Article(id=1156264266425553548, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156264148657886112, articleNumber=null, orderNo=null, doi=10.12404/j.issn.1671-1815.2402025, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1711209600000, receivedDateStr=2024-03-24, revisedDate=1733760000000, revisedDateStr=2024-12-10, acceptedDate=null, acceptedDateStr=null, onlineDate=1753604483467, onlineDateStr=2025-07-27, pubDate=1740672000000, pubDateStr=2025-02-28, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1753604483467, onlineIssueDateStr=2025-07-27, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1753604483467, creator=13701087609, updateTime=1753604483467, updator=13701087609, issue=Issue{id=1156264148657886112, tenantId=1146029695717560320, journalId=1146123166801305609, year='2025', volume='25', issue='6', pageStart='2193', pageEnd='2636', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1753604455388, creator=13701087609, updateTime=1753771257443, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1156963767234945803, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156264148657886112, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1156963767234945804, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156264148657886112, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=2397, endPage=2405, ext={EN=ArticleExt(id=1156264267528655502, articleId=1156264266425553548, tenantId=1146029695717560320, journalId=1146123166801305609, language=EN, title=Short Term Prediction of Wind Power Based on Error Compensation and IDBO-BiLSTM, columnId=1156262729162810294, journalTitle=Science Technology and Engineering, columnName=Papers·Automation and Computational Technology, runingTitle=null, highlight=null, articleAbstract=

Aiming at the problem of poor model accuracy caused by poor stability and strong randomness of wind power output. A short-term prediction model of wind power based on quadratic decomposition error compensation was proposed. Firstly, BiLSTM (bidirectional long short-term memory) prediction model is established to predict wind power and output prediction errors. Secondly, an IDBO (improved dung beetle optimizer) algorithm was used to initialize the population by using chaotic mapping, update the position of rolling dung beetles by introducing golden sine strategy, and update the position of thieving dung beetles by adding dynamic adaptive weight coefficient to optimize the parameters of the prediction model. Prevent the network from falling into the local optimal solution, and adaptively search the optimal parameter combination. Then, using the decomposition-reconstruction-decomposition strategy, CEEMDAN (complete ensemble empirical mode decomposition with adaptive noise) was used for the first decomposition. In addition, SE(sample entropy) and K-means are introduced to reconstruct the sequence according to frequency, and the high-frequency error sequence was decomposed into error sequences of different frequency bands by VMD(variational mode decomposition). Improve the prediction efficiency and accuracy of subsequent models. Finally, the input error compensation model of each component was used to predict and the Attention mechanism was introduced to learn the feature relationship of different time steps and give different weight values to enhance the attention to key information. Through the measured data of a wind farm in Xinjiang, the prediction accuracy of the proposed model is proved to be high and has significant advantages.

, correspAuthors=Xue-song JIANG, 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=Zhen-yu WEI, Xue-song JIANG, Li-fa YANG), CN=ArticleExt(id=1156264322650199033, articleId=1156264266425553548, tenantId=1146029695717560320, journalId=1146123166801305609, language=CN, title=基于误差补偿及IDBO-BiLSTM的风电功率短期预测, columnId=1156262729783567290, journalTitle=科学技术与工程, columnName=论文·自动化技术、计算机技术, runingTitle=null, highlight=null, articleAbstract=

针对风电出力稳定性差、随机性强而导致的模型精度差的问题。提出了一种基于二次分解误差补偿的风电功率短期预测模型。首先建立双向长短期记忆(bidirectional long short-term memory, BiLSTM)预测模型对风电功率进行预测并输出预测误差。其次,采用了一种利用混沌映射初始化种群、引入黄金正弦策略更新滚球蜣螂位置,并添加动态自适应性权重系数来更新偷窃蜣螂的位置的改进蜣螂优化算法(improved dung beetle optimizer, IDBO)对预测模型参数寻优,防止网络陷入局部最优解,自适应搜寻最优参数组合。然后,采用分解-重构-分解的策略,利用自适应噪声的完全集合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise, CEEMDAN)进行首次分解,并且引入样本熵(sample entropy, SE)与K均值(K-means)将序列按频率进行重构并通过变分模态分解(variational mode decomposition, VMD)将高频误差序列分解成不同频段的误差序列,提高后续模型的预测效率及预测精度。最后,将各分量输入误差补偿模型进行预测并引入Attention机制学习不同时间步的特征关系,并给与不同权重值,加强对关键信息的注意力。通过新疆达坂城风电场实测数据验证了所提模型预测精度高,具有显著优势。

, correspAuthors=姜雪松, authorNote=null, correspAuthorsNote=
* 姜雪松(1979—),男,汉族,黑龙江佳木斯人,博士,副教授。研究方向:工业工程与管理、智能制造工艺与装备。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=zXGJ0Us4BrM6dBbBQ0okIw==, magXml=tLckWyVAYhRdMVyEdmlPKQ==, pdfUrl=null, pdf=OfLPUUb/F5XQV48I94XmfQ==, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=HJvRTbGy7xyZ2rLvb+NEeQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=tYQEeM6D01fA8giDjOlwng==, mapNumber=null, authorCompany=null, fund=null, authors=

魏振宇(2000—),男,汉族,山东济宁人,硕士研究生。研究方向:短期风电功率预测。E-mail:

, authorsList=魏振宇, 姜雪松, 杨立发)}, authors=[Author(id=1233422548055150958, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=1634945056@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1233422548134842745, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, authorId=1233422548055150958, language=EN, stringName=Zhen-yu WEI, firstName=Zhen-yu, middleName=null, lastName=WEI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin 150040, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1233422548235506052, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, authorId=1233422548055150958, language=CN, stringName=魏振宇, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 东北林业大学机电工程学院, 哈尔滨 150040, bio={"content":"

魏振宇(2000—),男,汉族,山东济宁人,硕士研究生。研究方向:短期风电功率预测。E-mail:

"}, bioImg=null, bioContent=

魏振宇(2000—),男,汉族,山东济宁人,硕士研究生。研究方向:短期风电功率预测。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1233422547761549649, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, xref=1, ext=[AuthorCompanyExt(id=1233422547765743954, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547761549649, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin 150040, China), AuthorCompanyExt(id=1233422547774132563, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547761549649, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东北林业大学机电工程学院, 哈尔滨 150040)])]), Author(id=1233422548369723790, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=xuesongjiang@nefu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1233422548470387097, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, authorId=1233422548369723790, language=EN, stringName=Xue-song JIANG, firstName=Xue-song, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin 150040, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1233422548612993444, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, authorId=1233422548369723790, language=CN, stringName=姜雪松, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 东北林业大学机电工程学院, 哈尔滨 150040, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1233422547761549649, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, xref=1, ext=[AuthorCompanyExt(id=1233422547765743954, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547761549649, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin 150040, China), AuthorCompanyExt(id=1233422547774132563, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547761549649, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东北林业大学机电工程学院, 哈尔滨 150040)])]), Author(id=1233422548747211184, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, 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=1233422548881428926, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, authorId=1233422548747211184, language=EN, stringName=Li-fa YANG, firstName=Li-fa, middleName=null, lastName=YANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 703 Research Institute, China State Shipbuilding Corporation, Harbin 150783, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1233422550311686600, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, authorId=1233422548747211184, language=CN, stringName=杨立发, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 中国船舶集团有限公司第七O三研究所, 哈尔滨 150783, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1233422547891573086, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, xref=2, ext=[AuthorCompanyExt(id=1233422547908350304, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547891573086, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 703 Research Institute, China State Shipbuilding Corporation, Harbin 150783, China), AuthorCompanyExt(id=1233422547912544609, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547891573086, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 中国船舶集团有限公司第七O三研究所, 哈尔滨 150783)])])], keywords=[Keyword(id=1233422550559150562, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, orderNo=1, keyword=wind power short-term forecast), Keyword(id=1233422550676591088, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, orderNo=2, keyword=bidirectional long short-term memory network), Keyword(id=1233422550810808830, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, orderNo=3, keyword=improved dung beetle optimization algorithm), Keyword(id=1233422550928249354, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, orderNo=4, keyword=complete ensemble empirical mode decomposition), Keyword(id=1233422551062467098, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, orderNo=5, keyword=variational mode decomposition), Keyword(id=1233422551179907621, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, orderNo=1, keyword=风电功率短期预测), Keyword(id=1233422551318319670, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, orderNo=2, keyword=双向长短期记忆网络), Keyword(id=1233422551439954500, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, orderNo=3, keyword=改进蜣螂优化算法), Keyword(id=1233422551578366543, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, orderNo=4, keyword=完全集合经验模态分解), Keyword(id=1233422551716778588, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, orderNo=5, keyword=变分模态分解)], refs=[Reference(id=1233422556313736111, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=39, issue=11, pageStart=142, pageEnd=150, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=欧旭鹏, 唐云, 张凯, journalName=电网与清洁能源, refType=null, unstructuredReference=欧旭鹏, 唐云, 张凯, 等. 基于CEEMDAN-IDOA-BiLSTM的超短期风电功率预测[J]. 电网与清洁能源, 2023, 39(11): 142-150., articleTitle=基于CEEMDAN-IDOA-BiLSTM的超短期风电功率预测, refAbstract=null), Reference(id=1233422556422788028, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=39, issue=11, pageStart=142, pageEnd=150, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=Ou Xupeng, Tang Yun, Zhang Kai, journalName=Power Grids and Clean Energy, refType=null, unstructuredReference=Ou Xupeng, Tang Yun, Zhang Kai, et al. Ultra-short term wind power prediction based on CEEMDAN-IDOA-BiLSTM[J]. Power Grids and Clean Energy, 2023, 39(11): 142-150., articleTitle=Ultra-short term wind power prediction based on CEEMDAN-IDOA-BiLSTM, refAbstract=null), Reference(id=1233422556565394376, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2021, volume=45, issue=7, pageStart=28, pageEnd=36, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=Hu S, Xiang Y, Shen X D, journalName=Automation of Electric Power Systems, refType=null, unstructuredReference=Hu S, Xiang Y, Shen X D, et al. Wind power prediction model considering meteorological factor and spatial correlation of wind speed[J]. Automation of Electric Power Systems, 2021, 45(7): 28-36., articleTitle=Wind power prediction model considering meteorological factor and spatial correlation of wind speed, refAbstract=null), Reference(id=1233422556703806419, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=30, issue=8, pageStart=70, pageEnd=74, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=王泽华, 徐兴国, 叶子臣, journalName=技术与市场, refType=null, unstructuredReference=王泽华, 徐兴国, 叶子臣. 基于改进麻雀搜索算法优化SVM的风电功率预测研究[J]. 技术与市场, 2023, 30(8): 70-74, 79., articleTitle=基于改进麻雀搜索算法优化SVM的风电功率预测研究, refAbstract=null), Reference(id=1233422556934493157, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=30, issue=8, pageStart=70, pageEnd=74, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=Wang Zehua, Xu Xingguo, Ye Zichen, journalName=Technology & Field, refType=null, unstructuredReference=Wang Zehua, Xu Xingguo, Ye Zichen. Wind power prediction based on improved sparrow search algorithm and SVM optimization[J]. Technology & Field, 2023, 30(8): 70-74, 79., articleTitle=Wind power prediction based on improved sparrow search algorithm and SVM optimization, refAbstract=null), Reference(id=1233422557072905198, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2021, volume=44, issue=12, pageStart=63, pageEnd=69, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=谢波, 高建宇, 张惠娟, journalName=电子测量技术, refType=null, unstructuredReference=谢波, 高建宇, 张惠娟, 等. 短期风电功率预测中的IOFA-SVM算法实现[J]. 电子测量技术, 2021, 44(12): 63-69., articleTitle=短期风电功率预测中的IOFA-SVM算法实现, refAbstract=null), Reference(id=1233422557228094461, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2021, volume=44, issue=12, pageStart=63, pageEnd=69, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=Xie Bo, Gao Jianyu, Zhang Huijuan, journalName=Electronic Measurement Technology, refType=null, unstructuredReference=Xie Bo, Gao Jianyu, Zhang Huijuan, et al. Implementation of IOFA-SVM algorithm for short-term wind power prediction[J]. Electronic Measurement Technology, 2021, 44(12): 63-69., articleTitle=Implementation of IOFA-SVM algorithm for short-term wind power prediction, refAbstract=null), Reference(id=1233422557328756741, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=43, issue=12, pageStart=273, pageEnd=281, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=Yang G Q, Liu S I, Wang D Y, journalName=Acta Energiae Solaris Sinica, refType=null, unstructuredReference=Yang G Q, Liu S I, Wang D Y, et al. Short-term wind power forecasting based on attention-gru wind speed correction and stacking[J]. Acta Energiae Solaris Sinica, 2022, 43(12): 273-281., articleTitle=Short-term wind power forecasting based on attention-gru wind speed correction and stacking, refAbstract=null), Reference(id=1233422557462974477, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=43, issue=11, pageStart=140, pageEnd=147, url=null, language=null, rfNumber=[6], rfOrder=8, authorNames=Qi F Z, Zhuo K X, Cao J, journalName=Acta Energiae Solaris Sinica, refType=null, unstructuredReference=Qi F Z, Zhuo K X, Cao J. Short-term wind power probability density prediction method based on multi-level semantic attention mechanism[J]. Acta Energiae Solaris Sinica, 2022, 43(11): 140-147., articleTitle=Short-term wind power probability density prediction method based on multi-level semantic attention mechanism, refAbstract=null), Reference(id=1233422557622358041, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=35, pageStart=15091, pageEnd=15099, url=null, language=null, rfNumber=[7], rfOrder=9, authorNames=程杰, 严婷, 窦春霞, journalName=科学技术与工程, refType=null, unstructuredReference=程杰, 严婷, 窦春霞, 等. 基于GWO-CNN-BiLSTM的超短期风电预测[J]. 科学技术与工程, 2023, 23(35): 15091-15099., articleTitle=基于GWO-CNN-BiLSTM的超短期风电预测, refAbstract=null), Reference(id=1233422557769158695, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2019, volume=23, issue=35, pageStart=15091, pageEnd=15099, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=Cheng Jie, Yan Ting, Dou Chunxia, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Cheng Jie, Yan Ting, Dou Chunxia, et al. Ultra-short term wind power prediction based on GWO-CNN-BiLSTM[J]. Science Technology and Engineering, 2019, 23(35): 15091-15099., articleTitle=Ultra-short term wind power prediction based on GWO-CNN-BiLSTM, refAbstract=null), Reference(id=1233422559291691060, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=25, issue=4, pageStart=44, pageEnd=49, url=null, language=null, rfNumber=[8], rfOrder=11, authorNames=丁同, 傅晓锦, 刘明旺, journalName=扬州大学学报(自然科学版), refType=null, unstructuredReference=丁同, 傅晓锦, 刘明旺. 基于GA-VMD-BiLSTM算法的风电功率预测[J]. 扬州大学学报(自然科学版), 2022, 25(4): 44-49., articleTitle=基于GA-VMD-BiLSTM算法的风电功率预测, refAbstract=null), Reference(id=1233422559438491715, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=25, issue=4, pageStart=44, pageEnd=49, url=null, language=null, rfNumber=[8], rfOrder=12, authorNames=Ding Tong, Fu Xiaojin, Liu Mingwang, journalName=Journal of Yangzhou University (Natural Science Edition), refType=null, unstructuredReference=Ding Tong, Fu Xiaojin, Liu Mingwang. Wind power prediction based on GA-VMD-BiLSTM algorithm[J]. Journal of Yangzhou University (Natural Science Edition), 2022, 25(4): 44-49., articleTitle=Wind power prediction based on GA-VMD-BiLSTM algorithm, refAbstract=null), Reference(id=1233422559593680977, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=50, issue=4, pageStart=27, pageEnd=34, url=null, language=null, rfNumber=[9], rfOrder=13, authorNames=高鹭, 孔繁苗, 张飞, journalName=智慧力, refType=null, unstructuredReference=高鹭, 孔繁苗, 张飞, 等. 基于IPSO-BiLSTM-AM模型的超短期风电功率预测方法[J]. 智慧力, 2022, 50(4): 27-34., articleTitle=基于IPSO-BiLSTM-AM模型的超短期风电功率预测方法, refAbstract=null), Reference(id=1233422559765647455, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2012, volume=50, issue=4, pageStart=27, pageEnd=34, url=null, language=null, rfNumber=[9], rfOrder=14, authorNames=Gao Lu, Kong Fanmiao, Zhang Fei, journalName=Smart Power, refType=null, unstructuredReference=Gao Lu, Kong Fanmiao, Zhang Fei, et al. Ultrashort term wind power prediction method based on IPSO-BiLSTM-AM model[J]. Smart Power, 2012, 50(4): 27-34., articleTitle=Ultrashort term wind power prediction method based on IPSO-BiLSTM-AM model, refAbstract=null), Reference(id=1233422559874699366, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=44, issue=11, pageStart=239, pageEnd=246, url=null, language=null, rfNumber=[10], rfOrder=15, authorNames=Xiao L X, Zhang Y, Zhou H, journalName=Acta Energiae Solaris Sinica, refType=null, unstructuredReference=Xiao L X, Zhang Y, Zhou H, et al. Ultra short trem wind power prediction based on IAOA-VMD-LSTM[J]. Acta Energiae Solaris Sinica, 2023, 44(11): 239-246., articleTitle=Ultra short trem wind power prediction based on IAOA-VMD-LSTM, refAbstract=null), Reference(id=1233422560013111407, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=46, issue=3, pageStart=897, pageEnd=904, url=null, language=null, rfNumber=[11], rfOrder=16, authorNames=Sheng S Q, Jin H Y, Liu C R, journalName=Power System Technology, refType=null, unstructuredReference=Sheng S Q, Jin H Y, Liu C R. Short-term and mid-short-term wind power forecasting based on VMD-WSGRU[J]. Power System Technology, 2022, 46(3): 897-904., articleTitle=Short-term and mid-short-term wind power forecasting based on VMD-WSGRU, refAbstract=null), Reference(id=1233422560180883577, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=47, issue=12, pageStart=5015, pageEnd=8026, url=null, language=null, rfNumber=[12], rfOrder=17, authorNames=Chang Y F, Yang Z X, Pan F, journalName=Power System Technology, refType=null, unstructuredReference=Chang Y F, Yang Z X, Pan F, et al. Ultra-short-term wind power prediction based on CEEMDAN-PE-WPD and multi-objective optimization[J]. Power System Technology, 2023, 47(12): 5015-8026., articleTitle=Ultra-short-term wind power prediction based on CEEMDAN-PE-WPD and multi-objective optimization, refAbstract=null), Reference(id=1233422560378015880, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2022, volume=22, issue=12, pageStart=4828, pageEnd=4834, url=null, language=null, rfNumber=[13], rfOrder=18, authorNames=伍俊杰, 张倩, 李国丽, journalName=科学技术与工程, refType=null, unstructuredReference=伍俊杰, 张倩, 李国丽, 等. 计及误差修正的变分模态分解-长短期记忆神经网络短期负荷预测[J]. 科学技术与工程, 2022, 22(12): 4828-4834., articleTitle=计及误差修正的变分模态分解-长短期记忆神经网络短期负荷预测, refAbstract=null), Reference(id=1233422560575148182, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2002, volume=22, issue=12, pageStart=4828, pageEnd=4834, url=null, language=null, rfNumber=[13], rfOrder=19, authorNames=Wu Junjie, Zhang Qian, Li Guoli, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Wu Junjie, Zhang Qian, Li Guoli, et al. Variational mode decomposition with error correction for short-term load forecasting by long short-term memory neural networks[J]. Science Technology and Engineering, 2002, 22(12): 4828-4834., articleTitle=Variational mode decomposition with error correction for short-term load forecasting by long short-term memory neural networks, refAbstract=null), Reference(id=1233422560730337444, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=79, issue=7, pageStart=7305, pageEnd=7336, url=null, language=null, rfNumber=[14], rfOrder=20, authorNames=Xue J, Shen B, journalName=The Journal of Supercomputing, refType=null, unstructuredReference=Xue J, Shen B. Dung beetle optimizer: a new meta heuristic algorithm for global optimization[J]. The Journal of Supercomputing, 2023, 79(7): 7305-7336., articleTitle=Dung beetle optimizer: a new meta heuristic algorithm for global optimization, refAbstract=null), Reference(id=1233422560835195055, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2017, volume=17, issue=2, pageStart=71, pageEnd=79, url=null, language=null, rfNumber=[15], rfOrder=21, authorNames=Tanyildizi E, Demir G, journalName=Advances in Electrical and Computer Engineering, refType=null, unstructuredReference=Tanyildizi E, Demir G. Golden sine algorithm: a novel math inspired algorithm[J]. Advances in Electrical and Computer Engineering, 2017, 17(2): 71-79., articleTitle=Golden sine algorithm: a novel math inspired algorithm, refAbstract=null), Reference(id=1233422560977801400, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=6, pageStart=2428, pageEnd=2437, url=null, language=null, rfNumber=[16], rfOrder=22, authorNames=宋江涛, 崔双喜, 刘洪广, journalName=科学技术与工程, refType=null, unstructuredReference=宋江涛, 崔双喜, 刘洪广. 基于二次分解NGO-VMD残差项与长短时记忆神经网络的超短期风功率预测[J]. 科学技术与工程, 2023, 23(6): 2428-2437., articleTitle=基于二次分解NGO-VMD残差项与长短时记忆神经网络的超短期风功率预测, refAbstract=null), Reference(id=1233422561099436227, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2019, volume=23, issue=6, pageStart=2428, pageEnd=2437, url=null, language=null, rfNumber=[16], rfOrder=23, authorNames=Song Jiangtao, Cui Shuangxi, Liu Hongguang, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Song Jiangtao, Cui Shuangxi, Liu Hongguang. Ultra-short term wind power prediction based on quadratic decomposition NGO-VMD residual term and short term memory neural network[J]. Science Technology and Engineering, 2019, 23(6): 2428-2437., articleTitle=Ultra-short term wind power prediction based on quadratic decomposition NGO-VMD residual term and short term memory neural network, refAbstract=null), Reference(id=1233422561242042574, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2013, volume=41, issue=2, pageStart=144, pageEnd=149, url=null, language=null, rfNumber=[17], rfOrder=24, authorNames=周洪煜, 曾济贫, 王照阳, journalName=电力系统保护与控制, refType=null, unstructuredReference=周洪煜, 曾济贫, 王照阳, 等. 基于混沌DNA遗传算法与PSO组合优化的RNN短期风电功率预测[J]. 电力系统保护与控制, 2013, 41(2): 144-149., articleTitle=基于混沌DNA遗传算法与PSO组合优化的RNN短期风电功率预测, refAbstract=null), Reference(id=1233422561384648917, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, doi=null, pmid=null, pmcid=null, year=2013, volume=41, issue=2, pageStart=144, pageEnd=149, url=null, language=null, rfNumber=[17], rfOrder=25, authorNames=Zhou Hongyu, Zeng Jipin, Wang Zhaoyang, journalName=Power System Protection and Control, refType=null, unstructuredReference=Zhou Hongyu, Zeng Jipin, Wang Zhaoyang, et al. RNN short-term wind power prediction based on chaotic DNA Genetic algorithm and PSO combination optimization[J]. Power System Protection and Control, 2013, 41(2): 144-149., articleTitle=RNN short-term wind power prediction based on chaotic DNA Genetic algorithm and PSO combination optimization, refAbstract=null)], funds=[Fund(id=1233422555999163283, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, awardId=LH2019E001, language=CN, fundingSource=黑龙江省自然科学基金(LH2019E001), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1233422547761549649, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, xref=1, ext=[AuthorCompanyExt(id=1233422547765743954, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547761549649, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin 150040, China), AuthorCompanyExt(id=1233422547774132563, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547761549649, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东北林业大学机电工程学院, 哈尔滨 150040)]), AuthorCompany(id=1233422547891573086, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, xref=2, ext=[AuthorCompanyExt(id=1233422547908350304, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547891573086, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 703 Research Institute, China State Shipbuilding Corporation, Harbin 150783, China), AuthorCompanyExt(id=1233422547912544609, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, companyId=1233422547891573086, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 中国船舶集团有限公司第七O三研究所, 哈尔滨 150783)])], figs=[ArticleFig(id=1233422552018768504, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Fig.1, caption=Specific process of short-term prediction of wind power, figureFileSmall=R10cdGpeqYFzuwOUdFzsag==, figureFileBig=LWCyC0S91kmCO/+ceUYWZQ==, tableContent=null), ArticleFig(id=1233422552119431814, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=图1, caption=风电功率短期预测的具体流程, figureFileSmall=R10cdGpeqYFzuwOUdFzsag==, figureFileBig=LWCyC0S91kmCO/+ceUYWZQ==, tableContent=null), ArticleFig(id=1233422552228483733, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Fig.2, caption=Internal structure of LSTM model, figureFileSmall=XMNDsbMi46U0pXchEUQ0Pg==, figureFileBig=kJYAk4hR5kQOE6SiQPvI6Q==, tableContent=null), ArticleFig(id=1233422552329147040, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=图2, caption=LSTM 模型内部结构

C(t)和C(t-1)分别为t时刻更新后的新细胞状态与上一时刻的细胞状态;H(t)和H(t-1)分别为t时刻隐藏状态与上一时刻的隐藏状态

, figureFileSmall=XMNDsbMi46U0pXchEUQ0Pg==, figureFileBig=kJYAk4hR5kQOE6SiQPvI6Q==, tableContent=null), ArticleFig(id=1233422552421421738, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Fig.3, caption=Attention mechanism structure diagram, figureFileSmall=Hj880+WVlWtCID1QAPgmHw==, figureFileBig=qnpIjq3dcGjpc9oGKBKEug==, tableContent=null), ArticleFig(id=1233422552605971134, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=图3, caption=Attention机制结构图, figureFileSmall=Hj880+WVlWtCID1QAPgmHw==, figureFileBig=qnpIjq3dcGjpc9oGKBKEug==, tableContent=null), ArticleFig(id=1233422552715023050, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Fig.4, caption=CEEMDAN and VMD decomposition results, figureFileSmall=bxho/Y1lHNc449zuIgsLdg==, figureFileBig=+VdhOPRtXYaU58fopteLbw==, tableContent=null), ArticleFig(id=1233422552903766742, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=图4, caption=序列二次分解结果, figureFileSmall=bxho/Y1lHNc449zuIgsLdg==, figureFileBig=+VdhOPRtXYaU58fopteLbw==, tableContent=null), ArticleFig(id=1233422553063150312, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Fig.5, caption=Performance comparison of various optimization algorithms, figureFileSmall=WoRN4C8JX/wenFyBfyh/ig==, figureFileBig=xTJQXsENkKU1u6/sFouWag==, tableContent=null), ArticleFig(id=1233422553235116789, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=图5, caption=多种优化算法性能对比, figureFileSmall=WoRN4C8JX/wenFyBfyh/ig==, figureFileBig=xTJQXsENkKU1u6/sFouWag==, tableContent=null), ArticleFig(id=1233422554677957384, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Fig.6, caption=Comparison of prediction effects of some models, figureFileSmall=hptTo/+guqQ4r9+aGm8BnA==, figureFileBig=OcgMGKSam49hNAI0PgKtOg==, tableContent=null), ArticleFig(id=1233422554862506775, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=图6, caption=部分模型预测效果对比, figureFileSmall=hptTo/+guqQ4r9+aGm8BnA==, figureFileBig=OcgMGKSam49hNAI0PgKtOg==, tableContent=null), ArticleFig(id=1233422555042861858, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Table 1, caption=

Parameter optimization interval

, figureFileSmall=null, figureFileBig=null, tableContent=
超参数 寻优区间
第一层神经元数量 [32,128]
第二层神经元数量 [32,128]
学习率 [0.001,0.01]
样本批量 [20,60]
), ArticleFig(id=1233422555214828343, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=表1, caption=

参数优化区间

, figureFileSmall=null, figureFileBig=null, tableContent=
超参数 寻优区间
第一层神经元数量 [32,128]
第二层神经元数量 [32,128]
学习率 [0.001,0.01]
样本批量 [20,60]
), ArticleFig(id=1233422555328074563, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Table 2, caption=

Precision comparison of model prediction results

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE MMAPE/
%
R2/%
CEEMDAN-VMD-IDBO-BiLSTM-ATTENTION 5.173 5.44 99.21
CEEMDAN-IDBO-BiLSTM-ATTENTION 7.921 8.13 98.20
VMD-IDBO-BiLSTM-ATTENTION 10.351 8.78 96.41
), ArticleFig(id=1233422555470680915, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=表2, caption=

模型预测结果精度对比

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE MMAPE/
%
R2/%
CEEMDAN-VMD-IDBO-BiLSTM-ATTENTION 5.173 5.44 99.21
CEEMDAN-IDBO-BiLSTM-ATTENTION 7.921 8.13 98.20
VMD-IDBO-BiLSTM-ATTENTION 10.351 8.78 96.41
), ArticleFig(id=1233422555630064488, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=EN, label=Table 3, caption=

Comparison of accuracy of prediction results of some models

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE MMAPE/% R2/%
SVM 35.454 20.62 88.7
TCN-BiLSTM 12.129 8.17 96.56
CNN-BiLSTM 10.152 8.89 97.59
DBO-BiLSTM 11.317 9.04 96.64
本文模型 5.173 5.44 99.21
), ArticleFig(id=1233422555722339183, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156264266425553548, language=CN, label=表3, caption=

部分模型预测结果精度对比

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE MMAPE/% R2/%
SVM 35.454 20.62 88.7
TCN-BiLSTM 12.129 8.17 96.56
CNN-BiLSTM 10.152 8.89 97.59
DBO-BiLSTM 11.317 9.04 96.64
本文模型 5.173 5.44 99.21
)], attaches=null, journal=Journal(id=1146119176004939786, delFlag=0, nameCn=科学技术与工程, nameEn=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, issn=1671-1815, eissn=, cn=11-4688/T, coden=null, periodic=4, language=CN, oaType=是, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=UKU/O7GSka5polgCTkbIIw==, journalPrice=null, startedYear=null, abbrevIsoEn=Sci Technol Eng, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1754445529766, createdBy=null, updatedBy=13701087609, firstLetterCn=S, firstLetterEn=S, subjectCode=Natural Sciences, subjectName=自然科学, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=UKU/O7GSka5polgCTkbIIw==, picEn=5hwlULoNwcbj3xUmVi9MAQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1159791870395564357, language=CN, name=科学技术与工程, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.stae.com.cn/jsygc/home, createdTime=1754445529793, updatedTime=1754445529793, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.stae.com.cn/jsygc/site/menus/20090429150146001, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1159791870441701702, language=EN, name=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.stae.com.cn/jsygc/home, createdTime=1754445529804, updatedTime=1754445529804, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1146123166801305609, websiteList=[Website(id=1148243202391400884, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/kxjsygc/CN, language=CN, createTime=1751692112777, createBy=18614031015, updateTime=1753520965431, updateBy=18614031015, name=科学技术与工程-中文站点, tplId=1146099689490845704, title=科学技术与工程, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148622798802673703, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=articleTextType, value=kx, createTime=1751782615614, updateTime=1751782615614, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798781702180, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=banner, value=null, createTime=1751782615609, updateTime=1751782615609, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798769119267, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1751782615606, updateTime=1751782615606, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798794285094, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751782615612, updateTime=1751782615612, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798790090789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751782615611, updateTime=1751782615611, creator=18614031015, updator=18614031015)]), Website(id=1155914124811976731, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/kxjsygc/EN, language=EN, createTime=1753521003206, createBy=18614031015, updateTime=1753521003206, updateBy=18614031015, name=科学技术与工程-英文站点, tplId=1146101810881728533, title=Science Technology and Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155914371227308235, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=articleTextType, value=kx, createTime=1753521061952, updateTime=1753521061952, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371210531016, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=banner, value=null, createTime=1753521061947, updateTime=1753521061947, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371202142407, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1753521061945, updateTime=1753521061945, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371223113930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753521061950, updateTime=1753521061950, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371218919625, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753521061949, updateTime=1753521061949, creator=18614031015, updator=18614031015)])], journalTitle=科学技术与工程, weixinUrl=null, journalUrl=null, iacademicId=null, status=0, seqNo=null, journalTitleEn=Science Technology and Engineering, journalPhotoCn=UKU/O7GSka5polgCTkbIIw==, journalPhotoEn=5hwlULoNwcbj3xUmVi9MAQ==, journalFirstLetter=S, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=null, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2402025, detailUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/10.12404/j.issn.1671-1815.2402025, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/PDF/10.12404/j.issn.1671-1815.2402025, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/PDF/10.12404/j.issn.1671-1815.2402025, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于误差补偿及IDBO-BiLSTM的风电功率短期预测
收藏切换
PDF下载
魏振宇 1 , 姜雪松 1, * , 杨立发 2
科学技术与工程 | 论文·自动化技术、计算机技术 2025,25(6): 2397-2405
收起
收藏切换
科学技术与工程 | 论文·自动化技术、计算机技术 2025, 25(6): 2397-2405
基于误差补偿及IDBO-BiLSTM的风电功率短期预测
全屏
魏振宇1 , 姜雪松1, * , 杨立发2
作者信息
  • 1 东北林业大学机电工程学院, 哈尔滨 150040
  • 2 中国船舶集团有限公司第七O三研究所, 哈尔滨 150783
  • 魏振宇(2000—),男,汉族,山东济宁人,硕士研究生。研究方向:短期风电功率预测。E-mail:

通讯作者:

* 姜雪松(1979—),男,汉族,黑龙江佳木斯人,博士,副教授。研究方向:工业工程与管理、智能制造工艺与装备。E-mail:
Short Term Prediction of Wind Power Based on Error Compensation and IDBO-BiLSTM
Zhen-yu WEI1 , Xue-song JIANG1, * , Li-fa YANG2
Affiliations
  • 1 College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin 150040, China
  • 2 703 Research Institute, China State Shipbuilding Corporation, Harbin 150783, China
出版时间: 2025-02-28 doi: 10.12404/j.issn.1671-1815.2402025
文章导航
收藏切换

针对风电出力稳定性差、随机性强而导致的模型精度差的问题。提出了一种基于二次分解误差补偿的风电功率短期预测模型。首先建立双向长短期记忆(bidirectional long short-term memory, BiLSTM)预测模型对风电功率进行预测并输出预测误差。其次,采用了一种利用混沌映射初始化种群、引入黄金正弦策略更新滚球蜣螂位置,并添加动态自适应性权重系数来更新偷窃蜣螂的位置的改进蜣螂优化算法(improved dung beetle optimizer, IDBO)对预测模型参数寻优,防止网络陷入局部最优解,自适应搜寻最优参数组合。然后,采用分解-重构-分解的策略,利用自适应噪声的完全集合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise, CEEMDAN)进行首次分解,并且引入样本熵(sample entropy, SE)与K均值(K-means)将序列按频率进行重构并通过变分模态分解(variational mode decomposition, VMD)将高频误差序列分解成不同频段的误差序列,提高后续模型的预测效率及预测精度。最后,将各分量输入误差补偿模型进行预测并引入Attention机制学习不同时间步的特征关系,并给与不同权重值,加强对关键信息的注意力。通过新疆达坂城风电场实测数据验证了所提模型预测精度高,具有显著优势。

风电功率短期预测  /  双向长短期记忆网络  /  改进蜣螂优化算法  /  完全集合经验模态分解  /  变分模态分解

Aiming at the problem of poor model accuracy caused by poor stability and strong randomness of wind power output. A short-term prediction model of wind power based on quadratic decomposition error compensation was proposed. Firstly, BiLSTM (bidirectional long short-term memory) prediction model is established to predict wind power and output prediction errors. Secondly, an IDBO (improved dung beetle optimizer) algorithm was used to initialize the population by using chaotic mapping, update the position of rolling dung beetles by introducing golden sine strategy, and update the position of thieving dung beetles by adding dynamic adaptive weight coefficient to optimize the parameters of the prediction model. Prevent the network from falling into the local optimal solution, and adaptively search the optimal parameter combination. Then, using the decomposition-reconstruction-decomposition strategy, CEEMDAN (complete ensemble empirical mode decomposition with adaptive noise) was used for the first decomposition. In addition, SE(sample entropy) and K-means are introduced to reconstruct the sequence according to frequency, and the high-frequency error sequence was decomposed into error sequences of different frequency bands by VMD(variational mode decomposition). Improve the prediction efficiency and accuracy of subsequent models. Finally, the input error compensation model of each component was used to predict and the Attention mechanism was introduced to learn the feature relationship of different time steps and give different weight values to enhance the attention to key information. Through the measured data of a wind farm in Xinjiang, the prediction accuracy of the proposed model is proved to be high and has significant advantages.

wind power short-term forecast  /  bidirectional long short-term memory network  /  improved dung beetle optimization algorithm  /  complete ensemble empirical mode decomposition  /  variational mode decomposition
魏振宇, 姜雪松, 杨立发. 基于误差补偿及IDBO-BiLSTM的风电功率短期预测. 科学技术与工程, 2025 , 25 (6) : 2397 -2405 . DOI: 10.12404/j.issn.1671-1815.2402025
Zhen-yu WEI, Xue-song JIANG, Li-fa YANG. Short Term Prediction of Wind Power Based on Error Compensation and IDBO-BiLSTM[J]. Science Technology and Engineering, 2025 , 25 (6) : 2397 -2405 . DOI: 10.12404/j.issn.1671-1815.2402025
风能是目前应用最广泛的可再生能源之一,其在能源系统中的比重不断提高。然而实际生活中风力发电过程却因为风力的随机性而带来很大的不确定性[1],并且风电功率的大小也受风速外的气压、温度、湿度、风向等因素的影响,这加剧了风电功率预测的不稳定性。风力发电预测稳定性差、随机性强的特性对电网的调度和储能能力形成巨大考验[2],因此,迫切地需要提升风电功率预测精度,以最大程度地减轻风电波动所带来的挑战。
早期的风电功率预测主要采用人工神经网络、支持向量机[3-4]和极端梯度增强算法(extreme gradient boosting, xgboost)等,但这些模型无法感受与记忆输入时间有关的时序特征,导致预测精度有限。随着深度学习发展,长短期记忆神经网络 (long short-term memory, LSTM),时序卷积网络(convolutional neural networks, CNN),自注意力模型(transformer)等深度神经网络被广泛用于风电功率的预测问题。然而深度学习模型经常面临无法有效地关注时间序列的关键信息的问题。文献[5]提出一种基于Attention-GRU的数值天气预报风速修正和Stacking多算法融合预测模型,考虑多种气象因素,并通过Attention机制分配关键因素权重,提高计算效率,有效地提高了预测精度。文献[6]通过设计多层语义结合的注意力机制加强对特征向量的编码能力,并使用分位数回归和核密度估计方法处理模型输出。文献[7-9]分别通过灰狼优化算法、遗传算法、改进粒子群优化算法对预测模型进行改进,优化了参数的选取,提高了预测精度。上述模型仅通过组合深度学习模型进行风电功率预测,挖掘特征信息,提高了预测精度,并且结合了优化算法寻找模型最优超参数,提高了运算效率。但上述模型却没有考虑到风电功率预测稳定性差、随机性强的问题。文献[10]为了降低时间序列的非稳定特征,将原始数据输入变分模态分解(variational mode decomposition, VMD)进行分解,然后使用权值共享门控循环单元(weight sharing gate recurrent unit, WSGRU)对所有子分量进行快速建模预测。同样的,文献[11]提出了算术优化算法(improved arithmetic optimization algorithm, IAOA)-变分模态分解VMD-LSTM预测模型。利用IAOA对VMD的关键分解参数进行优化,有效提高了预测效率和预测精度。文献[12]通过完全经验模态分解(complete ensemble empirical mode decomposition with adaptive noise, CEEMDAN)排列熵(permutation entropy, PE)、小波包分解(wavelet packet decomposition, WPD)对信号进行处理,然后输入LSTM神经网络,并建立多目标优化损失函数,综合提升模型的预测性能。
利用深度学习的方法预测风电功率,不可避免地会造成预测误差[13]。同时如果能从预测误差中建立起误差与风电功率的关联网络,挖掘其内在联系,便能更进一步地提高风电功率预测的精度。然而上述研究只关注了对深度学习模型的改进和对气象特征的提取,却忽视了预测误差所带来的潜在价值。
综上所述,现提出一种基于二次分解误差补偿的风电功率短期预测模型。首先利用IDBO-BILSTM模型对风电功率进行初步预测,随后输出预测功率与实际功率之间的误差值作为误差补偿模型的训练数据。最后,叠加误差补偿数据和初步预测功率数据作为最终预测功率,以提升预测精度。考虑到风力发电具有稳定性差、随机性强的特征,简单地输入误差信息进行预测对精度的提升效果有限,因此在误差补偿之前加入自适应白噪声完备集成经验模态分解法和变分模态分解降低序列的不稳定性,并在误差补偿模型中引入注意力机制加强对关键信息的注意力。基于此,以新疆达坂城风电场功率实测数据为研究对象,构建一种基于二次分解误差补偿的风电功率短期预测模型,并进行仿真验证所提模型预测可行性与有效性。
对于一个特定的深度学习网络,其内部的运算原理、随机数种子、各个模块结构、参数权重等皆是定值。使用同样的样本数据多次预测,其结果始终相同。同样的,通过相同模型进行预测,所造成的预测误差也必然遵守模型的运算规则。因此,将初步预测功率所得到的预测误差再次投入深度学习网络进行训练,使模型建立起误差与用于功率预测的气象特征、功率数据与误差之间的结构关系网络,实现对误差的精准预测,从而提升风功率的预测精度。
风功率由于其强随机性和非平稳性,导致其预测精度一直较低,为提升功率预测精度,本文研究提出一种基于二次分解误差补偿的风电功率短期预测模型。具体流程如下。
(1)首先将输入预测模型的数据进行预处理,利用IDBO-LSTM预测模型对风功率进行快速的初步预测,并对比实际功率,得到风功率的初步预测误差。
(2)利用CEEMDAN将初步预测误差分解为多组模态分量,并计算各分量的样本熵,通过K-means聚类算法将分量进行重构。
(3)选出重构后的高频分量,利用VMD将高频预测误差序列分解为一组子分量。
(4)将各分量、气象特征、风功率输入BiLSTM-ATTENTION 预测模型。得到各分量风功率预测误差值
(5)将误差值与风功率初步预测功率值进行整合,作为最终风功率预测值。
(6)对预测结果进行评价,分析模型性能
风功率预测流程图如图1所示。
BiLSTM神经网络结构是由2个相互独立的LSTM神经网络组成,它同时考虑了过去和未来的信息,可以通过向前向和向后向两个方向传播。这使得BiLSTM网络不仅能捕捉过去的信息,还能预知未来的信息,通过双向处理序列信息加强了捕捉长期依赖关系的能力,从而在处理时间序列问题上有更好的表现。 LSTM模型内部结构如图2 所示。
LSTM模型单元的完整传播公式为
$f=\sigma \left({W}_{\mathrm{f}}\right[{h}_{t-1},{x}_{t}]+{b}_{\mathrm{f}})$
${i}_{t}=\sigma \left({W}_{\mathrm{i}}\right[{h}_{t-1},{x}_{t}]+{b}_{\mathrm{i}})$
${\stackrel{·}{C}}_{t}=\mathrm{R}\mathrm{e}\mathrm{l}\mathrm{u}\left({W}_{\mathrm{c}}\right[{h}_{t-1},{x}_{t}]+{b}_{\mathrm{c}})$
${C}_{t}={f}_{t}{C}_{\mathrm{t}-1}+{i}_{t}{C}_{t}$
${O}_{t}=\sigma \left({W}_{\mathrm{o}}\right[{h}_{t-1},{x}_{t}]+{b}_{\mathrm{o}})$
${h}_{t}={O}_{t}\mathrm{R}\mathrm{e}\mathrm{l}\mathrm{u}\left({C}_{t}\right)$
式中:ft为遗忘门;it为输入门;ot 为输出门;σ和Relu分别为Sigmoid函数和Relu激活函数;WfWiWcWo分别为各式的权重矩阵;bfbibcbo分别为各式的偏置向量;xt为当前模型输入信息;${\stackrel{·}{C}}_{t}$和Ct分别为隐藏层生成t时刻记忆单元的临时状态和新细胞状态;ht隐藏层最终输出值。
LSTM网络相比传统神经网络,能够捕捉数据中的长期依赖性、通过门控机制解决可能会遇到的梯度消失与梯度爆炸问题,同时LSTM特殊的细胞状态能够选择性的动态处理信息。BiLSTM网络增加了对前后信息处理的能力,能充分挖掘数据中的有用信息,在预测任务中具有出色的表现。
Attention机制能够在计算能力有限的情况下将资源分配给优先级更高的任务,从而提升模型的精度。在本文研究中,将注意力机制引入LSTM模型的输出上,使用注意力层来计算输入序列中每个时间步的权重,然后将这些权重应用于LSTM输出序列中的每个时间步来产生一个加权的表示。Attention机制的引入使得模型在信息处理过程中优先关注权重占比较大的信息,提高了信息处理效率,同时提升了结果的准确性。Attention机制结构如图3所示。
Attention机制计算公式为
${S}_{t}=F(Q,{K}_{t})$
${a}_{i}=\mathrm{s}\mathrm{o}\mathrm{f}\mathrm{t}\mathrm{m}\mathrm{a}\mathrm{x}\left({S}_{i}\right)=\frac{{\mathrm{e}}^{{S}_{i}}}{\stackrel{N}{\sum _{j=1}}}{\mathrm{e}}^{{S}_{j}}$
$\mathrm{A}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\left[\right(K,V\left)Q\right]=\stackrel{n}{\sum _{i=1}}{a}_{i}{h}_{i}$
式中:Q为查询;K为键;V为值;N为数据长度;Si为每个隐藏层的得分;ai为权重系数;hi为BiLSTM隐藏层输入值;Attention为最终所求的注意力值。
蜣螂优化算法是一种新型的仿生智能优化算法[14],通过模仿蜣螂滚球、跳舞、觅食、繁殖和偷窃等行为,可有效解决复杂的优化搜索问题。蜣螂优化算法相较于其他仿生算法,具有更强的全局搜索能力。但同时,DBO算法也会与其他寻优算法一样容易陷入局部最优解,而无法做到全局搜索。为了减少搜索的复杂性、加快算法收敛速度,同时提升全局搜索能力和跳出局部最优的能力,提出一种IDBO算法。首先引入Chebyshev映射,帮助改善种群的分布。同时,使用黄金正弦策略对滚球蜣螂进行位置更新。最后,增加动态权重系数调整窃取蜣螂位置。IDBO算法加快了收敛速度、增强了算法跳出局部最优能力,并且平衡全局搜索能力和局部开发的能力。
在基础DBO优化算法中采用的是随机种群初始化,会造成种群多样性不足、种群分布不均的情况,针对这种情况,本文模型采用Chebyshev混沌映射的种群初始化的方法。通过Chebyshev混沌映射使得初始种群分布得更加均衡,从而提高算法全局搜索效果。Chebyshev 混沌映射的公式为
${x}_{n+1}=\mathrm{c}\mathrm{o}\mathrm{s}\left(k\mathrm{a}\mathrm{r}\mathrm{c}\mathrm{c}\mathrm{o}\mathrm{s}{x}_{n}\right), x\in [-\mathrm{1,1}]$
式(10)中:k为阶次。
黄金正弦算法(Golden-SA)是一种新型的元启发式算法[15]。该算法利用正弦函数与单位圆的关系,能够遍历正弦函数上的所有点,应用在蜣螂滚球行为中,能够有效提高DBO算法的全局搜索能力。同时,在蜣螂滚球更新位置过程中引入黄金分割系数,充分搜索产生优质解的局部区域,从而有助于DBO算法跳出局部最优,快速收敛。位置更新公式为
$\left\{\begin{array}{l}{x}_{1}=-\mathrm{\pi }+(1-\tau )2\mathrm{\pi }\\ {x}_{2}=-\mathrm{\pi }+\tau 2\mathrm{\pi }\\ {X}_{i}(t+1)={X}_{i}\left(t\right)\left|\mathrm{s}\mathrm{i}\mathrm{n}{R}_{1}\right|+\\   {R}_{2}\mathrm{s}\mathrm{i}\mathrm{n}{R}_{1}|{x}_{1}{{X}^{\mathrm{b}}}_{t}-{x}_{2}{X}_{t}|\end{array}\right.$
式(11)中:τ为黄金分割系数;Xi为第i只滚球蜣螂的位置信息;t为迭代次数;Xb为最优位置;R1为一个介于0~2π的随机数;R2为一个介于0~π的随机数。
DBO算法中部分蜣螂会偷窃其他蜣螂的粪球,在这个过程中偷窃蜣螂会向其他蜣螂传递自身所获得的信息,方便其他蜣螂更快地收敛到更优解附近。通过引入自适应动态权重系数改进其位置,从而使得算法更有可能发现全局最优解,并且蜣螂之间的信息传递与共享也使得整个群体更快地搜索全局,提高算法的全局搜索能力。本文模型窃取蜣螂位置更新公式为
$\left\{\begin{array}{l}{X}_{i}(t+1)=\\   \frac{{w}_{1}{X}^{b}+{w}_{2}\mathrm{g}[\left|{\mathrm{x}}_{\mathrm{i}}\left(\mathrm{t}\right)-{\mathrm{X}}^{\mathrm{*}}\right|+\left|{\mathrm{x}}_{\mathrm{i}}\left(\mathrm{t}\right)-{\mathrm{X}}^{\mathrm{b}}\right|]}{2}\\ {w}_{1}=2{\mathrm{e}}^{-(2t/T)}\\ {w}_{2}=2{\mathrm{e}}^{-\left[2\right(T-t)/T]}\end{array}\right.$
式(12)中:X(t)为第i只偷窃蜣螂在第t次迭代时的位置信息;g表示服从正态分布的大小为1×D的随机向量;T为最大迭代次数。
固定的位置更新权重系数并不利于找寻全局最优解。在搜索初期,w1较大,蜣螂会在最优蜣螂附近进行搜索。随着迭代的不断进行,这时w2的影响力逐渐大于w1,蜣螂这时会跳出局部最优,一定程度上解决了算法局部开发与全局探索不平衡的问题。
神经网络预测模型的精度主要受学习率、隐藏层的神经元数量、迭代次数等参数的影响。为进一步提升预测模型的精度,克服神经网络参数难以调整的问题,选用IDBO为预测模型进行参数寻优。BiLSTM激活函数为Relu, Dropout层选取0.2,迭代次数为 70次。并且采用 LearningRateScheduler 调度器动态调整学习率。优化目标包括隐藏层的神经元数量、样本批量以及预测模型学习率。
IDBO优化BiLSTM-ATTENTION模型的具体优化过程流程如下:①通过混沌映射初始化蜣螂种群;②初始BLSTM参数并设置参数寻优区间;③计算当前所有蜣螂的适应度值;④更新所有蜣螂的位置;⑤每次迭代后重新计算蜣螂的适应度,引入动态学习因子,随着迭代的进行,自适应调整蜣螂位置;⑥判断是否满足最大迭代,若满足,则输出神经网络超参数,否则返回3继续迭代。
改进蜣螂算法寻优区间如表1所示。
CEEMDAN是EEMD的改进模型,通过向经验模态分解EMD(empirical mode decomposition),分解后产生的IMF本征模态函数(intrinsic mode function,IMF)分量中加入自适应噪声使分解结果更加准确,有效地减少了重构误差,提高了分解效率。CEEMDAN 算法的操作步骤如下。
${y}_{m}\left(t\right)=x\left(t\right)+\alpha {\delta }_{m}\left(t\right)$
$\mathrm{I}\mathrm{M}{\mathrm{F}}_{1}\left(t\right)=\frac{1}{M}\stackrel{M}{\sum _{m=1}}\mathrm{I}\mathrm{M}{\mathrm{F}}_{1}^{m}\left(t\right)$
${r}_{1}\left(t\right)=y\left(t\right)-\mathrm{I}\mathrm{M}{\mathrm{F}}_{1}\left(t\right)$
$\mathrm{I}\mathrm{M}{\mathrm{F}}_{n}\left(t\right)=\frac{1}{K}\stackrel{K}{\sum _{k=1}}{E}_{n-1}\left\{{r}_{n-1}\right(t)+{\alpha }_{n-1}{E}_{n-1}[{\delta }_{k}\left(t\right)\left]\right\}$
${r}_{n}\left(t\right)={r}_{n-1}\left(t\right)-\mathrm{I}\mathrm{M}{\mathrm{F}}_{n}\left(t\right)$
$x\left(t\right)=\stackrel{N}{\sum _{n=1}}\mathrm{I}\mathrm{M}{\mathrm{F}}_{n}+{r}_{N}\left(t\right)$
式中:${\delta }_{m}\left(t\right)$斯白噪声序列;ym(t)m次得分解子序列;x(t)为原始信号;α为白噪声权重系数。
m次高斯白噪声序列添加到初步预测误差序列中,得到待分解的m次子序列,使用EMD分解得分解子序列,对得到m个第一分量求平均值得到第一个模态分量IMF1及第一阶残差分量r1(t)。随后,通过上述公式持续分解,依次计算。直到余量误差无法继续分解停止。
为降低初步预测误差的非平稳行,将初步预测的功率误差进行信号分解,同时为了得到更精细的模态分量,将初步预测误差进行CEEMDAN-VMD二次分解,这虽能一定程度上提升预测精度,但这将导致模态分量过多,极大地影响预测效率。为了提升风电功率预测精度的同时有效地减少预测时间,提高计算效率,本文研究提出一种先重构在分解的方法。通过计算CEEMDAN分解的初步误差序列各模态分量的样本熵,利用样本熵评估各分量复杂度,然后按照各自分量的复杂度通过K-means聚类算法将各分量重构为高、中、低三种频率的模态分量。然后在对高频初步预测误差分量进行VMD分解,通过重构分量可以得到更加稳定且精细的子序列,同时也大大缩小了误差补偿预测模型的运算时间,极大地增加了预测效率和预测精度。
K-means算法由于它具有操作简单、收敛速度快的优点被广泛用于聚类运算[16],K-means算法的基本过程如下: ①随机选取K个点作为初始聚类的簇心;②分别计算各聚点到簇心的欧式距离,将离该点距离较近的点划分为同类;③更新簇心再次迭代;④簇心不再发生明显的变化时停止。
VMD是一种自适应、非递归的信号分解方法,能够将原始信号分解为一系列具有稀疏性质的平稳模态分量。它可以自适应性的匹配各分量的最佳中心频率和有限带宽,求得变分问题最优解。VMD分解过程如下。
变分模型约束表达式为
$\left\{\begin{array}{l}\underset{\left\{uk\right\}\left\{\omega k\right\}}{\mathrm{m}\mathrm{i}\mathrm{n}}\left\{\stackrel{K}{\sum _{k=1}}=\partial \left(t\right)\left\{\right[\delta \left(t\right)+\frac{\mathrm{j}}{\mathrm{\pi }t}\left]{u}_{k}\right(t\left)\right\}{\mathrm{e}}^{-\mathrm{j}{\omega }_{k}t}{=}_{2}^{2}\right.\\ \mathrm{s}.\mathrm{t}.\sum _{K}{u}_{k}=f\left(t\right)\end{array}\right.$
式(19)中:uk为各模态分量,k=1,2,…,K;ωk为各模态分量对应的中心频率;$\delta \left(t\right)$为狄拉克函数。
为求解上述约束表达式,引入二次惩罚函数和增广拉格朗日函数,公式为
$\begin{array}{l}L\left(\right\{{u}_{k}\},\{{\omega }_{k}\},\lambda )=\alpha \stackrel{K}{\sum _{k=1}}=\partial \left(t\right)\left\{\right[\delta \left(t\right)+\\   \frac{\mathrm{j}}{\mathrm{\pi }t}\left]{u}_{k}\right(t\left)\right\}{\mathrm{e}}^{-\mathrm{j}{\omega }_{k}t}={\mathrm{ }}_{2}^{2}+=f\left(t\right)-\stackrel{K}{\sum _{k=1}}{u}_{k}{=}_{2}^{2}+\\   <\lambda \left(t\right),f\left(t\right)-\stackrel{K}{\sum _{k=1}}{u}_{k}>\end{array}$
式(20)中:α为二次惩罚参数;λ为拉格朗日乘法算子。
使用交替方向乘子法计算各模态分量uk以及各模态分量对应的中心频率ωk,更新公式为
${\dot{\mathrm{u}}}_{k}^{n+1}\left(\omega \right)=\frac{\hat{f}\left(\omega \right)-\sum _{i\ne k}{\dot{\mathrm{u}}}_{i}\left(\omega \right)+\frac{\dot{\mathrm{\lambda }}\left(\omega \right)}{2}}{1+2\alpha (\omega -{\omega }_{k}{)}^{2}}$
${\omega }_{k}^{n+1}=\frac{{\int }_{0}^{\infty }\omega \left|{\dot{\mathrm{u}}}_{k}^{m+1}{\left(\omega \right)}^{2}\right|\mathrm{d}\omega }{{\int }_{0}^{\infty }\left|{\dot{\mathrm{u}}}_{k}^{m+1}{\left(\omega \right)}^{2}\right|\mathrm{d}\omega }$
式中:n为迭代次数;${\dot{\mathrm{u}}}_{k}^{n+1}\left(\omega \right)、\hat{f}\left(\omega \right)、{\dot{\mathrm{u}}}_{i}\left(\omega \right)、\dot{\mathrm{\lambda }}\left(\omega \right)$分别为${u}_{k}^{n+1}\left(t\right)、\hat{f}\left(t\right)、{\dot{\mathrm{u}}}_{i}\left(t\right)、\hat{\lambda }\left(t\right)$的傅里叶变换。
为验证本文模型的有效性,以新疆达坂城风电场实际监测数据集进行仿真验证,该数据集包括2019年1月1日—2月7日的风力发电功率数据。数据中的特征包括风电场风向、风速、实时气温、气压、湿度、时间,共6种特征。以15 min采样1次,每天采样96条,38 d共3 648条数据。为避免过程产生过拟合现象,本文研究采取7∶3的比例选取前25 d的2 400条数据作为训练数据,验证集取10 d共计960条数据,最后3 d共计288条数据作为预测集。实验结果将取10次实验结果的算术平均结果作为最终结果。
采用均方根误差(root mean square error, RMSE)、决定系数(R2)、以及基于绝对平均误差MAPE改进的指标(mean of actual values mean Absolute percentage error, MMAPE)[17],3个指标作为评价指标。其公式分别为
${E}_{\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E}}=\sqrt{\frac{1}{n}\stackrel{n}{\sum _{i=1}}({y}_{t}-{\dot{\mathrm{y}}}_{t}{)}^{2}}$
${R}^{2}=1-\frac{\stackrel{N}{\sum _{i=1}}({y}_{t}-{\dot{\mathrm{y}}}_{t}{)}^{2}}{\stackrel{N}{\sum _{i=1}}({\stackrel{-}{\mathrm{y}}}_{t}-{y}_{t}{)}^{2}}$
$\left\{\begin{array}{l}{E}_{\mathrm{M}\mathrm{M}\mathrm{A}\mathrm{P}\mathrm{E}}=\frac{1}{n}\stackrel{n}{\sum _{i=1}}\left|\frac{{y}_{t}-{\dot{\mathrm{y}}}_{t}}{{y}_{\mathrm{a}\mathrm{v}\mathrm{e}}}\right|\times 100\mathrm{\%}\\ {y}_{\mathrm{a}\mathrm{v}\mathrm{e}}=\frac{1}{n}\stackrel{n}{\sum _{i=1}}{y}_{t}\times 100\mathrm{\%}\end{array}\right.$
式中: yt为实测值;${\dot{\mathrm{y}}}_{t}$为预测值;ERMSEEMMAPE分别为预测结果的均方根误差和改进绝对平均误差;yave为实测值的均值。
通过CEEMDAN分解将初步预测误差序列分解,并计算分解后的IMF样本熵展示各分量的复杂度,最后将复杂度较高的分量通过VMD再次分解,误差序列二次分解结果如图4所示。
图4所示,CEEMDAN将初步预测误差序列分解成多个IMF,然后将其重构并将高频分量再使用VMD分解,形成最终的预测序列。为进一步验证本文所提二次分解方法的优越性,将单次CEEMDAN分解、单次VMD分解与本文模型所构建的误差分量进行精度预测比较,验证模型性能。预测结果如表2所示。
表2得出,将本文模型与使用VMD-IDBO-BiLSTM-ATTENTION和CEEMDAN-IDBO-BiLSTM-ATTENTION组合模型预测结果比较,RMSE分别降低了5.178、2.748,MMAPE分别降低了3.34%、2.69%,R2分别提高了2.80%、1.01%。表2给出了3种模型预测精度结果,从表2可以看出,经过CEEMDAN-VMD二次分解过后的误差序列的预测精度更高,风电功率短期预测曲线更加贴和真实曲线。
为验证IDBO优化算法的性能,将IDBO与灰狼优化算法GWO、麻雀优化算法SSA、 鲸鱼优化算法WOA、苍鹰优化算法NGO、以及蜣螂优化算法DBO进行比较。设置GWO的参数a=2-2/M,其中M为最大迭代次数;NGO的攻击半径R=0.02(1-t)/M,其中t为迭代次数;WOA常数b=1;SSA预警值ST= 0.8发现者比例 PD=0.2,警戒者比例SD=0.2,DBO滚球蜣螂、产卵蜣螂、小蜣螂、偷窃蜣螂的比例分别为0.2、0.4、0.2、0.2;且6种优化算法的种群数量均为30,最大迭代次数为200,6种模型寻优收敛曲线如图5所示。
图5可看出,GWO收敛速度最慢,寻优效果最差,WOA和NGO算法收敛速度与适应度值量级相差不大,NGO略优于WOA,但收敛速度较慢;DBO与SSA收敛速度优于除IDBO外的其他算法,适应度指也相对优秀。DBO、SSA、IDBO适应度值在一个量级上,但在6种优化算法中IDBO收敛速度最快,且适应度值最优。因此,在这几种搜索优化算法中,IDBO 寻优能力最佳。
为验证本文模型的有效性,建立SVM、 CNN-BiLSTM、IDBO-BiLSTM、TCN-LSTM 4种组合预测模型模型,并与将本文模型比较,评估模型性能。结果如表3所示。
表3得出,将新衍生特征集输入到预测模型后,相较于将新数据集输入SVM、TCN-LSTM、CNN-BiLSTM、IDBO-BiLSTM组合模型,RMSE分别降低了30.281、6.956、4.979、6.144,MMAPE分别降低了15.18%、2.73%、3.45%、3.60%,R2分别提高了10.51%、2.65%、1.62%、2.57%各模型对验证数据的预测结果如图6所示。
通过图6对比LSTM模型与SVM模型在谷峰位置的预测值与实际值有很大误差,没有很好的拟合效果。而加入了卷积神经网络后的预测模型精度有了很大提升,且CNN-BiLSTM对比TCN-BiLSTM组合模型预测精度上仍有较大提升,虽然TCN模型能够捕捉长期依赖关系,可能比CNN更适合处理长时序预测问题,而在风电功率预测问题中,风速与功率的相关性较强,局部风速的增加或减少会极大地影响功率的大小,而CNN模型可以很好地捕捉局部特征,这也导致了CNN模型所训练的模型精度更加准确。而进行误差补偿后的功率在平缓区和突变区的预测功率精度明显高于其他模型,同时IDBO-BiLSTM组合模型由于IDBO具有更强的全局搜索能力,在局部突变位置的补偿效果也优于DBO-BiLSTM模型,对比以DBO寻优的误差补偿模型MMAPE降低了3.60%。因此,IDBO-BiLSTM相较于其他的组合模型,在运行平稳区段和局部突变区段拟合效果更加突出,预测准确度明显提升。
针对风电功率预测精度低的问题。提出了一种二次分解的误差补偿修正模型提高风功率预测精度,通过使用算例仿真和与传统预测模型和其他深度学习预测模型相比得出如下结论。
(1)利用CEEMDAN和VMD对预测误差与风功率特征进行分解,可以有效地降低功率的非平稳性,增强模型的性能。
(2)针对深度预测模型参数多难选取的问题,采用收敛速度快全局搜索能力强的改进蜣螂算法对模型参数进行寻优,避免因人工寻优选取参数速度慢、参数选择效果差的问题,大大提高了效率和精度。
(3)通过对误差序列进行分解-重构-分解的策略,在对高频误差进行二次分解降低非平稳性的同时加快的风功率预测效率,提高了预测精度。
(4)通过与传统预测模型以及其他组合深度学习模型进行对比,本文模型具有较高的预测精度和泛化能力。
通过新疆达坂城风电场的实测数据仿真实验表明,本文模型具有可行性,且具有更优的预测能力。同时也为提高时序预测问题的精度提供了一定的思路。
  • 黑龙江省自然科学基金(LH2019E001)
参考文献 引证文献
排序方式:
[1]
欧旭鹏, 唐云, 张凯, 等. 基于CEEMDAN-IDOA-BiLSTM的超短期风电功率预测[J]. 电网与清洁能源, 2023, 39(11): 142-150.
Ou Xupeng, Tang Yun, Zhang Kai, et al. Ultra-short term wind power prediction based on CEEMDAN-IDOA-BiLSTM[J]. Power Grids and Clean Energy, 2023, 39(11): 142-150.
[2]
Hu S, Xiang Y, Shen X D, et al. Wind power prediction model considering meteorological factor and spatial correlation of wind speed[J]. Automation of Electric Power Systems, 2021, 45(7): 28-36.
[3]
王泽华, 徐兴国, 叶子臣. 基于改进麻雀搜索算法优化SVM的风电功率预测研究[J]. 技术与市场, 2023, 30(8): 70-74, 79.
Wang Zehua, Xu Xingguo, Ye Zichen. Wind power prediction based on improved sparrow search algorithm and SVM optimization[J]. Technology & Field, 2023, 30(8): 70-74, 79.
[4]
谢波, 高建宇, 张惠娟, 等. 短期风电功率预测中的IOFA-SVM算法实现[J]. 电子测量技术, 2021, 44(12): 63-69.
Xie Bo, Gao Jianyu, Zhang Huijuan, et al. Implementation of IOFA-SVM algorithm for short-term wind power prediction[J]. Electronic Measurement Technology, 2021, 44(12): 63-69.
[5]
Yang G Q, Liu S I, Wang D Y, et al. Short-term wind power forecasting based on attention-gru wind speed correction and stacking[J]. Acta Energiae Solaris Sinica, 2022, 43(12): 273-281.
[6]
Qi F Z, Zhuo K X, Cao J. Short-term wind power probability density prediction method based on multi-level semantic attention mechanism[J]. Acta Energiae Solaris Sinica, 2022, 43(11): 140-147.
[7]
程杰, 严婷, 窦春霞, 等. 基于GWO-CNN-BiLSTM的超短期风电预测[J]. 科学技术与工程, 2023, 23(35): 15091-15099.
Cheng Jie, Yan Ting, Dou Chunxia, et al. Ultra-short term wind power prediction based on GWO-CNN-BiLSTM[J]. Science Technology and Engineering, 2019, 23(35): 15091-15099.
[8]
丁同, 傅晓锦, 刘明旺. 基于GA-VMD-BiLSTM算法的风电功率预测[J]. 扬州大学学报(自然科学版), 2022, 25(4): 44-49.
Ding Tong, Fu Xiaojin, Liu Mingwang. Wind power prediction based on GA-VMD-BiLSTM algorithm[J]. Journal of Yangzhou University (Natural Science Edition), 2022, 25(4): 44-49.
[9]
高鹭, 孔繁苗, 张飞, 等. 基于IPSO-BiLSTM-AM模型的超短期风电功率预测方法[J]. 智慧力, 2022, 50(4): 27-34.
Gao Lu, Kong Fanmiao, Zhang Fei, et al. Ultrashort term wind power prediction method based on IPSO-BiLSTM-AM model[J]. Smart Power, 2012, 50(4): 27-34.
[10]
Xiao L X, Zhang Y, Zhou H, et al. Ultra short trem wind power prediction based on IAOA-VMD-LSTM[J]. Acta Energiae Solaris Sinica, 2023, 44(11): 239-246.
[11]
Sheng S Q, Jin H Y, Liu C R. Short-term and mid-short-term wind power forecasting based on VMD-WSGRU[J]. Power System Technology, 2022, 46(3): 897-904.
[12]
Chang Y F, Yang Z X, Pan F, et al. Ultra-short-term wind power prediction based on CEEMDAN-PE-WPD and multi-objective optimization[J]. Power System Technology, 2023, 47(12): 5015-8026.
[13]
伍俊杰, 张倩, 李国丽, 等. 计及误差修正的变分模态分解-长短期记忆神经网络短期负荷预测[J]. 科学技术与工程, 2022, 22(12): 4828-4834.
Wu Junjie, Zhang Qian, Li Guoli, et al. Variational mode decomposition with error correction for short-term load forecasting by long short-term memory neural networks[J]. Science Technology and Engineering, 2002, 22(12): 4828-4834.
[14]
Xue J, Shen B. Dung beetle optimizer: a new meta heuristic algorithm for global optimization[J]. The Journal of Supercomputing, 2023, 79(7): 7305-7336.
[15]
Tanyildizi E, Demir G. Golden sine algorithm: a novel math inspired algorithm[J]. Advances in Electrical and Computer Engineering, 2017, 17(2): 71-79.
[16]
宋江涛, 崔双喜, 刘洪广. 基于二次分解NGO-VMD残差项与长短时记忆神经网络的超短期风功率预测[J]. 科学技术与工程, 2023, 23(6): 2428-2437.
Song Jiangtao, Cui Shuangxi, Liu Hongguang. Ultra-short term wind power prediction based on quadratic decomposition NGO-VMD residual term and short term memory neural network[J]. Science Technology and Engineering, 2019, 23(6): 2428-2437.
[17]
周洪煜, 曾济贫, 王照阳, 等. 基于混沌DNA遗传算法与PSO组合优化的RNN短期风电功率预测[J]. 电力系统保护与控制, 2013, 41(2): 144-149.
Zhou Hongyu, Zeng Jipin, Wang Zhaoyang, et al. RNN short-term wind power prediction based on chaotic DNA Genetic algorithm and PSO combination optimization[J]. Power System Protection and Control, 2013, 41(2): 144-149.
2025年第25卷第6期
PDF下载
388
154
引用本文
BibTeX
文章信息
doi: 10.12404/j.issn.1671-1815.2402025
  • 接收时间:2024-03-24
  • 首发时间:2025-07-27
  • 出版时间:2025-02-28
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-03-24
  • 修回日期:2024-12-10
基金
黑龙江省自然科学基金(LH2019E001)
作者信息
    1 东北林业大学机电工程学院, 哈尔滨 150040
    2 中国船舶集团有限公司第七O三研究所, 哈尔滨 150783

通讯作者:

* 姜雪松(1979—),男,汉族,黑龙江佳木斯人,博士,副教授。研究方向:工业工程与管理、智能制造工艺与装备。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2402025
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