 1.427 7 2.555 0
BP 3.027 3 3.940 5
少云 CNN-SE-BiLSTM 1.392 5 2.502 2
LSTM 2.283 1 3.975 2
BiLSTM 1.968 9 3.320 7
BP 2.153 7 3.767 8
阴雨 CNN-SE-BiLSTM 1.697 2 3.544 3
LSTM 2.474 7 7.359 0
BiLSTM 2.124 3 3.715 5
BP 2.848 0 3.977 9
极端天气 CNN-SE-BiLSTM 0.765 6 1.845 9
LSTM 1.446 0 2.460 1
BiLSTM 1.152 0 1.447 3
BP 1.428 6 3.208 3
)], attaches=null, journal=Journal(id=1149695973913305096, delFlag=0, nameCn=电气传动, nameEn=Electric Drive, nameHistory1=null, nameHistory2=null, issn=1001-2095, eissn=null, cn=12-1067/TP, 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=MX9kQyxauhRIcWSHz1lsfg==, journalPrice=null, startedYear=null, abbrevIsoEn=null, journalRemark=null, publicationField=null, createdTime=1752038480501, updatedTime=1761735785885, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=E, firstLetterEn=E, subjectCode=Engineering, subjectName=Engineering, subjectCodeEn=Engineering, subjectNameEn=null, picCn=MX9kQyxauhRIcWSHz1lsfg==, picEn=R8RQJTk995s1dk9HSLANAQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1190369420775096487, language=CN, name=电气传动, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1761735785907, updatedTime=1761735785907, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://tg.au365.cn/Login/index.html, submissionEditorUrl=https://tg.au365.cn/Login/index/cate/3.html, submissionReviewUrl=https://tg.au365.cn/Login/index/cate/2.html, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1190369420829622440, language=EN, name=Electric Drive, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1761735785920, updatedTime=1761735785920, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://tg.au365.cn/Login/index.html, submissionEditorUrl=https://tg.au365.cn/Login/index/cate/3.html, submissionReviewUrl=https://tg.au365.cn/Login/index/cate/2.html, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1189987059142926344, websiteList=[Website(id=1189988338468713189, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1189987059142926344, 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/dqcd/CN, language=CN, createTime=1761644928805, createBy=18614031015, updateTime=1761644944247, updateBy=18614031015, name=电气传动-中文, tplId=1146099689490845704, title=电气传动, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1190236467705254391, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=articleTextType, value=kx, createTime=1761704087425, updateTime=1761704087425, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467671699956, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=banner, value=null, createTime=1761704087417, updateTime=1761704087417, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467763974650, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=grayFlag, value=0, createTime=1761704087439, updateTime=1761704087439, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467663311347, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=logo, value=https://castjournals.cast.org.cn/joweb/dqcd/CN/file/pic?fileId=C+51VGe7It8gHMdS0MAlXg==, createTime=1761704087415, updateTime=1761704087415, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467784946172, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=minRunFlag, value=0, createTime=1761704087444, updateTime=1761704087444, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467692671478, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/dqcd/CN/file/pic, createTime=1761704087422, updateTime=1761704087422, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467776557563, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=silenceFlag, value=0, createTime=1761704087442, updateTime=1761704087442, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467680088565, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1761704087419, updateTime=1761704087419, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467713643000, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=themeColor, value=null, createTime=1761704087427, updateTime=1761704087427, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236467726225913, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338468713189, code=themeStyle, value=null, createTime=1761704087430, updateTime=1761704087430, creator=18614031015, updator=18614031015)]), Website(id=1189988338527433447, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1189987059142926344, 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/dqcd/EN, language=EN, createTime=1761644928819, createBy=18614031015, updateTime=1761644963810, updateBy=18614031015, name=电气传动-英文, tplId=1146101810881728533, title=Electric Drive, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1190236496637563393, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=articleTextType, value=kx, createTime=1761704094323, updateTime=1761704094323, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496612397566, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=banner, value=null, createTime=1761704094317, updateTime=1761704094317, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496666923524, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=grayFlag, value=0, createTime=1761704094330, updateTime=1761704094330, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496604008957, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=logo, value=https://castjournals.cast.org.cn/joweb/dqcd/EN/file/pic?fileId=C+51VGe7It8gHMdS0MAlXg==, createTime=1761704094315, updateTime=1761704094315, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496683700742, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=minRunFlag, value=0, createTime=1761704094334, updateTime=1761704094334, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496629174784, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/dqcd/EN/file/pic, createTime=1761704094321, updateTime=1761704094321, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496675312133, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=silenceFlag, value=0, createTime=1761704094332, updateTime=1761704094332, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496620786175, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1761704094319, updateTime=1761704094319, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496650146306, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=themeColor, value=null, createTime=1761704094326, updateTime=1761704094326, creator=18614031015, updator=18614031015), WebsiteProps(id=1190236496658534915, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1189988338527433447, code=themeStyle, value=null, createTime=1761704094328, updateTime=1761704094328, creator=18614031015, updator=18614031015)])], journalTitle=电气传动, weixinUrl=null, journalUrl=https://www.au365.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Electric Drive, journalPhotoCn=MX9kQyxauhRIcWSHz1lsfg==, journalPhotoEn=R8RQJTk995s1dk9HSLANAQ==, journalFirstLetter=E, 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/dqcd/CN/10.19457/j.1001-2095.dqcd25932, detailUrlEn=https://castjournals.cast.org.cn/joweb/dqcd/EN/10.19457/j.1001-2095.dqcd25932, pdfUrlCn=https://castjournals.cast.org.cn/joweb/dqcd/CN/PDF/10.19457/j.1001-2095.dqcd25932, pdfUrlEn=https://castjournals.cast.org.cn/joweb/dqcd/EN/PDF/10.19457/j.1001-2095.dqcd25932, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于CGAN和CNN-SE-BiLSTM的极端天气光伏功率超短期预测
收藏切换
PDF下载
唐岚 1 , 黄力文 1 , 王成磊 2
电气传动 | 综合能源与现代电网 2025,55(8): 58-69
收起
收藏切换
电气传动 | 综合能源与现代电网 2025, 55(8): 58-69
基于CGAN和CNN-SE-BiLSTM的极端天气光伏功率超短期预测
全屏
唐岚1 , 黄力文1, 王成磊2
作者信息
  • 1.昆明理工大学 电力工程学院,云南 昆明 650051
  • 2.云南电网有限责任公司 电力科学研究院,云南 昆明 650217
  • 唐岚(1977—),男,博士,副教授,主要研究方向为电力系统优化与运行,Email:

通讯作者:

王成磊(1997—),男,硕士,主要研究方向为人工智能在电力系统中的运用,Email:
Extreme Weather Photovoltaic Power Ultra-short-term Forecasting Based on CGAN and CNN-SE-BiLSTM
Lan TANG1 , Liwen HUANG1, Chenglei WANG2
Affiliations
  • 1. School of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650051,Yunnan,China
  • 2. Yunnan Electric Power Research Institute,Yunnan Power Grid Co.,Ltd.,Kunming 650217,Yunnan,China
出版时间: 2025-08-20 doi: 10.19457/j.1001-2095.dqcd25932
文章导航
收藏切换

针对因极端天气出现概率较低导致的光伏发电数据不平衡的问题,提出一种K-means聚类算法和基于Wasserstein距离含梯度惩罚项的条件生成对抗网络实现极端天气数据的分类扩充,并提出了一种结合双向长短期记忆网络与卷积神经网络并融入通道注意力机制的预测方法,旨在通过整合时空特征和动态调节特征通道重要性来提升光伏功率预测性能。首先,使用相关性分析和K-means算法对多种环境因素进行筛选,并对其进行划分以及添加标签。其次,选择聚类后数量较少的极端天气标签,使用CWGAN-GP对其进行样本扩充。最后,将扩充后的数据集作为训练集训练CNN-SE-BiLSTM预测模型,实现极端天气的光伏功率预测。以某光伏电站数据进行仿真建模,结果表明:使用CGAN-GP对原始极端天气训练集进行扩充有助于提高模型的预测精度。同时,CNN-SE-BiLSTM在五类天气中的预测误差较其他传统模型有更高的预测进度,说明所提方法适用于光伏功率超短期预测。

光伏功率预测  /  极端天气生成  /  双向长短期记忆神经网络  /  Wasserstein距离含梯度惩罚项的条件生成对抗网络  /  K-means聚类算法

A K-means clustering algorithm was proposed and a conditional Wasserstein generative adversarial network with gradient penalty(CWGAN-GP)to address the problem of imbalanced photovoltaic generation data caused by the low occurrence probability of extreme weather. A prediction approach combining bidirectional long short-term memory(BFLSTM)with convolutional neural network was introduced and incorporating channel attention mechanism to enhance the PV power prediction performance by integrating spatio-temporal features and dynamically adjusting the importance of feature channels. Firstly,correlation analysis and K-means algorithm were utilized to select and label various environmental factors. Then,extreme weather labels with fewer samples after clustering were selected,and CWGAN-GP was used for data augmentation.Finally,the augmented dataset was used to train the CNN-SE-BiLSTM prediction model for PV power prediction under extreme weather conditions.Simulation modeling was conducted using data from a certain PV power station,and the results demonstrate that augmenting the original extreme weather training set with CGAN-GP helps improve the prediction accuracy of the model. Moreover,CNN-SE-BiLSTM shows higher prediction accuracy among five weather categories compared to other traditional models,indicating that the proposed method is suitable for ultra-short-term photovoltaic power prediction.

photovoltaic power prediction  /  extreme weather generation  /  bidirectional long short-term memory(BiLSTM)  /  conditional Wasserstein generative adversarial network with gradient penalty(CWGAN-GP)  /  K-means clustering algorithm
唐岚, 黄力文, 王成磊. 基于CGAN和CNN-SE-BiLSTM的极端天气光伏功率超短期预测. 电气传动, 2025 , 55 (8) : 58 -69 . DOI: 10.19457/j.1001-2095.dqcd25932
Lan TANG, Liwen HUANG, Chenglei WANG. Extreme Weather Photovoltaic Power Ultra-short-term Forecasting Based on CGAN and CNN-SE-BiLSTM[J]. Electric Drive, 2025 , 55 (8) : 58 -69 . DOI: 10.19457/j.1001-2095.dqcd25932
“双碳”背景下,构建高比例可再生能源的新型电力系统是推进能源结构变革的重要举措[1]。随着光伏技术的不断创新和成本的持续下降,大规模的光伏发电系统并网,为能源转型提供了坚实的基础[2]。然而,光伏发电因受到所处地理环境和气候因素的影响,其发电功率表现出较强的间歇性以及非平稳性,同时不同天气的功率曲线差距明显,加剧了光伏发电的随机性[3]。大规模应用光伏发电并网对维持电网的稳定安全造成了极大的影响[4]。因此,电网需要精准的功率预测为其安全稳定运行提供可靠的数据支撑来应对光伏发电的随机波动性和间歇性带来的挑战,降低电力系统发生故障的风险,确保电力供应的可靠性和持续性。
功率预测方法主要包括数学统计预测法[5]、人工智能预测法和混合预测法[6]。数学统计预测法通常依赖于历史数据的模式识别和统计分析。如利用马尔科夫链模型预测未来数据[7],但这类方法可能无法捕捉到天气突变或设备故障等突发事件或不可预见的变化[8]。人工智能预测方法的原理是利用机器学习和数据挖掘等技术,从大量的历史数据中学习其潜在的分布规律,并将其应用于未来的预测[9]。采用两种或多种模型的混合方法可以在一定程度上提高预测的准确性。文献[10]使用完全集成经验模态分解将历史数据分解为多个序列,使用CNN-LSTM模型预测分解序列。文献[11]提出了MBI-PBI-ResNet深度学习网络模型用于光伏发电的超短期预测,并使用自适应仿射传播聚类进行相似日聚类,实现了分天气的预测任务。基于人工智能的预测方法需要大量高质量的历史发电数据作为训练预测模型的基础。但在实际工程中,由于技术故障或成本限制导致的数据缺失和因极端天气的罕见性导致的数据匮乏,削弱了预测模型训练的基础,导致预测性能下滑。因此,发电数据的不完整性和不稳定性给人工智能预测模型的训练和优化带来了极大的挑战。
随着机器学习在电力系统中的广泛应用,生成模型因在处理缺失数据、生成新的数据样本等方面的优秀性能,备受学者关注[12]。针对新能源预测采用的数据生成模型主要有迁移学习[13]和对抗生成网络[14]。但使用迁移学习进行数据增强可能会因源领域和目标领域之间可能存在的特征差异,导致源领域模型无法捕捉到目标领域的特征,影响模型的泛化能力。与迁移学习相比,对抗生成网络无需事先准备大量标记数据进行模型训练,在数据生成任务中具有更大的灵活性。文献[15]以Wasserstein距离作为损失函数,并增加梯度惩罚项的对抗生成网络增强少量真实数据,用于预测发电功率。然而,由于传统GAN以随机噪声作为输入,导致生成的数据类型难以预测,无法根据实际需求生成特定类型的数据[16]。为解决此类问题,条件对抗生成网络(conditional generative adversarial network,CGAN)在训练过程中增加一个标签变量作为辅助信息,来指导训练过程,增强模型的稳定性和学习能力。文献[17]以负荷关键值作为标签变量、相似曲线作为输入训练集,构建基于CGAN的负荷曲线生成模型,用于扩充预测模型的训练数据。文献[18]为解决负荷侧缺少样本数据和隐私泄露的问题,应用深度卷积条件生成对抗网络生成了大量逼近真实的用户用电数据,且可用于各类分类模型作为训练数据。
综上所述,为提高预测模型在极端天气下的预测精度,本文基于Wasserstein距离含梯度惩罚项的条件生成对抗网络(CWGAN-GP)和K-means聚类算法提出一种增强极端天气光伏数据的生成方法。首先通过K-means聚类得到各类天气的标签,作为条件输入训练CWGAN-GP,从而生成大量高质量的光伏发电数据;随后,提出一种基于CNN-SE-BiLSTM组合模型的超短期光伏功率预测方法,以满足不同天气类型的预测任务。最后,在算例分析中使用该预测模型验证了K-means-CWGAN-GP生成模型的可行性,并通过与其他预测模型的比较,说明了本文所提的预测方法的有效性。
由于光伏发电功率受周围多种环境变量影响,不同环境变量对发电功率的贡献度不同。为了避免某些贡献度较低的变量输入造成信息冗余干扰模型训练的效果,本文对各环境变量进行相关性分析,使用Pearson相关系数计算各环境变量相对于发电功率的相关系数以评估各个气候参数对发电功率的具体影响程度[18]。对于2个n维线性连续信号xy,其Pearson相关系数表达式如下式:
P x y = i = 1 n ( x i - x - ) ( y i - y - ) i = 1 n ( x i - x - ) 2 i = 1 n ( y i - y - ) 2
式中: P x y为Pearson相关系数; x i y i为第i个时刻信号; x - y -分别为信号xy的均值。
Pearson相关系数 P x y的值在[-1,1]之间,正值越大,说明两个变量正相关性越强;若相关性为负数,则说明两个变量负相关。
基于相关性分析的结果,舍弃与光伏发电功率相关度较低的环境变量,选择高度相关的环境变量作为模型的输入特征,有助提高模型的泛化能力和训练效率[20]
考虑到不同天气下的气候环境因素不同,每日的光伏发电功率曲线差异明显,因此对发电数据进行不同天气类别的聚类,针对不同聚类类别分别计划预测工作能有效提高预测准确率[21]
K-means聚类算法在处理数据分布均匀、簇结构规则的数据集时,通常能取得良好的聚类效果,且具有较强的可解释性。其基本的工作原理为:将样本集划分为k类,每一类对应一个聚类中心,通过迭代来实现簇的划分,最终目标是使样本离其聚类中心距离最近。其算法的目标是最小化所有数据点到其所属簇中心的距离之和,该距离数学表达式如下式所示:
J = i = 1 k j = 1 n | | x j ( i ) - μ i | | 2
式中:k为簇的数量;n为数据点的个数; x j ( i )为属于第i个簇的第j个数据点; μ i为第i个簇的中心点。
由于聚类的目标是使相同簇分类中相似度尽可能最高,而不同簇的样本差异性较大,因此本文用轮廓系数(silhouette coefficient,SC)衡量聚类结果的紧密度,其计算过程如下:
s ( i ) = b ( i ) - a ( i ) m a x { a ( i ) , b ( i ) }
式中: a ( i )为样本点i到相同簇内与其他样本点的平均距离; b ( i )为样本点i到其他簇与其他样本点的平均距离; s ( i )为所有数据点的轮廓系数的均值,即整个数据集的轮廓系数。
s ( i )的值越接近于1时,表示聚类结果非常紧密且明确,与不同聚类的分离度越大。
GAN的架构由两个网络组成:一个生成器(generator,G)和一个判别器(discriminator,D)[22]。G输入随机噪声,生成尽可能逼近真实数据分布的生成数据;D则通过学习真实数据的特征,判别输入数据来自真实样本还是G。在训练过程中,当G和D通过相互竞争找到一个纳什均衡状态时,则代表此时GAN的G可以生成能够准确地反映真实数据内在特征的生成数据,使得D无法区分样本的真实性,此时,G所生成的新的数据样本将同真实数据极度相似。
在GAN生成光伏发电出力的场景中,将光伏发电功率的历史数据作为训练集x。G以随机噪声z为输入,作为G的输入数据Gz);D的输入则为G的生成数据和真实数据。GAN的整体损耗由两部分组成:一部分来自G,另一部分来自D。GAN中G和D的损失函数如下式所示:
L G = E z ~ P z { D [ G ( z ) ] } L D = - E x ~ P x [ D ( x ) ] + E z ~ P z { D [ G ( z ) ] }
式中:LGLD分别为G和D的损失函数;E为对应样本的期望值。
D的目标是使判别器能够区分真实样本x和生成样本Gz)之间的差异,相反,G旨在生成的生成数据Gz)尽可能地接近真实数据x;在G和D的博弈过程中,需要得到一个纳什均衡点,使得两者的损失函数最小,其中极大极小化博弈模型如下式所示:
V ( G , D ) = E x ~ P x [ D ( x ) ] - E z ~ P z { D [ G ( z ) ] }
式中: V ( G , D )为GAN的价值函数。
在GAN的优化过程中,生成器和判别器的损失函数无法指导训练过程,容易出现模式崩溃的现象,而且作为输入的随机噪声z使得GAN无法根据实际要求生成特定标签的数据[23]。针对训练难以收敛的问题,WGAN-GP在传统GAN的基础上引入了Wasserstein距离,作为判别器评估生成数据分布与真实数据分布之间差异的新指标,直接计算判别器输出对于真实样本和生成样本期望值的差异,并引入了梯度惩罚(gradient penalty,GP)机制来改进对抗生成网络的训练过程,通过在真实数据分布和生成数据分布之间的线性插值点上添加一个额外的惩罚项,确保判别器的梯度在整个数据空间上变化平缓,从而间接实现了Lipschitz连续性,提高了训练的稳定性。其中Wasserstein距离的计算方法和梯度惩罚项如下:
W ( P x , P z ) = s u p L 1 E x ~ P x [ D ( x ) ] - E z ~ P z { D [ G ( z ) ] }
G P | x = E x ~ P x ( x D ( x ) 2 - 1 ) 2
式中: W ( P x , P z )为真实数据与生成数据两个分布的Wasserstein距离;L为1-Lipschitz的期望值差值的上界,L 1。
针对GAN缺乏可控性和生成数据的不确定性,CGAN在G和D的输入中引入额外的条件变量,使得GAN能够在训练过程中考虑这些条件,从而实现按需生成具有特定属性或标签的逼真样本,增强了生成模型的可控性和多样性[24]。CGAN中G和D的损失函数也做了改进,其数学表达式如下:
L G = E z ~ P z { D [ G ( z | y ) ] } L D = - E x ~ P x [ D ( x | y ) ] + E z ~ P z { D [ G ( z | y ) ] }
本文使用的CWGAN-GP网络结合了WGAN-GP和CGAN的特点,生成器G和判别器D可以稳定、有效地提取光伏发电特征;同时,可在训练过程中加入不同的标签y,充分挖掘不同聚类特征的相关性,得到不同标签下的风光出力数据集。根据式(4)和式(1)可以得出CWGAN-GP训练过程中博弈模型的目标函数,如下式所示:
V ( G , D ) = - E x ~ P x [ D ( x | y ) ] + E z ~ P z { D [ G ( z | y ) ] } + λ × G P
式中: λ为梯度惩罚系数。
CWGAN-GP的训练过程如图1所示。首先,标签信息被添加到输入数据中,作为控制其输出的附加条件,即真实的或生成的样本与其标签信息被组合成一个统一的向量,供给CWGAN-GP的判别器。然后,判别器对输入向量输出判断,只有当生成的数据和实际数据以及它们的标签相似且生成数据符合实际分布特征时,才能判定为真。
在传统的神经网络中所有的特征通常被同等对待,导致网络投入同样的算力处理对当前任务贡献度不大的特征,不利于预测模型的训练效率。甚至可能由于没有针对性地加强关键特征,导致训练过程中忽略了重要信息,从而影响最终的预测效果[25]。针对这类问题,本文使用通道注意力机制(squeeze-and-excitation networks,SENet)来分配不同通道之间的权重。如图2所示,通道注意力机制就是在训练过程中学习不同通道中的信息,自动获得各通道的权重,网络根据该权重增强重要特征,抑制贡献度相对较弱的通道特征。SENet的工作流程主要包括两个步骤[26]
1)压缩(squeeze)操作。压缩操作是运用全局池化压缩卷积层的输出特征图。把大小为H×W的二维矩阵映射为一个一维的通道权重,如下式所示:
z c = 1 H × W i = 1 H j = 1 W X c
式中: z c ( 1,1 , C )为每个通道的权重向量; X c ( H , W , C )为给定输入特征图;HWC分别为特征图的高度、宽度和通道数。
2)激励(excitation)操作。激励操作是对压缩后的特征 z c应用两层全连接层进行非线性转换,以学习通道间的权重,其计算过程如下:
α c = σ t [ W 2 R ( W 1 z c ) ]
σ t ( x ) = e x - e - x e x + e - x
R ( x ) = m a x ( 0 ,   x )
式中:W1W2分别为第1层和第2层全连接层的权重参数; σ tR分别为Sigmoid和ReLU激活函数。
最后,将激励输出 α c作为每个特征通道重要性的权重,然后对原始特征进行重加权,得到最终输出 Y c,如下式所示:
Y c = α c X c
卷积神经网络(convolutional neural network,CNN)是在实际应用中最为成功的网络,尤其是其在提取空间特征方面的能力极为出色。并且卷积神经网络可以通过简单串联不同大小卷积核的卷积层实现不同维度的高级抽象特征的提取。本文使用一维卷积层与ReLU激活函数串联作为一个卷积单元,其中一维卷积层提取光伏发电数据的一维非线性局部特征,ReLU函数则为神经网络提供非线性转换能力,有助于提高模型的泛化能力和解释性,同时减少不必要的计算资源消耗,卷积单元如图3所示。
双向长短期记忆神经网络(bidirectional long short-term memory,BiLSTM)是对传统单向长短期记忆网络(long short-term memory,LSTM)结构和性能的优化升级,通过整合两个方向的LSTM层来增强信息处理能力:一个是向前传播的LSTM层,另一个是向后传播的LSTM层。两个方向的层共同作用于输出结果,确保模型不仅能充分利用前向序列中的输入信息,还能捕捉后向序列中的历史背景信息,进而全面考量过去与未来的数据关联。这一特性使得BiLSTM在预测发电功率的任务上展现出更高的精确度[27]。具体来说,在预测光伏发电功率应用场景中,BiLSTM所对应的数学表达式如下式所示:
z i = W i n x i + W r e c a i - 1 z i = W i n x i + W r e c a i - 1 a i = σ t ( z i ) a i = σ t ( z i ) y i = W o u t a i + W o u t a i
式中:xi为第i个输入数据; z i z i 为前向传播层和反向传播层的加权输入; σ t ( x )为Tanh激活函数; W i n W i n 分别为输入层到前向传播层和反向传播层的权重矩阵; W r e c 为前一时刻的前向传播层输出到当前时刻前向传播层的权重矩阵; W r e c 为前一时刻的反向传播层输出到当前时刻反向传播层的权重矩阵; W o u t W o u t 分别为从前向和后向传播层到输出层的权重矩阵;yi为BiLSTM的输出。
BiLSTM结构图如图4所示。
由于光伏发电属于受环境条件高度影响的可再生能源发电方式,其功率输出具有很强的非线性和时变特性,所以本文使用了两个BiLSTM层串联,下层的BiLSTM可能捕捉到基础的局部和短程依赖,而上层的BiLSTM可以在此基础上学习更复杂、更全局的上下文信息。
基于K-means聚类的结果,将不同天气条件下的光伏发电量作为输入特征,并将聚类类别作为输入标签y。在CWGAN-GP中,判别器D的输入为96个采样点的光伏功率和环境因素构造的96×4矩阵,即一天的光伏出力特性,以及相应的一维聚类标签;输出则为一维的判别结果。生成器G的输入由符合标准正态分布的100×1维噪声和重构后的100×1维的聚类标签组成。根据判别器D的反馈和标签特征,在迭代学习中,生成器G学习光伏出力的潜在特征。最终,生成器G与判别器D达到纳什平衡,使得判别器D无法区分生成数据和真实数据的区别,即CWGAN-GP训练收敛。生成器G和判别器D的网络结构如表1表2所示,本文采用的训练超参数如表3所示。
本文结合K-means聚类算法和CWGAN-GP改进对抗生成网络,提出的一种光伏发电数据的生成方法。具体过程如下:
1)通过K-means聚类得到发电数据的聚类标签,将发电数据和对应的标签作为训练集。
2)将训练集作为输入,设置相关的训练参数后,训练CWGAN-GP网络。
3)根据聚类结果,选择极端天气的标签,将其作为训练完毕的生成器输入,生成极端天气数据。
4)将生成的极端天气发电数据,与真实数据结合,作为扩充数据集。
在本文的预测工作中,采用分天气类型预测并设置预测步长n=6,即模型基于前6个时刻的数据预测下一个时刻的数据。在实验中选取对应天气类型的最后3 d作为测试集,其余天数数据则作为训练集。在训练过程中,损失函数为均方误差(mean squared error,MSE),优化器选用Adam。由于通道注意力机制是一种嵌入式模块,专注于调整输入数据中不同通道之间的权重,故通道注意力的参数可以被独立调节。组合模型中的CNN和BiLSTM数据根据文献[28]设置,通道注意力机制的结构和主要参数如表4所示。此外,为保证模型性能可比性,CNN-BiLSTM,BiLSTM,LSTM和BP的主要参数与SE-CNN-BiLSTM一致。
为了验证预测模型的性能,本文分别用均方根误差(RMSE)和平均绝对误差(MAE)对预测结果进行评估,2种指标的表达式如下:
M A E = 1 n i = 1 n | P i - P ^ i | R M S E = 1 n i = 1 n ( P i - P ^ i ) 2
在CNN-SE-BiLSTM模型中,光伏发电数据在经过一个卷积单元和通道注意力后,进行空间特征提取和全局特征的重新校准,有选择地强调贡献度较高的特征,得到不同通道的权重后,与对应的特征通道相乘得到新的特征图,然后经过第二个卷积单元进行更高维的特征提取。再通过压平层调整数据结构,进入双层BiLSTM层,对数据的时序进行提取和学习。最后,进入全连接层和回归层,输出发电功率的预测值。CNN-SE-BiLSTM组合预测模型结构图如图5所示。
本文首先利用Pearson相关系数对数据集中的各环境变量进行相关性分析,确定与光伏发电功率相关性较强的环境变量作为CWGAN-GP对抗生成模型和CNN-SE-BiLSTM预测模型的输入。然后将一天96个数据点的发电功率作为一个特征向量,利用K-means聚类算法将一点的数据分为5类,并将其聚类类别作为CWGAN-GP的条件输入,与一天的发电数据合并作为CWGAN-GP的训练集,随后利用训练完毕的生成器生成极端天气的发电数据,具体做法是将极端天气的聚类类别作为条件输入,与随机噪声合并后输入到网络中,生成具有极端天气特征的发电数据。最后,将扩充后的数据作为CNN-SE-BiLSTM预测模型的训练集,CNN卷积层和SENet通道注意力机制SENet提取发电数据的空间特征,BiLSTM则提取发电数据的时间特征,结合前6个时间步长的数据,预测下一个时间步长的数据,具体实验过程如图6所示。
为了验证本文所提预测方法的有效性,以中国南方某大型光伏发电站为研究对象,所使用的数据集包括从2021年1月1日至2022年1月1日的5类关键环境因素实测数据和光伏发电功率实测值,时间间隔为15 min,每天一共96个数据点。首先,采用插值方法填补缺失值和修正异常值[29],然后对经过清洗的数据进行相关性分析,分别计算6类环境变量和光伏发电功率之间的Pearson系数,计算结果如图7所示。
图7可知,温度、辐照度和风速与光伏发电功率相关性较强,而与大气压的相关性较低,与云量呈负相关,故不能直接影响光伏发电的出力。考虑到预测和对抗生成模型的计算效率,故选择正相关性最高的三个环境因素,即辐照度、温度和风速,作为预测模型的输入特征,以提高预测的准确度。
本文使用每天的光伏发电功率96个数据点作为一个特征向量,将其作为K-means的输入。为了全面捕捉光伏发电功率在不同环境条件下的特征变化,选择光伏功率聚类为五个类别。聚类结果如图8所示。
在本次聚类分析中,依据K-means算法生成的五个类别分别包含不同数量的样本点,5个聚类类别的数量分别为122,68,82,54和39。第一类功率曲线呈现出较为平缓且稳定的特征,其发电能力在所有类别中达到最大值,因此将此类曲线归类为“晴天天气类型”。第二类功率曲线表现出较大的波动性,尽管其最大发电功率与晴天类别相近,但仍显示出与晴天状况有所区别的动态特性,故将其命名为“多云天气类型”。第三类功率曲线虽然有波动,但更加接近与晴天的大致波形,说明云量较少,光伏板受阳光直射更多,故将第三类曲线命名为“少云天气类型”。第四类功率曲线峰值相对较小且整体波动幅度较大,故命名为“阴雨天气类型”。第五类曲线幅值最低,并且在数据集中出现的频次最少,故将此类特殊曲线定义为“极端天气类型”。聚类结果的轮廓系数为0.697 2,说明设置K-means簇中心的个数为5可以有效地区分不同天气的发电功率曲线,具有较高的聚类质量。
为了验证CWGAN-GP模型生成的极端样本在预测模型中的可用性,本节对极端天气条件下的发电数据进行了不同程度的增强,并将增强后的数据集用作预测模型的训练集。首先,使用真实发电数据及其聚类标签来训练CWGAN-GP模型。训练完成并收敛后,将极端天气的标签作为CWGAN-GP的输入,以生成相应的极端天气发电数据。出于篇幅考虑,随机选择30 d的生成数据展示,如图9所示。
图9可以看出,经过训练的CWGAN-GP模型能够利用光伏真实数据集中各类数据的内在特征,从而隐式地学习数据的各种未知特性,并生成大量接近真实数据的合成数据。所得数据集能够很好地反映真实值的分布情况。同时,生成数据的平均值与测试集真实值的平均值呈现相似的趋势。然后利用CWGAN-GP生成的数据对原始发电数据进行不同大小的扩充,扩充量分别为5 d,10 d,15 d和20 d。将生成的发电数据和原始数据拼接作为训练集,以原始数据集的最后三天的数据作为测试集。将训练集作为输入,训练CNN-SE-BiLSTM预测模型,图10为扩充不同天数训练集的预测曲线,表5为预测误差指标。
图10表5可知,经过扩充的原始极端天气训练集能够有效提高预测模型的预测精度,并且随着训练集扩充比例的增加,预测模型对发电功率的预测精度也随之提高。与未扩充的原始训练集相比,扩充天数为5 d,10 d,15 d和20 d在测试集的MAE分别降低了47.45%,49.10%,49.56%和49.91%。RMSE分别降低了47.47%,48.72%,50.33%和50.51%。
为验证本文所提CNN-SE-BiLSTM预测模型在光伏功率预测方面具有优越性能,分别构建BiLSTM网络、LSTM网络和反向传播神经网络(backpropagation(BP)neural network)4种模型作为对比模型,以每种天气类型的后三天作为预测模型的测试集,其余作为训练集,对比4种模型预测精度,结果分别如图11表6所示。
图11所示,不同预测模型的预测曲线与实际光伏发电功率曲线之间的拟合程度存在差异。由于光伏功率具有较强的波动性和随机性,且训练数据集时间跨度较长,因此结构简单的预测模型的预测结果的误差较大。本文所提的CNN-SE-BiLSTM在五类天气中的预测曲线较其他模型都更加与实际发电曲线拟合。
表6得到:
1)与BP和LSTM模型相比,BiLSTM预测模型在五类天气中的误差都较低,其中在多云天气的预测任务中最为明显,BiLSTM模型的测试集误差MAE较BP模型和LSTM模型分别降低了52.84%,39.94%,RMSE分别降低了54.23%,21.49%,这表明BiLSTM模型在光伏发电预测这样长时间序列预测任务中更具优势。
2)与BiLSTM模型相比,CNN-SE-BiLSTM在各类天气的预测误差都有不同程度的降低,在五类天气中,MAE分别降低了12.40%,10.70%,29.27%,20.11%和33.54%。RMSE分别降低了46.56%,2.05%,32.71%,4.61%和21.59%。这说明通道注意力机制和卷积网络能有效增强数据特征提取能力,提高了组合预测模型的预测精度。
本文针对极端天气光伏发电数据匮乏的问题,提出了基于数据驱动的CWGAN-GP和CNN-SE-BiLSTM超短期预测模型。通过算例分析得到如下结论:
1)使用K-means聚类得到各天气作为输入标签,初步区分不同天气下的光伏功率曲线。CWGAN-GP可以学习与各天气下真实光伏发电数据的分布规律,并综合各天气的特征生成逼近真实数据的极端天气数据,实现样本增强。通过实验证明,样本扩充的数量越多,预测模型的误差越小,说明CWGAN-GP生成的数据可以被用于各类人工智能预测模型,且CWGAN-GP可以通过改变输入标签,对不同天气类型的数据进行扩充,以解决数据样本不平衡的问题,稳定并提升预测模型在各种天气下的预测精度。
2)通过对比分析及实验验证,本文提出的CNN-SE-BiLSTM的光伏发电功率模型,较其他传统预测模型有更低的预测误差,且在不同的天气下都表现出了较高的预测精度,说明本文所提的预测模型能有效应用于光伏功率的超短期预测任务。
  • 云南省重大科技专项计划(202002AF080001)
参考文献 引证文献
排序方式:
[1]
江秀臣, 许永鹏, 李曜丞, 等. 新型电力系统背景下的输变电数字化转型[J]. 高电压技术, 2022, 48(1):1-10.
JIANG Xiuchen, XU Yongpeng, LI Yaocheng, et al. Digitalization transformation of power transmission and transformation under the background of new power system[J]. High Voltage Engineering, 2022, 48(1):1-10.
[2]
肖瑶, 钮文泽, 魏高升, 等. 太阳能光伏/光热技术研究现状与发展趋势综述[J]. 发电技术, 2022, 43(3):392-404.
XIAO Yao, NIU Wenze, WEI Gaosheng, et al. Review on research status and developing tendency of solar photovoltaic/thermal technology[J]. Power Generation Technology, 2022, 43(3):392-404.
[3]
卢俊杰, 蔡涛, 郎建勋, 等. 基于集群划分的光伏电站集群发电功率短期预测方法[J]. 高电压技术, 2022, 48(5):1943-1951.
LU Junjie, CAI Tao, LANG Jianxun, et al. Short-term power output forecasting of clustered photovoltaic solar plants based on cluster partition[J]. High Voltage Engineering, 2022, 48(5):1943-1951.
[4]
石文辉, 屈姬贤, 罗魁, 等. 高比例新能源并网与运行发展研究[J]. 中国工程科学, 2022, 24(6):52-63.
SHI Wenhui, QU Jixian, LUO Kui, et al. Grid-integration and operation of high proportioned new energy[J]. Strategic Study of CAE, 2022, 24(6):52-63.
[5]
谭津, 邓长虹, 杨威, 等. 微电网光伏发电的Adaboost天气聚类超短期预测方法[J]. 电力系统自动化, 2017, 41(21):33-39.
TAN Jin, DENG Changhong, YANG Wei, et al. Ultra-short-term photovoltaic power forecasting in microgrid based on Adaboost clustering[J]. Automation of Electric Power Systems, 2017, 41(21):33-39.
[6]
王洪坤, 葛磊蛟, 李宏伟, 等. 分布式光伏发电的特性分析与预测方法综述[J]. 电力建设, 2017, 38(7):1-9.
WANG Hongkun, GE Leijiao, LI Hongwei, et al. A review on characteristic analysis and prediction method of distributed PV[J]. Electric Power Construction, 2017, 38(7):1-9.
[7]
王继拓, 王万成, 陈宏伟. 基于回归——马尔科夫链的光伏发电功率预测[J]. 电测与仪表, 2019, 56(1):76-81.
WANG Jituo, WANG Wancheng, CHEN Hongwei. Photovoltaic power generation forecasting based on regression——Markov chain[J]. Electrical Measurement & Instrumentation, 2019, 56(1):76-81.
[8]
李安寿, 陈琦, 王子才, 等. 光伏发电系统功率预测方法综述[J]. 电气传动, 2016, 46(6):93-96.
LI Anshou, CHEN Qi, WANG Zicai, et al. Review of power forecast methods for photovoltaic generating system[J]. Electric Drive, 2016, 46(6):93-96.
[9]
HU Z, GAO Y, JI S, et al. Improved multistep ahead photovol-taic power prediction model based on LSTM and self-attention with weather forecast data[J]. Applied Energy, 2024, 359(2):122709.
[10]
GAO B, HUANG X, SHI J, et al. Hourly forecasting of solar irradiance based on CEEMDAN and multi-strategy CNN-LSTM neural networks[J]. Renewable Energy, 2020, 162(2):1665-1683.
[11]
黄泽, 毕贵红, 谢旭, 等. 基于MBI-PBI-ResNet的超短期光伏功率预测[J]. 电力系统保护与控制, 2024, 52(2):165-176.
HUANG Ze, BI Guihong, XIE Xu, et al. Ultra-short-term PV power prediction based on MBI-PBI-ResNet[J]. Power System Protection and Control, 2024, 52(2):165-176.
[12]
邵振国, 张承圣, 陈飞雄, 等. 生成对抗网络及其在电力系统中的应用综述[J]. 中国电机工程学报, 2023, 43(3):987-1004.
SHAO Zhenguo, ZHANG Chengsheng, CHEN Feixiong, et al. A review on generative adversarial networks for power system applications[J]. Proceedings of the CSEE, 2023, 43(3):987-1004.
[13]
HU Q H, ZHANG R J, ZHOU Y C. Transfer learning for short-term wind speed prediction with deep neural networks[J]. Renewable Energy, 2016, 85(1):83-95.
[14]
WANG Fei, ZHANG Zhanyao, LIU Chun, et al. Generative adversarial networks and convolutional neural networks based weather classification model for day ahead short-term photovoltaic power forecasting[J]. Energy Conversion and Management, 2019, 181(2):443-462.
[15]
殷豪, 张铮, 丁伟锋, 等. 基于生成对抗网络和LSTM-CSO的少样本光伏功率短期预测[J]. 高电压技术, 2022, 48(11):4342-4351.
YIN Hao, ZHANG Zheng, DING Weifeng, et al. Short-term prediction of small-sample photovoltaic power based on generative adversarial network and LSTM-CSO[J]. High Voltage Engineering, 2022, 48(11):4342-4351.
[16]
董骁翀, 孙英云, 蒲天骄. 基于条件生成对抗网络的可再生能源日前场景生成方法[J]. 中国电机工程学报, 2020, 40(17):5527-5536.
DONG Xiaochong, SUN Yingyun, PU Tianjiao. Day-ahead scenario generation of renewable energy based on conditional GAN[J]. Proceedings of the CSEE, 2020, 40(17):5527-5536.
[17]
孙浩, 万灿, 曹照静, 等. 基于条件生成对抗网络曲线生成的短期负荷概率预测[J]. 电力系统自动化, 2023, 47(23):189-199.
SUN Hao, WAN Can, CAO Zhaojing, et al. Short-term load probabilistic forecasting based on conditional generative adversarial network curve generation[J]. Automation of Electric Power Systems 2023, 47(23):189-199.
[18]
LI Jianbin, CHEN Zhiqiang, CHENG Long, et al. Energy data generation with Wasserstein deep convolutional generative adversarial networks[J]. Energy, 2022, 257(15):1-17.
[19]
贾鹏云. 基于深度学习算法的短期光伏发电功率预测研究[D]. 北京: 华北电力大学(北京), 2021.
JIA Pengyun. Research on short term photovoltaic power prediction based on deep learning[D]. Beijing: North China Electric Power University (Beijing), 2021.
[20]
陈嘉铭. 基于深度学习和强化学习的光伏发电功率预测研究[D]. 广州: 广东工业大学, 2022.
CHEN Jiaming. Research on photovoltaic power forecasting based on deep learning and reinforcement learning[D]. Guangzhou: Guangdong University of Technology, 2022.
[21]
孟安波, 陈嘉铭, 黎湛联, 等. 基于相似日理论和CSO-WGPR的短期光伏发电功率预测[J]. 高电压技术, 2021, 47(4):1176-1184.
MENG Anbo, CHEN Jiaming, LI Zhanlian, et al. Short-term photovoltaic power generation prediction based on similar day theory and CSO-WGPR[J]. High Voltage Engineering, 2021, 47(4):1176-1184.
[22]
QIAO J, PU T J, WANG X Y. Renewable scenario generation using controllable generative adversarial networks with transparent latent space[J]. CSEE Journal of Power and Energy Systems, 2021, 7(1):66-77.
[23]
刘云鹏, 许自强, 和家慧, 等. 基于条件式Wasserstein生成对抗网络的电力变压器故障样本增强技术[J]. 电网技术, 2020, 44(4):1505-1513.
LIU Yunpeng, XU Ziqiang, HE Jiahui, et al. Data augmentation method for power transformer fault diagnosis based on conditional Wasserstein generative adversarial network[J]. Power System Technology, 2020, 44(4):1505-1513.
[24]
王成磊, 匡熠, 濮永现, 等. 基于改进生成对抗网络场景生成的长-短期储能优化配置[J]. 高电压技术, 2024, 50(3):1131-1144.
WANG Chenglei, KUANG Yi, PU Yongxian, et al. Optimal configuration of long-term and short-term energy storage based on improved generative adversarial network scenarios[J]. High Voltage Engineering, 2024, 50(3):1131-1144.
[25]
任欢, 王旭光. 注意力机制综述[J]. 计算机应用, 2021, 41(S1):1-6.
REN Huan, WANG Xuguang. Review of attention mechanism[J]. Journal of Computer Applications, 2021, 41(S1):1-6.
[26]
HU J, SHENG L, SUN G. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8):2011-2023.
[27]
毕贵红, 赵鑫, 陈臣鹏, 等. 基于多通道输入和PCNN-BiLSTM的光伏发电功率超短期预测[J]. 电网技术, 2022, 46(9):3463-3476.
BI Guihong, ZHAO Xin, CHEN Chenpeng. Ultra-short-term prediction of photovoltaic power generation based on multi-channel input and PCNN-BiLSTM[J]. Power System Technology, 2022, 46(9):3463-3476.
[28]
王开艳, 杜浩东, 贾嵘, 等. 基于相似日聚类和QR-CNN-BiLSTM模型的光伏功率短期区间概率预测[J]. 高电压技术, 2022, 48(11):4372-4388.
WANG Kaiyan, DU Haodong, JIA Rong, et al. Short-term interval probability prediction of photovoltaic power based on similar daily clustering and QR-CNN-BiLSTM model[J]. High Volta-ge Engineering, 2022, 48(11):4372-4388.
[29]
乔依林. 风电运行数据预处理技术及其应用研究[D]. 北京: 华北电力大学(北京), 2019.
QIAO Yilin. Research on preprocessing of wind power operation data and its application[J]. Beijing: North China Electric Power University (Beijing), 2019.
2025年第55卷第8期
PDF下载
374
178
引用本文
BibTeX
文章信息
doi: 10.19457/j.1001-2095.dqcd25932
  • 接收时间:2024-05-13
  • 首发时间:2025-10-29
  • 出版时间:2025-08-20
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-05-13
  • 修回日期:2024-07-01
基金
云南省重大科技专项计划(202002AF080001)
作者信息
    1.昆明理工大学 电力工程学院,云南 昆明 650051
    2.云南电网有限责任公司 电力科学研究院,云南 昆明 650217

通讯作者:

王成磊(1997—),男,硕士,主要研究方向为人工智能在电力系统中的运用,Email:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/dqcd/CN/10.19457/j.1001-2095.dqcd25932
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