Article(id=1322506210474050302, tenantId=1146029695717560320, journalId=1281212831689347082, issueId=1322506166672937516, articleNumber=null, orderNo=null, doi=10.14075/j.jgg.2024.10.469, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1728662400000, receivedDateStr=2024-10-12, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1793239651554, onlineDateStr=2026-10-29, pubDate=1755187200000, pubDateStr=2025-08-15, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1793239651554, onlineIssueDateStr=2026-10-29, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1793239651554, creator=13701087609, updateTime=1793239651554, updator=13701087609, issue=Issue{id=1322506166672937516, tenantId=1146029695717560320, journalId=1281212831689347082, year='2025', volume='45', issue='8', pageStart='771', pageEnd='880', issueExtLink='null', onlineDate='null', pubDate='1755187200000', pubDateStr='2025-08-15', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1793239641110, creator='13701087609', updateTime=1793239641110, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext=null, issueFiles=null, downloadFileDto=null}, startPage=781, endPage=790, ext={EN=ArticleExt(id=1322506210750874367, articleId=1322506210474050302, tenantId=1146029695717560320, journalId=1281212831689347082, language=EN, title=Multi-Scale Prediction of GNSS Elevation Time Series of CMONOC Based on CEEMDAN and TCN-Attention, columnId=null, journalTitle=Journal of Geodesy and Geodynamics, columnName=null, runingTitle=null, highlight=null, articleAbstract=

A multi-scale prediction model(referred to as C-TCN-A) based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN) and temporal convolutional network-attention mechanism(TCN-Attention) algorithms is proposed, which can be effectively applied to missing data imputation and future trend prediction of GNSS elevation time series. The model first employs CEEMDAN for multi-scale decomposition of time series, then utilizes TCN-Attention for prediction and reconstruction of different scale components to obtain final results. To validate the model's performance, 12 observation stations were selected for 1-day and 5-day predictions, with comparisons made against multiple other models. Results demonstrate that in 1-day prediction, C-TCN-A reduces RMSE and MAE by 35%-40% and 36%-41% respectively while improving correlation coefficient R by 25%-29%. For 5-day prediction, it achieves reductions of 20%-26% in RMSE and 20%-28% in MAE, with R increasing by 26%-33%. To verify the model's universality, C-TCN-A was applied to 99 stations from the crustal movement observation network of China(CMONOC) for 1-day and 5-day predictions. Results indicate generally favorable RMSE and MAE metrics with concentrated error distribution, where most errors remain below 4 mm. Spatial analysis reveals regional performance differences, with optimal results achieved in the northwestern region.

, authors=Yiyong LUO1, Aowen ZHAN1, *, Xiaohuan FENG1, authorsList=Yiyong LUO, Aowen ZHAN, Xiaohuan FENG, authorCompany=null, correspAuthors=Aowen ZHAN, authorNote=

LUO Yiyong, PhD, professor, majors in surveying and mapping processing, E-mail:

, correspAuthorsNote=null, copyrightStatement=All rights reserved. Unauthorized reproduction is prohibited., 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, fund=null), CN=ArticleExt(id=1322506215859536659, articleId=1322506210474050302, tenantId=1146029695717560320, journalId=1281212831689347082, language=CN, title=基于CEEMDAN 与TCN-Attention 的陆态网络GNSS高程时间序列多尺度预测, columnId=1322506169105633839, journalTitle=大地测量与地球动力学, columnName=现代大地测量, runingTitle=null, highlight=null, articleAbstract=

提出一种基于自适应噪声完备集合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise, CEEMDAN)和时间卷积网络-注意力机制(temporal convolutional network-attention mechanism, TCN-Attention)算法的多尺度预测模型(简称C-TCN-A),该模型可有效应用于GNSS高程时间序列缺失数据的插补和未来趋势的预测。该模型利用CEEMDAN对时间序列进行多尺度分解,然后基于TCN-Attention对不同尺度分量进行预测和重构得到预测结果。为验证模型的性能,选取12个观测站进行1 d与5 d预测,并与其他多种模型进行对比。结果表明,在1 d预测中,C-TCN-A的RMSE和MAE分别降低35%~40%和36%~41%,相关系数R提高25%~29%;在5 d预测中,C-TCN-A的RMSE和MAE分别降低20%~26%和20%~28%,相关系数R提高26%~33%。为验证模型的普适性,利用C-TCN-A对陆态网络99个观测站进行1 d与5 d预测。结果表明,RMSE和MAE指标总体上结果较好,误差分布集中,大多数误差小于4 mm; 预测精度存在一定的空间分布差异,西北地区效果最佳。

, authors=罗亦泳1, 占奥文1, *, 冯小欢1, authorsList=罗亦泳, 占奥文, 冯小欢, authorCompany=null, correspAuthors=占奥文, authorNote=

罗亦泳,博士,教授,主要研究方向为测绘数据处理,E-mail:

, correspAuthorsNote=

占奥文,硕士生,主要研究方向为测绘数据处理,E-mail:

ZHAN Aowen, postgraduate, majors in surveying and mapping processing, E-mail:

, copyrightStatement=版权所有,未经授权,不得转载。, copyrightOwner=大地测量与地球动力学编辑部, extLink=null, articleAbsUrl=null, sourceXml=BUeDY6K/pzYYZ2xPlTpjqA==, magXml=q6bQIvAIwiOjGeJGgrthUQ==, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=RXX9fha/5HFr9/YdW5Y7Yg==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=hbb/nZ/23f0h5nuY8KASwQ==, mapNumber=null, fund=null)}, authors=[Author(id=1322506218317398808, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ecityyluo@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322506218745217818, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506218317398808, language=EN, stringName=Yiyong LUO, firstName=Yiyong, middleName=null, lastName=LUO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China, bio={"content":"

LUO Yiyong, PhD, professor, majors in surveying and mapping processing, E-mail:

"}, bioImg=null, bioContent=

LUO Yiyong, PhD, professor, majors in surveying and mapping processing, E-mail:

, aboutCorrespAuthor=null), CN=AuthorExt(id=1322506219152065307, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506218317398808, language=CN, stringName=罗亦泳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013, bio={"content":"

罗亦泳,博士,教授,主要研究方向为测绘数据处理,E-mail:

"}, bioImg=null, bioContent=

罗亦泳,博士,教授,主要研究方向为测绘数据处理,E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])]), Author(id=1322506219210785565, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=19870023681@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322506219638604575, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506219210785565, language=EN, stringName=Aowen ZHAN, firstName=Aowen, middleName=null, lastName=ZHAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322506219965760288, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506219210785565, language=CN, stringName=占奥文, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])]), Author(id=1322506220032869154, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, 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=1322506220389384996, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506220032869154, language=EN, stringName=Xiaohuan FENG, firstName=Xiaohuan, middleName=null, lastName=FENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322506220460688165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506220032869154, language=CN, stringName=冯小欢, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])])], keywords=[Keyword(id=1322506220573934374, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, orderNo=1, keyword=GNSS elevation time series), Keyword(id=1322506220674597671, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, orderNo=2, keyword=CMONOC), Keyword(id=1322506220884312872, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, orderNo=3, keyword=improved empirical mode decomposition), Keyword(id=1322506220959810345, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, orderNo=4, keyword=temporal convolutional network), Keyword(id=1322506221039502122, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, orderNo=1, keyword=GNSS高程时间序列), Keyword(id=1322506222910161707, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, orderNo=2, keyword=陆态网络), Keyword(id=1322506223023407916, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, orderNo=3, keyword=改进经验模态分解), Keyword(id=1322506223333786413, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, orderNo=4, keyword=时间卷积网络)], refs=[Reference(id=1322506227989463891, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2016, volume=46, issue=6, pageStart=834, pageEnd=844, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=明锋, 杨元喜, 曾安敏, journalName=中国科学: 地球科学, refType=null, unstructuredReference=明锋, 杨元喜, 曾安敏, 等. 中国区域IGS站高程时间序列季节性信号及长期趋势分析[J]. 中国科学: 地球科学, 2016, 46 (6): 834- 844, articleTitle=中国区域IGS站高程时间序列季节性信号及长期趋势分析, refAbstract=null), Reference(id=1322506228048184148, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2016, volume=46, issue=6, pageStart=834, pageEnd=844, url=null, language=null, rfNumber=1, rfOrder=1, authorNames=Ming Feng, Yang Yuanxi, Zeng Anmin, journalName=null, refType=null, unstructuredReference= Ming Feng , Yang Yuanxi , Zeng Anmin , et al. Analysis of Seasonal Signals and Long-Term Trends in the Height Time Series of IGS Sites in China[J]. Scientia Sinica Terrae, 2016, 46 (6): 834- 844, articleTitle=null, refAbstract=null), Reference(id=1322506228127875925, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2023, volume=15, issue=18, pageStart=null, pageEnd=null, url=http://www.keyanzhidian.com/doc/detail?id=2074000763, language=null, rfNumber=2, rfOrder=2, authorNames=Li Z, Lu T D, Yu K G, journalName=Remote Sensing, refType=null, unstructuredReference= Li Z , Lu T D , Yu K G , et al. Interpolation of GNSS Position Time Series Using GBDT, XGBoost, and RF Machine Learning Algorithms and Models Error Analysis[J]. Remote Sensing, 2023, 15 (18), articleTitle=Interpolation of GNSS Position Time Series Using GBDT, XGBoost, and RF Machine Learning Algorithms and Models Error Analysis, refAbstract=null), Reference(id=1322506228190790486, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2021, volume=13, issue=19, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=3, authorNames=Crocetti L, Schartner M, Soja B, journalName=Remote Sensing, refType=null, unstructuredReference= Crocetti L , Schartner M , Soja B . Discontinuity Detection in GNSS Station Coordinate Time Series Using Machine Learning[J]. Remote Sensing, 2021, 13 (19), articleTitle=Discontinuity Detection in GNSS Station Coordinate Time Series Using Machine Learning, refAbstract=null), Reference(id=1322506228274676567, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2022, volume=96, issue=10, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=4, authorNames=Gao W Z, Li Z, Chen Q S, journalName=Journal of Geodesy, refType=null, unstructuredReference= Gao W Z , Li Z , Chen Q S , et al. Modelling and Prediction of GNSS Time Series Using GBDT, LSTM and SVM Machine Learning Approaches[J]. Journal of Geodesy, 2022, 96 (10), articleTitle=Modelling and Prediction of GNSS Time Series Using GBDT, LSTM and SVM Machine Learning Approaches, refAbstract=null), Reference(id=1322506228350174040, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=10.14075/j.jgg.2022.09.004, pmid=null, pmcid=null, year=2022, volume=42, issue=9, pageStart=898, pageEnd=903, url=null, language=null, rfNumber=5, rfOrder=5, authorNames=鲁铁定, 李祯, journalName=大地测量与地球动力学, refType=null, unstructuredReference=鲁铁定, 李祯. 基于Prophet-XGBoost模型的GNSS高程时间序列预测[J]. 大地测量与地球动力学, 2022, 42 (9): 898- 903, articleTitle=基于Prophet-XGBoost模型的GNSS高程时间序列预测, refAbstract=null), Reference(id=1322506228425671513, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=10.14075/j.jgg.2022.09.004, pmid=null, pmcid=null, year=2022, volume=42, issue=9, pageStart=898, pageEnd=903, url=null, language=null, rfNumber=5, rfOrder=6, authorNames=Lu Tieding, Li Zhen, journalName=null, refType=null, unstructuredReference= Lu Tieding , Li Zhen . Prediction of GNSS Vertical Coordinate Time Series Based on Prophet-XGBoost Model[J]. Journal of Geodesy and Geodynamics, 2022, 42 (9): 898- 903, articleTitle=null, refAbstract=null), Reference(id=1322506228484391770, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=10.14075/j.jgg.2021.02.002, pmid=null, pmcid=null, year=2021, volume=41, issue=2, pageStart=116, pageEnd=121, url=null, language=null, rfNumber=6, rfOrder=7, authorNames=李威, 鲁铁定, 贺小星, journalName=大地测量与地球动力学, refType=null, unstructuredReference=李威, 鲁铁定, 贺小星, 等. 基于Prophet-RF模型的GNSS高程坐标时间序列预测分析[J]. 大地测量与地球动力学, 2021, 41 (2): 116- 121, articleTitle=基于Prophet-RF模型的GNSS高程坐标时间序列预测分析, refAbstract=null), Reference(id=1322506228568277851, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=10.14075/j.jgg.2021.02.002, pmid=null, pmcid=null, year=2021, volume=41, issue=2, pageStart=116, pageEnd=121, url=null, language=null, rfNumber=6, rfOrder=8, authorNames=Li Wei, Lu Tieding, He Xiaoxing, journalName=null, refType=null, unstructuredReference= Li Wei , Lu Tieding , He Xiaoxing , et al. Prediction and Analysis of GNSS Vertical Coordinate Time Series Based on Prophet-RF Model[J]. Journal of Geodesy and Geodynamics, 2021, 41 (2): 116- 121, articleTitle=null, refAbstract=null), Reference(id=1322506228660552540, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2023, volume=28, issue=1, pageStart=null, pageEnd=null, url=http://link.springer.com/doi/10.1007/s10291-023-01544-z, language=null, rfNumber=7, rfOrder=9, authorNames=Jiang W P, Wang J, Li Z, journalName=GPS Solutions, refType=null, unstructuredReference= Jiang W P , Wang J , Li Z , et al. A New Deep Self-Attention Neural Network for GNSS Coordinate Time Series Prediction[J]. GPS Solutions, 2023, 28 (1), articleTitle=A New Deep Self-Attention Neural Network for GNSS Coordinate Time Series Prediction, refAbstract=null), Reference(id=1322506228723467101, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2021, volume=13, issue=16, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=10, authorNames=Wang J, Jiang W P, Li Z, journalName=Remote Sensing, refType=null, unstructuredReference= Wang J , Jiang W P , Li Z , et al. A New Multi-Scale Sliding Window LSTM Framework(MSSW-LSTM): A Case Study for GNSS Time-Series Prediction[J]. Remote Sensing, 2021, 13 (16), articleTitle=A New Multi-Scale Sliding Window LSTM Framework(MSSW-LSTM): A Case Study for GNSS Time-Series Prediction, refAbstract=null), Reference(id=1322506228815741790, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2018, volume=43, issue=12, pageStart=2112, pageEnd=2123, url=null, language=null, rfNumber=9, rfOrder=11, authorNames=姜卫平, 王锴华, 李昭, journalName=武汉大学学报: 信息科学版, refType=null, unstructuredReference=姜卫平, 王锴华, 李昭, 等. GNSS坐标时间序列分析理论与方法及展望[J]. 武汉大学学报: 信息科学版, 2018, 43 (12): 2112- 2123, articleTitle=GNSS坐标时间序列分析理论与方法及展望, refAbstract=null), Reference(id=1322506228882850655, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2018, volume=43, issue=12, pageStart=2112, pageEnd=2123, url=null, language=null, rfNumber=9, rfOrder=12, authorNames=Jiang Weiping, Wang Kaihua, Li Zhao, journalName=null, refType=null, unstructuredReference= Jiang Weiping , Wang Kaihua , Li Zhao , et al. Prospect and Theory of GNSS Coordinate Time Series Analysis[J]. Geomatics and Information Science of Wuhan University, 2018, 43 (12): 2112- 2123, articleTitle=null, refAbstract=null), Reference(id=1322506228991902560, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2023, volume=52, issue=8, pageStart=1235, pageEnd=1244, url=null, language=null, rfNumber=10, rfOrder=13, authorNames=鲁铁定, 李祯, 贺小星, journalName=测绘学报, refType=null, unstructuredReference=鲁铁定, 李祯, 贺小星, 等. 融合VMD和XGBoost算法的GNSS高程时间序列预测方法[J]. 测绘学报, 2023, 52 (8): 1235- 1244, articleTitle=融合VMD和XGBoost算法的GNSS高程时间序列预测方法, refAbstract=null), Reference(id=1322506229054817121, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2023, volume=52, issue=8, pageStart=1235, pageEnd=1244, url=null, language=null, rfNumber=10, rfOrder=14, authorNames=Lu Tieding, Li Zhen, He Xiaoxing, journalName=null, refType=null, unstructuredReference= Lu Tieding , Li Zhen , He Xiaoxing , et al. GNSS Vertical Time Series Prediction Method Integrating VMD and XGBoost Algorithms[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52 (8): 1235- 1244, articleTitle=null, refAbstract=null), Reference(id=1322506229121925986, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2023, volume=15, issue=14, pageStart=null, pageEnd=null, url=http://openurl.ebsco.com/contentitem/doi:10.3390%2Frs15143694?sid=ebsco:plink:crawler&id=ebsco:doi:10.3390%2Frs15143694, language=null, rfNumber=11, rfOrder=15, authorNames=Chen H K, Lu T D, Huang J H, journalName=Remote Sensing, refType=null, unstructuredReference= Chen H K , Lu T D , Huang J H , et al. An Improved VMD-LSTM Model for Time-Varying GNSS Time Series Prediction with Temporally Correlated Noise[J]. Remote Sensing, 2023, 15 (14), articleTitle=An Improved VMD-LSTM Model for Time-Varying GNSS Time Series Prediction with Temporally Correlated Noise, refAbstract=null), Reference(id=1322506229189034851, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2022, volume=32, issue=增2, pageStart=174, pageEnd=177, url=null, language=null, rfNumber=12, rfOrder=16, authorNames=符娇, 宋圆, 段炬奎, journalName=安徽地质, refType=null, unstructuredReference=符娇, 宋圆, 段炬奎, 等. 基于EMD-Prophet模型的GPS坐标时间序列预测[J]. 安徽地质, 2022, 32 (增2): 174- 177, articleTitle=基于EMD-Prophet模型的GPS坐标时间序列预测, refAbstract=null), Reference(id=1322506229272920932, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2022, volume=32, issue=S2, pageStart=174, pageEnd=177, url=null, language=null, rfNumber=12, rfOrder=17, authorNames=Fu Jiao, Song Yuan, Duan Jukui, journalName=null, refType=null, unstructuredReference= Fu Jiao , Song Yuan , Duan Jukui , et al. Prediction of GPS Coordinate Time Series Based on EMD-Prophet Model[J]. Geology of Anhui, 2022, 32 (S2): 174- 177, articleTitle=null, refAbstract=null), Reference(id=1322506229335835493, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2021, volume=35, issue=12, pageStart=11106, pageEnd=11115, url=http://www.xueshufan.com/publication/3177318507, language=null, rfNumber=13, rfOrder=18, authorNames=Zhou H Y, Zhang S H, Peng J Q, journalName=Proceedings of the AAAI Conference on Artificial Intelligence, refType=null, unstructuredReference= Zhou H Y , Zhang S H , Peng J Q , et al. Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2021, 35 (12): 11106- 11115, articleTitle=Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting, refAbstract=null), Reference(id=1322506229449081702, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Bai S J, Zico Kolter J, Koltun V. An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling[EB/OL]. https://arxiv.org/abs/1803.01271, 2018-04-19, articleTitle=null, refAbstract=null), Reference(id=1322506229516190567, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Torres M E, Colominas M A, Schlotthauer G, et al. A Complete Ensemble Empirical Mode Decomposition with Adaptive Noise[C]. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP), Prague, 2011, articleTitle=null, refAbstract=null), Reference(id=1322506229587493736, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2024, volume=305, issue=null, pageStart=null, pageEnd=null, url=http://www.sciencedirect.com/science/article/pii/S0360544224020024, language=null, rfNumber=16, rfOrder=21, authorNames=Sun X Y, Liu H Z, journalName=Energy, refType=null, unstructuredReference= Sun X Y , Liu H Z . Multivariate Short-Term Wind Speed Prediction Based on PSO-VMD-SE-ICEEMDAN Two-Stage Decomposition and Att-S2S[J]. Energy, 2024, 305, articleTitle=Multivariate Short-Term Wind Speed Prediction Based on PSO-VMD-SE-ICEEMDAN Two-Stage Decomposition and Att-S2S, refAbstract=null), Reference(id=1322506229671379817, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2024, volume=43, issue=3, pageStart=192, pageEnd=200, url=null, language=null, rfNumber=17, rfOrder=22, authorNames=张文乾, 刘金凤, 江军, journalName=电力工程技术, refType=null, unstructuredReference=张文乾, 刘金凤, 江军, 等. 基于CEEMDAN和TCN的变压器油中溶解气体含量预测[J]. 电力工程技术, 2024, 43 (3): 192- 200, articleTitle=基于CEEMDAN和TCN的变压器油中溶解气体含量预测, refAbstract=null), Reference(id=1322506229742682986, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2024, volume=43, issue=3, pageStart=192, pageEnd=200, url=null, language=null, rfNumber=17, rfOrder=23, authorNames=Zhang Wenqian, Liu Jinfeng, Jiang Jun, journalName=null, refType=null, unstructuredReference= Zhang Wenqian , Liu Jinfeng , Jiang Jun , et al. Prediction of Concentration for Dissolved Gas in Oil Based on CEEMDAN and TCN[J]. Electric Power Engineering Technology, 2024, 43 (3): 192- 200, articleTitle=null, refAbstract=null), Reference(id=1322506229834957675, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=Bahdanau D, Cho K, Bengio Y. Neural Machine Translation by Jointly Learning to Align and Translate[EB/OL]. https://arxiv.org/abs/1409.0473, 2014-05-19, articleTitle=null, refAbstract=null), Reference(id=1322506229914649452, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2020, volume=8, issue=5, pageStart=85, pageEnd=91, url=null, language=null, rfNumber=19, rfOrder=25, authorNames=周建营, 陈国恒, 刘文建, journalName=导航定位学报, refType=null, unstructuredReference=周建营, 陈国恒, 刘文建. GAMIT/GLOBK多系统GNSS数据处理分析[J]. 导航定位学报, 2020, 8 (5): 85- 91, articleTitle=GAMIT/GLOBK多系统GNSS数据处理分析, refAbstract=null), Reference(id=1322506230023701357, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2020, volume=8, issue=5, pageStart=85, pageEnd=91, url=null, language=null, rfNumber=19, rfOrder=26, authorNames=Zhou Jianying, Chen Guoheng, Liu Wenjian, journalName=null, refType=null, unstructuredReference= Zhou Jianying , Chen Guoheng , Liu Wenjian . GAMIT/GLOBK Multi-System GNSS Data Processing Analysis[J]. Journal of Navigation and Positioning, 2020, 8 (5): 85- 91, articleTitle=null, refAbstract=null), Reference(id=1322506231248438126, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, doi=null, pmid=null, pmcid=null, year=2017, volume=21, issue=3, pageStart=1389, pageEnd=1394, url=http://www.keyanzhidian.com/doc/detail?id=2031684016, language=null, rfNumber=20, rfOrder=27, authorNames=Wu D C, Yan H M, Shen Y C, journalName=GPS Solutions, refType=null, unstructuredReference= Wu D C , Yan H M , Shen Y C . TSAnalyzer, a GNSS Time Series Analysis Software[J]. GPS Solutions, 2017, 21 (3): 1389- 1394, articleTitle=TSAnalyzer, a GNSS Time Series Analysis Software, refAbstract=null)], funds=[Fund(id=1322506227842663249, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, awardId=41861058, language=EN, fundingSource=National Natural Science Foundation of China(41861058), fundOrder=null, country=null), Fund(id=1322506227901383506, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, awardId=41861058, language=CN, fundingSource=国家自然科学基金(41861058), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])], figs=[ArticleFig(id=1322506223518335790, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 1, caption=TCN-Attention model, figureFileSmall=ejELYiQW9uXGb8kods49yw==, figureFileBig=RXX9fha/5HFr9/YdW5Y7Yg==, tableContent=null), ArticleFig(id=1322506223593833263, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图1, caption=TCN-Attention模型, figureFileSmall=ejELYiQW9uXGb8kods49yw==, figureFileBig=RXX9fha/5HFr9/YdW5Y7Yg==, tableContent=null), ArticleFig(id=1322506223828714288, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 2, caption=Model prediction flow chart, figureFileSmall=CGlv+GH+YbYox6/budzVMg==, figureFileBig=42fnVt082Z6xJwEjYSaCfg==, tableContent=null), ArticleFig(id=1322506223887434545, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图2, caption=模型预测流程, figureFileSmall=CGlv+GH+YbYox6/budzVMg==, figureFileBig=42fnVt082Z6xJwEjYSaCfg==, tableContent=null), ArticleFig(id=1322506223954543410, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 3, caption=CEEMDAN decomposes HRBN site data, figureFileSmall=QljSV9BO5LZY5ZsEG3B5pw==, figureFileBig=qa0kfIABU+HSwwDQOswNcw==, tableContent=null), ArticleFig(id=1322506224030040883, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图3, caption=CEEMDAN分解HRBN站点数据, figureFileSmall=QljSV9BO5LZY5ZsEG3B5pw==, figureFileBig=qa0kfIABU+HSwwDQOswNcw==, tableContent=null), ArticleFig(id=1322506224092955444, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 4, caption=Prediction results of recombination component modeling, figureFileSmall=XvdT6MRt4LyiTnf+TN87qQ==, figureFileBig=DsI60VpAQu6r9oeWLvB8bQ==, tableContent=null), ArticleFig(id=1322506224168452917, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图4, caption=重组分量建模预测结果, figureFileSmall=XvdT6MRt4LyiTnf+TN87qQ==, figureFileBig=DsI60VpAQu6r9oeWLvB8bQ==, tableContent=null), ArticleFig(id=1322506224227173174, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 5, caption=Comparative analysis of high-frequency component prediction effect, figureFileSmall=WyKj8MSjbqKHLKoSgo5scw==, figureFileBig=Z73Iyr8UxU7rWbUEz3hLnA==, tableContent=null), ArticleFig(id=1322506224306864951, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图5, caption=高频分量预测效果对比分析, figureFileSmall=WyKj8MSjbqKHLKoSgo5scw==, figureFileBig=Z73Iyr8UxU7rWbUEz3hLnA==, tableContent=null), ArticleFig(id=1322506224365585208, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 6, caption=GNSS sites distribution, figureFileSmall=aVNGtKlV9X2sVyuP8NQlmg==, figureFileBig=mtcLUifpeFWe42djRFu/wQ==, tableContent=null), ArticleFig(id=1322506224428499769, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图6, caption=GNSS测站分布, figureFileSmall=aVNGtKlV9X2sVyuP8NQlmg==, figureFileBig=mtcLUifpeFWe42djRFu/wQ==, tableContent=null), ArticleFig(id=1322506224508191546, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 7, caption=Comparison of prediction results of different models, figureFileSmall=f9ys/VtYzehrddfEVPuYWw==, figureFileBig=a7ml5qzYFJsYSbgqpTNu3A==, tableContent=null), ArticleFig(id=1322506224566911803, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图7, caption=各模型预测结果对比, figureFileSmall=f9ys/VtYzehrddfEVPuYWw==, figureFileBig=a7ml5qzYFJsYSbgqpTNu3A==, tableContent=null), ArticleFig(id=1322506224650797884, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 8, caption=Spatial distribution of 1-day prediction error of CMONOC elevation time series, figureFileSmall=LDB8gS99ovslWk8VDJIVUQ==, figureFileBig=B9w0uQ3QrBcTsClfLMZYqg==, tableContent=null), ArticleFig(id=1322506224717906749, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图8, caption=陆态网络高程时间序列1 d预测误差空间分布

底图使用审图号GS(2016)1552号, 无修改

, figureFileSmall=LDB8gS99ovslWk8VDJIVUQ==, figureFileBig=B9w0uQ3QrBcTsClfLMZYqg==, tableContent=null), ArticleFig(id=1322506224814375742, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Fig. 9, caption=Statistical analysis of 1-day prediction error of CMONOC elevation time series, figureFileSmall=uSCiPHMCaDIVOtnEr8y/4g==, figureFileBig=chnG6Lz2koueJ/J2j+59yA==, tableContent=null), ArticleFig(id=1322506224885678911, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=图9, caption=陆态网络高程时间序列1 d预测误差统计分析, figureFileSmall=uSCiPHMCaDIVOtnEr8y/4g==, figureFileBig=chnG6Lz2koueJ/J2j+59yA==, tableContent=null), ArticleFig(id=1322506224944399168, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 1, caption=

Partial core parameters of the model

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 模型核心参数
XGboost 学习率: 0.001;迭代次数: 100;批大小: 64
CNN 卷积核: 64;卷积核大小: 3;丢弃率: 0.5;激活函数: ReLU; 批大小: 64;迭代次数: 100
LSTM 隐藏层层数: 3;隐藏层神经元: 64;丢弃率: 0.5;激活函数: ReLU; 批大小: 64;迭代次数: 100
CNN-LSTM CNN卷积核: 64;卷积核大小: 3;LSTM隐藏层层数: 3;隐藏层神经元: 64;丢弃率: 0.5;激活函数: ReLU; 批大小: 64;迭代次数: 100
TCN 学习率: 0.001;卷积核大小: 5;丢弃率: 0.5;步长: 1;膨胀系数: [1, 2, 4, 8]; 迭代次数: 100;批大小: 64
informer 学习率: 0.001;批大小: 32;迭代次数: 100;丢弃率: 0.5;激活函数: GELU
C-TCN-A CEEMDAN噪声添加次数: 100;信噪比: 0.005
TCN-A学习率: 0.001;卷积核大小: 5;丢弃率: 0.5;步长: 1;膨胀系数: [1, 2, 4, 8]; 迭代次数: 100;批大小: 64
), ArticleFig(id=1322506225003119425, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表1, caption=

模型部分核心参数

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 模型核心参数
XGboost 学习率: 0.001;迭代次数: 100;批大小: 64
CNN 卷积核: 64;卷积核大小: 3;丢弃率: 0.5;激活函数: ReLU; 批大小: 64;迭代次数: 100
LSTM 隐藏层层数: 3;隐藏层神经元: 64;丢弃率: 0.5;激活函数: ReLU; 批大小: 64;迭代次数: 100
CNN-LSTM CNN卷积核: 64;卷积核大小: 3;LSTM隐藏层层数: 3;隐藏层神经元: 64;丢弃率: 0.5;激活函数: ReLU; 批大小: 64;迭代次数: 100
TCN 学习率: 0.001;卷积核大小: 5;丢弃率: 0.5;步长: 1;膨胀系数: [1, 2, 4, 8]; 迭代次数: 100;批大小: 64
informer 学习率: 0.001;批大小: 32;迭代次数: 100;丢弃率: 0.5;激活函数: GELU
C-TCN-A CEEMDAN噪声添加次数: 100;信噪比: 0.005
TCN-A学习率: 0.001;卷积核大小: 5;丢弃率: 0.5;步长: 1;膨胀系数: [1, 2, 4, 8]; 迭代次数: 100;批大小: 64
), ArticleFig(id=1322506225061839682, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 2, caption=

Comparison of prediction accuracy of different models at different stations (1 day)

, figureFileSmall=null, figureFileBig=null, tableContent=
测站 模型 RMSE/mm MAE/mm R t/s 测站 RMSE/mm MAE/mm R t/s
HRBN XGboost 5.58 4.16 0.75 32.26 BJFS 4.92 3.82 0.82 34.46
CNN 5.42 4.04 0.76 8.47 5.46 4.38 0.82 8.26
LSTM 5.31 3.98 0.77 45.92 5.15 4.07 0.82 56.08
CNN-LSTM 5.39 4.07 0.77 31.32 4.94 3.87 0.83 32.64
informer 5.20 3.81 0.78 31.50 4.92 3.80 0.82 26.17
TCN 5.18 3.78 0.78 12.13 4.87 3.77 0.83 14.69
C-TCN-A 3.03 2.16 0.93 105.81 3.32 2.41 0.93 104.90
XJSH XGboost 4.84 4.16 0.78 35.30 XJBC 4.86 3.89 0.68 33.28
CNN 5.08 4.04 0.80 7.85 4.66 3.56 0.68 8.92
LSTM 4.76 3.98 0.80 53.75 4.58 3.48 0.69 50.55
CNN-LSTM 4.60 4.07 0.80 31.68 4.57 3.49 0.69 34.52
informer 4.58 3.49 0.80 30.51 4.66 3.50 0.69 41.33
TCN 4.56 3.78 0.80 14.14 4.56 3.48 0.70 14.23
C-TCN-A 3.03 2.16 0.92 103.93 3.03 2.41 0.93 103.23
YNLJ XGboost 5.36 4.18 0.81 32.12 SCSP 5.61 3.94 0.66 31.24
CNN 5.56 4.37 0.82 9.08 5.85 4.23 0.68 8.38
LSTM 5.32 4.15 0.82 40.65 5.58 4.06 0.70 44.18
CNN-LSTM 5.47 4.28 0.81 31.72 5.54 4.02 0.70 45.21
informer 5.36 4.18 0.82 43.23 5.37 3.84 0.70 41.39
TCN 5.35 4.18 0.82 14.82 5.21 3.56 0.70 15.39
C-TCN-A 3.15 2.43 0.94 103.38 3.29 2.25 0.89 105.14
AHBB XGboost 5.92 4.33 0.61 34.32 GDSG 6.72 5.07 0.44 31.46
CNN 6.10 4.47 0.60 8.32 7.13 5.45 0.43 8.56
LSTM 6.07 4.67 0.67 45.60 6.88 5.28 0.43 47.91
CNN-LSTM 5.69 4.15 0.66 32.62 6.85 5.25 0.44 45.94
informer 5.53 4.03 0.66 41.43 6.69 5.04 0.42 29.86
TCN 5.40 3.89 0.63 12.34 6.67 5.01 0.44 12.41
C-TCN-A 3.47 2.64 0.87 105.81 4.29 3.26 0.82 104.78
XJQM XGboost 4.54 3.45 0.67 32.68 NMEJ 4.76 3.60 0.73 32.16
CNN 4.61 3.57 0.69 12.61 5.27 4.14 0.68 8.11
LSTM 4.48 3.42 0.70 45.03 4.76 3.63 0.69 45.73
CNN-LSTM 4.42 3.34 0.70 34.63 4.69 3.52 0.69 34.41
informer 4.43 3.34 0.69 26.25 4.66 3.50 0.69 26.11
TCN 4.35 3.27 0.70 12.26 4.56 3.42 0.70 11.72
C-TCN-A 2.83 2.23 0.88 108.4 3.23 2.46 0.93 106.41
HLMH XGboost 5.25 4.01 0.62 40.32 LHAZ 5.81 4.30 0.84 36.28
CNN 5.50 4.24 0.63 8.01 5.94 4.30 0.84 11.29
LSTM 5.39 4.16 0.62 40.54 5.78 4.24 0.85 45.27
CNN-LSTM 5.33 4.10 0.63 32.78 5.84 4.26 0.85 39.97
informer 5.22 3.93 0.63 41.48 5.84 4.24 0.83 41.33
TCN 5.18 3.92 0.63 12.98 5.68 4.11 0.85 15.57
C-TCN-A 3.53 2.51 0.84 105.21 3.96 2.88 0.94 101.97
), ArticleFig(id=1322506225145725764, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表2, caption=

各测站不同模型预测精度比较(1 d)

, figureFileSmall=null, figureFileBig=null, tableContent=
测站 模型 RMSE/mm MAE/mm R t/s 测站 RMSE/mm MAE/mm R t/s
HRBN XGboost 5.58 4.16 0.75 32.26 BJFS 4.92 3.82 0.82 34.46
CNN 5.42 4.04 0.76 8.47 5.46 4.38 0.82 8.26
LSTM 5.31 3.98 0.77 45.92 5.15 4.07 0.82 56.08
CNN-LSTM 5.39 4.07 0.77 31.32 4.94 3.87 0.83 32.64
informer 5.20 3.81 0.78 31.50 4.92 3.80 0.82 26.17
TCN 5.18 3.78 0.78 12.13 4.87 3.77 0.83 14.69
C-TCN-A 3.03 2.16 0.93 105.81 3.32 2.41 0.93 104.90
XJSH XGboost 4.84 4.16 0.78 35.30 XJBC 4.86 3.89 0.68 33.28
CNN 5.08 4.04 0.80 7.85 4.66 3.56 0.68 8.92
LSTM 4.76 3.98 0.80 53.75 4.58 3.48 0.69 50.55
CNN-LSTM 4.60 4.07 0.80 31.68 4.57 3.49 0.69 34.52
informer 4.58 3.49 0.80 30.51 4.66 3.50 0.69 41.33
TCN 4.56 3.78 0.80 14.14 4.56 3.48 0.70 14.23
C-TCN-A 3.03 2.16 0.92 103.93 3.03 2.41 0.93 103.23
YNLJ XGboost 5.36 4.18 0.81 32.12 SCSP 5.61 3.94 0.66 31.24
CNN 5.56 4.37 0.82 9.08 5.85 4.23 0.68 8.38
LSTM 5.32 4.15 0.82 40.65 5.58 4.06 0.70 44.18
CNN-LSTM 5.47 4.28 0.81 31.72 5.54 4.02 0.70 45.21
informer 5.36 4.18 0.82 43.23 5.37 3.84 0.70 41.39
TCN 5.35 4.18 0.82 14.82 5.21 3.56 0.70 15.39
C-TCN-A 3.15 2.43 0.94 103.38 3.29 2.25 0.89 105.14
AHBB XGboost 5.92 4.33 0.61 34.32 GDSG 6.72 5.07 0.44 31.46
CNN 6.10 4.47 0.60 8.32 7.13 5.45 0.43 8.56
LSTM 6.07 4.67 0.67 45.60 6.88 5.28 0.43 47.91
CNN-LSTM 5.69 4.15 0.66 32.62 6.85 5.25 0.44 45.94
informer 5.53 4.03 0.66 41.43 6.69 5.04 0.42 29.86
TCN 5.40 3.89 0.63 12.34 6.67 5.01 0.44 12.41
C-TCN-A 3.47 2.64 0.87 105.81 4.29 3.26 0.82 104.78
XJQM XGboost 4.54 3.45 0.67 32.68 NMEJ 4.76 3.60 0.73 32.16
CNN 4.61 3.57 0.69 12.61 5.27 4.14 0.68 8.11
LSTM 4.48 3.42 0.70 45.03 4.76 3.63 0.69 45.73
CNN-LSTM 4.42 3.34 0.70 34.63 4.69 3.52 0.69 34.41
informer 4.43 3.34 0.69 26.25 4.66 3.50 0.69 26.11
TCN 4.35 3.27 0.70 12.26 4.56 3.42 0.70 11.72
C-TCN-A 2.83 2.23 0.88 108.4 3.23 2.46 0.93 106.41
HLMH XGboost 5.25 4.01 0.62 40.32 LHAZ 5.81 4.30 0.84 36.28
CNN 5.50 4.24 0.63 8.01 5.94 4.30 0.84 11.29
LSTM 5.39 4.16 0.62 40.54 5.78 4.24 0.85 45.27
CNN-LSTM 5.33 4.10 0.63 32.78 5.84 4.26 0.85 39.97
informer 5.22 3.93 0.63 41.48 5.84 4.24 0.83 41.33
TCN 5.18 3.92 0.63 12.98 5.68 4.11 0.85 15.57
C-TCN-A 3.53 2.51 0.84 105.21 3.96 2.88 0.94 101.97
), ArticleFig(id=1322506225246389061, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 3, caption=

Average evaluation index of 1-day prediction results of each model

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE/mm RMSE改善率/% MAE/mm MAE改善率/% R R改善率/% t/s
XGboost 5.35 38 4.06 39 0.70 29 33.82
CNN 5.55 40 4.23 41 0.70 29 8.99
LSTM 5.34 37 4.09 39 0.71 27 46.77
CNN-LSTM 5.28 37 4.04 39 0.71 27 35.62
informer 5.21 36 3.89 36 0.71 27 35.65
TCN 5.13 35 3.85 36 0.72 25 13.56
C-TCN-A 3.34 - 2.48 - 0.90 - 104.91
), ArticleFig(id=1322506225409966918, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表3, caption=

各模型1 d预测结果评价指标平均值

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE/mm RMSE改善率/% MAE/mm MAE改善率/% R R改善率/% t/s
XGboost 5.35 38 4.06 39 0.70 29 33.82
CNN 5.55 40 4.23 41 0.70 29 8.99
LSTM 5.34 37 4.09 39 0.71 27 46.77
CNN-LSTM 5.28 37 4.04 39 0.71 27 35.62
informer 5.21 36 3.89 36 0.71 27 35.65
TCN 5.13 35 3.85 36 0.72 25 13.56
C-TCN-A 3.34 - 2.48 - 0.90 - 104.91
), ArticleFig(id=1322506225489658695, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 4, caption=

Average evaluation index for 5-day prediction results of each model

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE/mm RMSE改善率/% MAE/mm MAE改善率/% R R改善率/% t/s
XGboost 6.26 25 4.89 27 0.58 33 35.46
CNN 6.17 24 4.78 25 0.59 31 9.63
LSTM 6.36 26 5.00 28 0.59 31 44.58
CNN-LSTM 6.26 25 4.88 27 0.59 31 36.53
informer 5.91 20 4.54 21 0.60 28 24.79
TCN 5.84 20 4.49 20 0.61 26 16.20
C-TCN-A 4.70 - 3.58 - 0.77 - 103.03
), ArticleFig(id=1322506225552573256, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表4, caption=

各模型5 d预测结果评价指标平均值

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 RMSE/mm RMSE改善率/% MAE/mm MAE改善率/% R R改善率/% t/s
XGboost 6.26 25 4.89 27 0.58 33 35.46
CNN 6.17 24 4.78 25 0.59 31 9.63
LSTM 6.36 26 5.00 28 0.59 31 44.58
CNN-LSTM 6.26 25 4.88 27 0.59 31 36.53
informer 5.91 20 4.54 21 0.60 28 24.79
TCN 5.84 20 4.49 20 0.61 26 16.20
C-TCN-A 4.70 - 3.58 - 0.77 - 103.03
), ArticleFig(id=1322506227205129033, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 5, caption=

1-day prediction accuracy index of CMONOC elevation time series

, figureFileSmall=null, figureFileBig=null, tableContent=
最大RMSE/mm 平均RMSE/mm 最小RMSE/mm 最大MAE/mm 平均MAE/mm 最小MAE/mm 最大R 平均R 最小R
6.46 3.71 2.27 5 2.8 1.7 0.96 0.89 0.82
), ArticleFig(id=1322506227293209418, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表5, caption=

陆态网络高程时间序列1 d预测精度指标

, figureFileSmall=null, figureFileBig=null, tableContent=
最大RMSE/mm 平均RMSE/mm 最小RMSE/mm 最大MAE/mm 平均MAE/mm 最小MAE/mm 最大R 平均R 最小R
6.46 3.71 2.27 5 2.8 1.7 0.96 0.89 0.82
), ArticleFig(id=1322506227372901195, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 6, caption=

5-day prediction accuracy index of CMONOC elevation time series

, figureFileSmall=null, figureFileBig=null, tableContent=
最大RMSE/mm 平均RMSE/mm 最小RMSE/mm 最大MAE/mm 平均MAE/mm 最小MAE/mm 最大R 平均R 最小R
7.88 5.24 3.56 5.85 3.98 2.67 0.93 0.76 0.55
), ArticleFig(id=1322506227448398668, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表6, caption=

陆态网络高程时间序列5 d预测精度指标

, figureFileSmall=null, figureFileBig=null, tableContent=
最大RMSE/mm 平均RMSE/mm 最小RMSE/mm 最大MAE/mm 平均MAE/mm 最小MAE/mm 最大R 平均R 最小R
7.88 5.24 3.56 5.85 3.98 2.67 0.93 0.76 0.55
), ArticleFig(id=1322506227532284749, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 7, caption=

Comparison of accuracy between C-TCN-A method and reference [7]

, figureFileSmall=null, figureFileBig=null, tableContent=
模型测站 C-TCN-A DSANN 提高百分比/%
RMSE/mm MAE/mm RMSE/mm MAE/mm RMSE MAE
XJBC 3.03 2.27 3.9 2.9 22 22
XJKL 5.16 3.67 5.6 4.1 8 10
XJQM 2.83 2.23 4.1 3.1 31 28
XJSH 2.77 2.13 3.7 2.8 25 24
XJZS 2.97 2.34 3.9 3.0 24 22
XJSS 2.67 2.12 3.3 2.4 19 12
), ArticleFig(id=1322506227591005006, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表7, caption=

C-TCN-A方法与文献[7]精度对比

, figureFileSmall=null, figureFileBig=null, tableContent=
模型测站 C-TCN-A DSANN 提高百分比/%
RMSE/mm MAE/mm RMSE/mm MAE/mm RMSE MAE
XJBC 3.03 2.27 3.9 2.9 22 22
XJKL 5.16 3.67 5.6 4.1 8 10
XJQM 2.83 2.23 4.1 3.1 31 28
XJSH 2.77 2.13 3.7 2.8 25 24
XJZS 2.97 2.34 3.9 3.0 24 22
XJSS 2.67 2.12 3.3 2.4 19 12
), ArticleFig(id=1322506227653919567, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=EN, label=Tab. 8, caption=

Comparison of accuracy between C-TCN-A method and reference [10]

, figureFileSmall=null, figureFileBig=null, tableContent=
模型测站 C-TCN-A VMD+XGBoost
RMSE/mm MAE/mm RMSE/mm MAE/mm
BJGB 3.39 2.55 4.32 3.27
XJSS 2.67 2.12 5.54 4.44
HEZL 3.02 2.32 4.08 3.16
), ArticleFig(id=1322506227716834128, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, language=CN, label=表8, caption=

C-TCN-A方法与文献[10]精度对比

, figureFileSmall=null, figureFileBig=null, tableContent=
模型测站 C-TCN-A VMD+XGBoost
RMSE/mm MAE/mm RMSE/mm MAE/mm
BJGB 3.39 2.55 4.32 3.27
XJSS 2.67 2.12 5.54 4.44
HEZL 3.02 2.32 4.08 3.16
)], attaches=null, journal=Journal(id=1281211523964080133, delFlag=0, nameCn=大地测量与地球动力学, nameEn=Journal of Geodesy and Geodynamics, nameHistory1=null, nameHistory2=null, issn=1671-5942, eissn=null, cn=42-1655/P, 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=BeWo6OWOseISH6uPOmc2hA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Geodesy and Geodynamics, journalRemark=null, publicationField=null, createdTime=1783394231074, updatedTime=1788951966472, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=D, firstLetterEn=D, subjectCode=Science, subjectName=null, subjectCodeEn=Science, subjectNameEn=null, picCn=BeWo6OWOseISH6uPOmc2hA==, picEn=YHLfVa8gITbOo0FeIjbLDg==, jcr=null, cjcr=null, exts=[JournalExt(id=1304522356412666698, language=CN, name=大地测量与地球动力学, 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=1788951966622, updatedTime=1788951966622, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.jgg09.com/journalx_ddcl/authorLogOn.action, submissionEditorUrl=http://www.jgg09.com/journalx_ddcl/editorLogOn.action, submissionReviewUrl=http://www.jgg09.com/journalx_ddcl/expertLogOn.action, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1304522356576244555, language=EN, name=Journal of Geodesy and Geodynamics, 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=1788951966661, updatedTime=1788951966661, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.jgg09.com/journalx_ddcl/authorLogOn.action, submissionEditorUrl=http://www.jgg09.com/journalx_ddcl/editorLogOn.action, submissionReviewUrl=http://www.jgg09.com/journalx_ddcl/expertLogOn.action, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1281212831689347082, websiteList=[Website(id=1281216327125799907, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1281212831689347082, 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/ddcl/CN, language=CN, createTime=1783395376235, createBy=18614031015, updateTime=1783395439358, updateBy=18614031015, name=大地测量与地球动力学-中文, tplId=1146099689490845704, title=大地测量与地球动力学, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1281216715371548758, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=articleTextType, value=kx, createTime=1783395468800, updateTime=1783395468800, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715346382931, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=banner, value=null, createTime=1783395468794, updateTime=1783395468794, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715400908889, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=grayFlag, value=0, createTime=1783395468807, updateTime=1783395468807, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715333800018, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=logo, value=https://castjournals.cast.org.cn/joweb/ddcl/CN/file/pic?fileId=87adUdwolRb6UqGo0QjlJQ==, createTime=1783395468791, updateTime=1783395468791, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715421880411, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=minRunFlag, value=0, createTime=1783395468812, updateTime=1783395468812, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715363160149, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/ddcl/CN/file/pic, createTime=1783395468798, updateTime=1783395468798, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715409297498, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=silenceFlag, value=0, createTime=1783395468809, updateTime=1783395468809, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715358965844, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1783395468797, updateTime=1783395468797, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715384131671, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=themeColor, value=null, createTime=1783395468803, updateTime=1783395468803, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216715392520280, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327125799907, code=themeStyle, value=null, createTime=1783395468805, updateTime=1783395468805, creator=18614031015, updator=18614031015)]), Website(id=1281216327713002474, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1281212831689347082, 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/ddcl/EN, language=EN, createTime=1783395376375, createBy=18614031015, updateTime=1783395435863, updateBy=18614031015, name=大地测量与地球动力学-英文, tplId=1146101810881728533, title=Journal of Geodesy and Geodynamics, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1281216680483328066, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=articleTextType, value=kx, createTime=1783395460482, updateTime=1783395460483, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680453967935, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=banner, value=null, createTime=1783395460475, updateTime=1783395460475, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680521076805, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=grayFlag, value=0, createTime=1783395460491, updateTime=1783395460491, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680437190718, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=logo, value=https://castjournals.cast.org.cn/joweb/ddcl/EN/file/pic?fileId=87adUdwolRb6UqGo0QjlJQ==, createTime=1783395460471, updateTime=1783395460471, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680542048327, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=minRunFlag, value=0, createTime=1783395460496, updateTime=1783395460496, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680474939457, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/ddcl/EN/file/pic, createTime=1783395460480, updateTime=1783395460480, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680533659718, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=silenceFlag, value=0, createTime=1783395460494, updateTime=1783395460494, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680462356544, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1783395460477, updateTime=1783395460477, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680495910979, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=themeColor, value=null, createTime=1783395460485, updateTime=1783395460485, creator=18614031015, updator=18614031015), WebsiteProps(id=1281216680508493892, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327713002474, code=themeStyle, value=null, createTime=1783395460488, updateTime=1783395460488, creator=18614031015, updator=18614031015)])], journalTitle=大地测量与地球动力学, weixinUrl=null, journalUrl=http://www.jgg09.com/CN/home, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Geodesy and Geodynamics, journalPhotoCn=BeWo6OWOseISH6uPOmc2hA==, journalPhotoEn=YHLfVa8gITbOo0FeIjbLDg==, journalFirstLetter=D, 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, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/ddcl/CN/10.14075/j.jgg.2024.10.469, detailUrlEn=https://castjournals.cast.org.cn/joweb/ddcl/EN/10.14075/j.jgg.2024.10.469, pdfUrlCn=https://castjournals.cast.org.cn/joweb/ddcl/CN/PDF/10.14075/j.jgg.2024.10.469, pdfUrlEn=https://castjournals.cast.org.cn/joweb/ddcl/EN/PDF/10.14075/j.jgg.2024.10.469, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1755187200000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于CEEMDAN 与TCN-Attention 的陆态网络GNSS高程时间序列多尺度预测
收藏切换
罗亦泳 1 , 占奥文 1, * , 冯小欢 1
大地测量与地球动力学 | 现代大地测量 2025,45(8): 781-790
收起
收藏切换
大地测量与地球动力学 |现代大地测量 2025 , 45 (8) : 781 -790
基于CEEMDAN 与TCN-Attention 的陆态网络GNSS高程时间序列多尺度预测
全屏
[Author(id=1322506218317398808, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ecityyluo@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322506218745217818, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506218317398808, language=EN, stringName=Yiyong LUO, firstName=Yiyong, middleName=null, lastName=LUO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China, bio={"content":"

LUO Yiyong, PhD, professor, majors in surveying and mapping processing, E-mail:

"}, bioImg=null, bioContent=

LUO Yiyong, PhD, professor, majors in surveying and mapping processing, E-mail:

, aboutCorrespAuthor=null), CN=AuthorExt(id=1322506219152065307, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506218317398808, language=CN, stringName=罗亦泳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013, bio={"content":"

罗亦泳,博士,教授,主要研究方向为测绘数据处理,E-mail:

"}, bioImg=null, bioContent=

罗亦泳,博士,教授,主要研究方向为测绘数据处理,E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])]), Author(id=1322506219210785565, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=19870023681@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322506219638604575, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506219210785565, language=EN, stringName=Aowen ZHAN, firstName=Aowen, middleName=null, lastName=ZHAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322506219965760288, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506219210785565, language=CN, stringName=占奥文, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])]), Author(id=1322506220032869154, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, 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=1322506220389384996, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506220032869154, language=EN, stringName=Xiaohuan FENG, firstName=Xiaohuan, middleName=null, lastName=FENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322506220460688165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, authorId=1322506220032869154, language=CN, stringName=冯小欢, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322506216396407572, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, xref=null, ext=[AuthorCompanyExt(id=1322506216597734165, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China), AuthorCompanyExt(id=1322506216610317078, tenantId=1146029695717560320, journalId=1281212831689347082, articleId=1322506210474050302, companyId=1322506216396407572, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013)])])]
罗亦泳1 , 占奥文1, * , 冯小欢1
作者信息
  • 1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013
通讯作者:

占奥文,硕士生,主要研究方向为测绘数据处理,E-mail:

ZHAN Aowen, postgraduate, majors in surveying and mapping processing, E-mail:

作者简介:

罗亦泳,博士,教授,主要研究方向为测绘数据处理,E-mail:

Multi-Scale Prediction of GNSS Elevation Time Series of CMONOC Based on CEEMDAN and TCN-Attention
Yiyong LUO1 , Aowen ZHAN1, * , Xiaohuan FENG1
Affiliations
  • 1 School of Surveying and Geoinformation Engineering, East China University of Technology, 418 Guanglan Road, Nanchang 330013, China
  • LUO Yiyong, PhD, professor, majors in surveying and mapping processing, E-mail:

出版时间: 2025-08-15 doi: 10.14075/j.jgg.2024.10.469
文章导航
收藏切换

提出一种基于自适应噪声完备集合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise, CEEMDAN)和时间卷积网络-注意力机制(temporal convolutional network-attention mechanism, TCN-Attention)算法的多尺度预测模型(简称C-TCN-A),该模型可有效应用于GNSS高程时间序列缺失数据的插补和未来趋势的预测。该模型利用CEEMDAN对时间序列进行多尺度分解,然后基于TCN-Attention对不同尺度分量进行预测和重构得到预测结果。为验证模型的性能,选取12个观测站进行1 d与5 d预测,并与其他多种模型进行对比。结果表明,在1 d预测中,C-TCN-A的RMSE和MAE分别降低35%~40%和36%~41%,相关系数R提高25%~29%;在5 d预测中,C-TCN-A的RMSE和MAE分别降低20%~26%和20%~28%,相关系数R提高26%~33%。为验证模型的普适性,利用C-TCN-A对陆态网络99个观测站进行1 d与5 d预测。结果表明,RMSE和MAE指标总体上结果较好,误差分布集中,大多数误差小于4 mm; 预测精度存在一定的空间分布差异,西北地区效果最佳。

GNSS高程时间序列  /  陆态网络  /  改进经验模态分解  /  时间卷积网络

A multi-scale prediction model(referred to as C-TCN-A) based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN) and temporal convolutional network-attention mechanism(TCN-Attention) algorithms is proposed, which can be effectively applied to missing data imputation and future trend prediction of GNSS elevation time series. The model first employs CEEMDAN for multi-scale decomposition of time series, then utilizes TCN-Attention for prediction and reconstruction of different scale components to obtain final results. To validate the model's performance, 12 observation stations were selected for 1-day and 5-day predictions, with comparisons made against multiple other models. Results demonstrate that in 1-day prediction, C-TCN-A reduces RMSE and MAE by 35%-40% and 36%-41% respectively while improving correlation coefficient R by 25%-29%. For 5-day prediction, it achieves reductions of 20%-26% in RMSE and 20%-28% in MAE, with R increasing by 26%-33%. To verify the model's universality, C-TCN-A was applied to 99 stations from the crustal movement observation network of China(CMONOC) for 1-day and 5-day predictions. Results indicate generally favorable RMSE and MAE metrics with concentrated error distribution, where most errors remain below 4 mm. Spatial analysis reveals regional performance differences, with optimal results achieved in the northwestern region.

GNSS elevation time series  /  CMONOC  /  improved empirical mode decomposition  /  temporal convolutional network
罗亦泳, 占奥文, 冯小欢. 基于CEEMDAN 与TCN-Attention 的陆态网络GNSS高程时间序列多尺度预测. 大地测量与地球动力学, 2025 , 45 (8) : 781 -790 . DOI: 10.14075/j.jgg.2024.10.469
Yiyong LUO, Aowen ZHAN, Xiaohuan FENG. Multi-Scale Prediction of GNSS Elevation Time Series of CMONOC Based on CEEMDAN and TCN-Attention[J]. Journal of Geodesy and Geodynamics, 2025 , 45 (8) : 781 -790 . DOI: 10.14075/j.jgg.2024.10.469
中国大陆构造环境监测网络(简称陆态网络)由260余个连续观测的GNSS基准站构成,积累了近20 a的高精度连续观测数据,为大地测量学、地球动力学等研究提供了重要的数据基础。GNSS基准站的水平位移主要由板块运动主导,其受地球物理因素的影响远小于高程方向,而高程时间序列则因多因素耦合作用呈现更复杂的动态特征[1]。这类序列不仅蕴含板块运动信息,还关联着冰川均衡调整、地表负荷形变等地球物理过程,在地壳运动监测、气候变化研究和工程形变预警等领域具有重要的科学价值。然而,2011—2021年的观测数据显示,陆态网络中99个观测条件较好的基准站平均数据缺失天数达38 d,其余站点数据缺失更为严重[2-3]。数据间断性直接影响时间序列分析的完整性,因此开展GNSS高程时间序列预测研究对数据插补和趋势分析具有重要意义,这也是近年来的研究热点。
近年来,机器学习方法较好地应用于GNSS高程时间序列预测领域,主要包括长短期记忆网络(LSTM)[4]、支持向量机(SVM)[4]、Prophet[5-6]、极限梯度提升(XGBoost)[5]、随机森林(RF)[6]、深度自注意力神经网络(DSANN)[7]、LSTM神经网络[8]等。GNSS时间序列受多种因素影响,具有显著的非平稳性[9],单一模型难以捕捉其中的不稳定波动。为解决这一问题,有学者基于经验模态分解(EMD)、变分模态分解(VMD)结合以上预测方法构建了GNSS时间序列预测方法[10-12],取得了一定效果。但EMD存在较为明显的模态混叠和端点效应,VMD存在分解参数设置难的问题,且上述方法主要利用少数GNSS站点的高程时间序列数据分析预测效果,方法的普适性较差,空间差异分析较少,预测结果的可靠性有待进一步研究。
基于Transformer架构构建的informer模型通过引入ProbSparse自注意力机制来处理长序列数据,具有较好的并行计算能力[13],在时间序列预测中取得了良好的效果,但存在超参数选择要求较高、对噪声鲁棒性不足等问题。时间卷积网络(temporal convolutional network, TCN)通过引入空洞卷积和因果卷积来适应时间序列预测问题,具有可变长输入、高计算效率等优点[14],相较于informer,其具有更加灵活的架构,易于结合其他模型,在多个领域已经表现出了优于CNN、LSTM等其他模型的预测精度。为了克服EMD的端点效应和模态混叠的问题,对EMD进行改进建立了CEEMDAN算法[15],该方法能够有效处理非线性和非平稳信号,显著提高了时间序列的分解精度。针对GNSS高程时间序列的特性,本文提出一种融合CEEMDAN与TCN-Attention的GNSS高程时间序列预测方法C-TCN-A。该方法利用CEEMDAN将时间序列分解为多个本征模态分量(intrinsic mode function,IMF),然后通过计算各分量的样本熵(sample entropy,SE)[16],将IMF分量进行重构,得到新的分量IMFs,以提高模型的计算效率和稳定性; 同时引入注意力机制提高TCN对关键特征的捕捉能力,进一步提升模型的性能; 最后利用TCN-Attention对重构分量IMFs分别进行训练和预测,基于CEEMDAN算法的重构特性,将重构分量的预测结果进行叠加重构,实现GNSS高程时间序列的多尺度预测。
CEEMDAN算法可用于处理非线性和非平稳信号[15]。该方法能够有效减少分解GNSS高程时间序列时可能出现的模态混叠现象,由于采用多次叠加高斯白噪声求平均的方法,经过分解的IMF分量重构后与原始序列的误差接近于0,可实现原始信号的精确重构[17]。CEEMDAN对非线性、非平稳的GNSS高程时间序列进行分解,可挖掘具有物理意义的高程时间序列变化规律,有效提高后续预测模型的精度。为减少建模的复杂性,利用样本熵法分析CEEMDAN分解的IMF分量,将样本熵相近的IMF分量进行融合。
时间卷积网络是一种用于处理时间序列数据的卷积神经网络。通过引入因果卷积、空洞卷积和残差连接等模块,增加网络对于不同长度的时间序列依赖信息的学习能力,解决传统循环神经网络在处理长序列数据时存在的梯度消失或爆炸问题; 另外,由于卷积操作可以并行执行,TCN在处理时间序列时具有显著的计算效率优势。
人的大脑在接收到大量信息时,通常会将注意力集中分配到重要的信息上,而不会把注意力放到其他无关的信息上。基于这种思想,Bahdanau等[18]提出了注意力机制,其核心是动态地为输入序列中的每个元素分配权重,这些权重表示每个输入元素对于当前输出状态的重要程度。本文通过引入注意力机制,为不同时间步的隐藏状态分配不同的权重,使深度学习模型能够关注到时间序列中更重要的信息,改善TCN模型的预测效果。TCN-Attention模型的架构如图 1所示。
基于CEEMDAN-TCN-Attention的GNSS高程时间序列多尺度预测模型的具体步骤如下:
1) 数据预处理(剔除异常值和离群值)。采用Z-core方法确定数据是否属于异常值(通常认为Z-score绝对值大于3的数据为异常值),如判定为异常值,用序列的平均值进行代替,输入模型前,对数据进行归一化,以加快模型的收敛速度。
2) CEEMDAN分解。对预处理后的时间序列进行CEEMDAN分解,得到m个IMF分量。
3) 样本熵计算。计算各个分量的样本熵,将样本熵相近的分量进行叠加,重构得到新的n个IMFs分量。
4) 构建数据集并进行预测。n个IMFs分别建立样本集,首先按照8∶2的比例将其划分为训练集和测试集,然后构建滑动窗口样本集,滑动窗口通过试算确定(1 d预测滑动窗口为7,5 d预测滑动窗口为10)。将滑动窗口样本集依次输入TCN-Attention模型中进行训练,然后将测试集输入到训练好的模型中进行预测。
5) 将各个分量的预测序列进行叠加,得到的序列即为预测结果,最后对模型进行精度评估。
建模流程如图 2所示。
以陆态网络HRBN站高程时间序列提前1 d预测为例进行建模。选取2011—2020年GNSS高程时间序列共有3 500个历元数据,前2 800个作为训练集,后700个作为测试集,GNSS高程原始数据如图 3所示。利用CEEMDAN对时间序列进行分解,得到10个IMF分量,分量按照频率由高到低排列(图 3)。由图 3可知,前3个IMF分量为高频分量,包含显著的噪声信息,样本熵数值明显高于其他分量,分别作为IMFs1、IMFs2、IMFs3。IMF4、IMF5的样本熵较为接近,数据复杂性和周期性较为接近,相加重构为IMFs4。IMF6~IMF10熵值接近并且明显小于其他分量熵值,主要包括明显的长周期、趋势信息,相加重构为IMFs5。基于数据特征重构IMFs1~IMFs5,测试集的重构数据如图 3红色曲线所示。对IMFs分别建立训练样本和测试样本(前7 d为输入,第8天为输出),基于TCN-Attention模型对训练集进行学习和预测测试集,将各IMFs的预测结果相加构建GNSS高程时间序列多尺度预测新模型,各IMFs预测结果和多尺度预测结果如图 4蓝色曲线所示。样本集和TCN-Attention模型具体参数后续详细罗列。由图 4可知,C-TCN-A模型能较好地挖掘、学习和预测IMFs,最终较好地预测出GNSS高程时间序列。
为进一步分析C-TCN-A方法的有效性,统计2011—2020年HRBN站高程解算中误差(简称sigu)[19]和高频分量(IMFs1+IMFs2+IMFs3)的绝对值(图 5(a))。由图 5(a)可知,sigu表现出明显的年周期,主要在5~12.5 mm波动; 高频分量绝对值与sigu具有相似的周期,波动范围主要集中在1~12.5 mm,少部分历元的高频分量绝对值分布在12.5~20 mm。GNSS高程时间序列中包含的随机噪声和其他不确定信息主要集中在高频分量,同时也可能隐含频率较高且有规律的有用信息。基于CEEMDAN与TCN-Attention建立高程时间序列多尺度预测模型,利用CEEMDAN对高程时间序列进行分解,提取高频分量,并单独对高频分量进行学习和预测,可提取高频分量中隐含的高频变形信息,降低其他相对低频分量对高频有用信息预测的干扰,进而提高最终预测效果。为分析该方法的有效性,绘制高频分量预测残差与去除高频分量预测残差的绝对值曲线(图 5(b))。由图 5(b)可知,相较于去除高频分量,对高频分量进行训练预测所得的残差值明显降低,表明TCN-Attention较好地学习和预测了高频分量中隐含的有用信息。
为验证C-TCN-A模型对GNSS高程时间序列的预测能力,分别对陆态网络GNSS高程时间序列进行1 d和5 d的预测。首先从陆态网络不同区域均匀选取12个GNSS测站,将新方法预测结果与常用的XGboost、CNN、LSTM等方法进行比较分析,各模型核心参数如表 1所示。为更好地验证C-TCN-A模型的稳定性和有效性,基于C-TCN-A对陆态网络99个站点进行预测,并统计误差和空间分布特征。
陆态网络的GNSS高程时间序列数据来自中国地震局GNSS数据产品服务平台。中国地震局采用10.4版本的GAMIT/GLOBK软件,对所有的陆态网络基准站进行单日松弛解处理,再进行七参数相似变换将其转入特定的参考框架,得到GNSS高程时间序列,同时获得高程时间序列解算的中误差,有助于分析预测结果误差分布特征。本文根据纬度从高到低划分陆态网络,考虑到观测时长和数据缺失情况,选取2011—2020年观测条件较好的99个GNSS基准站高程时间序列为研究对象,共有3 500个历元数据,取前2 800个作为训练集,后700个作为测试集,基准站分布如图 6所示(红色测站为均匀选取的12个用于模型对比的GNSS站)。受地震、火山喷发、仪器设备更换和升级等因素的影响,GNSS坐标时间序列会产生阶跃或震后弛豫位移现象[1, 9]。99个基准站中,有10个测站产生了明显的坐标阶跃现象,会对后续的数据处理和分析产生影响,需要对阶跃值进行校正。采用Wu等[20]研发的TSAnalyzer软件校正GNSS时间序列的阶跃。
利用表 1中7种模型分别对陆态网络中均匀选取的12个GNSS站高程时间序列进行预测,并对预测结果进行对比分析。
首先建立样本集,针对每个IMFs建立样本集$\left\{\left(x_l, y_l\right), l=1, 2, \cdots n\right\}$,其中xl=[IMFsj-d, …, IMFsj-1]为样本输入,通过试算确定样本输入维数d=7,yl=IMFsj为样本输出,n为样本数量。2011—2020年每个站共有3 500个历元数据,前2 800个作为训练集,后700个作为测试集。使用上述7种模型对训练集进行学习以确定各模型超参数,并对测试集进行预测,结果如图 7所示,为更直观地展示各模型的预测效果,将预测结果的部分区域放大(图中矩形框)。由图 7可知,7种方法均能很好地预测GNSS高程时间序列的变化趋势,且C-TCN-A对GNSS高程时间序列的预测效果显著优于其他模型。12个GNSS高程时间序列预测结果表明,C-TCN-A模型利用CEEMDAN有效地挖掘了高程时间序列的变化规律,能够较好地从多尺度上对这些变化特征进行学习和预测,预测效果显著提高。
统计各模型对各测站测试集预测结果的RMSE、MAE、R和t指标(表 2)。由表 2可知,C-TCN-A的RMSE、MAE、R指标均明显优于其他6种方法。由于C-TCN-A模型采用了多种方法进行有机融合,其预测消耗的时间相对有所增加,但这不会对实际应用产生显著影响。
统计各模型对12个测站预测效果的平均值和C-TCN-A相对于其他模型预测效果的提升情况(表 3)。由表 3可知,C-TCN-A各项精度指标均优于其他方法,其RMSE相对于其他方法改善35%~40%,MAE改善36%~41%,R提高25%~29%。
为验证新方法对不同时长高程时间序列的预测能力,基于上述7种模型对GNSS高程时间序列进行5 d预测,并统计12个GNSS站各模型预测结果的平均精度指标(表 4)。
对比表 3和表 4可知,随着预测时间跨度的增加,各模型的预测精度均有所降低。但C-TCN-A的RMSE、MAE、R仍然明显优于其他方法,其中RMSE较其他方法降低20%~26%,MAE降低20%~28%,R提高26%~33%。
为进一步验证新方法的普适性,将新模型应用于陆态网络99个观测条件较好的GNSS测站点。利用RMSE、MAE、R等精度指标评价模型预测效果,表 5和表 6为陆态网络高程时间序列1 d与5 d预测精度指标统计结果。
为直观地展示新方法的预测结果,绘制GNSS高程时间序列预测1 d RMSE和MAE的空间分布(图 8)。由图 8(a)和8(b)可知,本文模型适用于各个测站,具有较强的稳定性,纬度约在35°以上的测站预测精度多呈现蓝紫色,即RMSE处于2.2~3.6 mm,MAE处于1.6~2.6 mm; 纬度在35°以下的测站预测精度多呈现绿色,即RMSE处于3.6~4.8 mm,MAE处于2.6~3.6 mm; 少部分测站预测精度呈现橘红色,即RMSE处于5 mm以上,MAE处于3.6 mm以上。中国西北地区预测精度最高,北方地区绝大部分区域精度低于西北地区,但是高于南方地区和青藏地区,精度相对较低的区域为西藏和云南部分区域。为了更好地分析中国陆态网络GNSS高程时间序列的预测结果,绘制误差分布直方图(图 9)。由图 9(a)和9(b)可知,RMSE值普遍分布于2.2~4.2 mm,MAE值普遍分布于1.6~3.6 mm,预测精度较为稳定。此外,绘制各历元误差直方分布图(图 9(c)),可以看出,模型预测误差(绝对值)普遍分布于0~4 mm,占全部数据集的80%,少量分布于4~8 mm,占全部数据集的16%,其他异常值共占4%。图 9(d)为预测值与观测值的散点分布图,可以看出,预测值拟合直线与y=x直线吻合度很好,并且散点相对于y=x分布集中和密集,单独的离散点很少,表明预测模型具有很好的精度和可靠性。
5 d预测误差的空间分布和1 d预测误差分布较为一致。为了定量分析新方法的5 d预测效果,统计测站5 d预测误差分布情况,发现模型预测RMSE普遍位于3.4~5.4 mm,MAE普遍位于2.6~4.2 mm,历元预测误差(绝对值)普遍分布于4 mm以内,占比76%。拟合预测值与真实值散点,散点集中分布在y=x直线附近,拟合直线斜率为0.921。
为了更好地验证C-TCN-A方法的优势,将该方法与文献[7]和[10]提出的方法进行比较。文献[7]对陆态网络新疆地区的10个GNSS站高程时间序列进行预测,其中XJBC、XJKL、XJQM、XJSH、XJZS、XJSS等6个站与本文站点相同。因此,将C-TCN-A方法与文献[7]提出的DSANN算法的预测精度进行对比,如表 7所示。由表 7可知,C-TCN-A方法较DSANN方法RMSE约改善8%~31%,MAE约改善10%~28%。表明本文方法能有效挖掘信息,同时避免不同尺度信息对预测模型的干扰。
文献[10]利用VMD算法对GNSS高程时间序列进行预处理,并在此基础上利用深度学习算法进行预测,最终建立GNSS高程时间序列预测模型(VMD-XGBoost),较传统方法预测精度得到较好的提升。为了对比C-TCN-A方法与VMD-XGBoost方法的预测效果,选取3个共同GNSS站(BJGB、XJSS、HEZL)进行对比,精度指标如表 8所示。VMD-XGBoost方法由于VMD自身具有去噪功能,利用VMD对时间序列进行分解和重构可能导致GNSS高频信息中的有用信息丢失,并且VMD核心参数的确定会对预测结果产生显著影响,最终影响GNSS高程时间序列预测效果。同时,待预测站周围的若干GNSS站选取可能受到站间距离、环境差异、空间分布等影响,导致应用效果的差异。由表 8可知,C-TCN-A的RMSE、MAE指标明显优于VMD-XGBoost,验证了该方法的有效性。
1) C-TCN-A对GNSS基准站高程时间序列1 d预测具有较高的预测精度。分析12个测站的预测效果可知,该方法相比XGboost、CNN、LSTM、CNN-LSTM、informer、TCN等方法,RMSE改善35%~40%,MAE改善36%~41%,R提高25%~29%,能更好地预测高程时间序列的快速变化; 与DSANN和VMD-XGBoost方法相比,该方法的RMSE、MAE指标也改善显著。此外,为验证C-TCN-A的普适性,基于C-TCN-A对陆态网络99个GNSS站的高程时间序列进行预测,并统计误差分布特征。结果表明,新方法总体效果较好,不同区域的预测误差存在差异,西北区域预测效果最佳,西藏南部、云南部分区域和广东部分区域的预测误差相对较大。C-TCN-A方法的RMSE普遍位于2.2~4.2 mm之间,MAE普遍位于1.6~3.6 mm之间。误差分布集中,表明C-TCN-A方法具有很好的可靠性和有效性。
2) C-TCN-A对陆态网络进行5 d预测仍然具有较好的效果。分析12个测站的预测结果可知,C-TCN-A模型较其他方法RMSE降低20%~26%,MAE降低20%~28%,R提升26%~33%。陆态网络5 d预测结果误差的空间分布与1 d预测相似,总体上具有较好的预测效果。统计分析陆态网络99个站点的预测结果可知,C-TCN-A在5 d预测中,模型预测RMSE普遍为3.4~5.4 mm,MAE普遍为2.6~4.2 mm; 误差分布集中,76%的误差小于4 mm。
  • 国家自然科学基金(41861058)
参考文献 引证文献
排序方式:
1
明锋, 杨元喜, 曾安敏, 等. 中国区域IGS站高程时间序列季节性信号及长期趋势分析[J]. 中国科学: 地球科学, 2016, 46 (6): 834- 844
Ming Feng , Yang Yuanxi , Zeng Anmin , et al. Analysis of Seasonal Signals and Long-Term Trends in the Height Time Series of IGS Sites in China[J]. Scientia Sinica Terrae, 2016, 46 (6): 834- 844
2
Li Z , Lu T D , Yu K G , et al. Interpolation of GNSS Position Time Series Using GBDT, XGBoost, and RF Machine Learning Algorithms and Models Error Analysis[J]. Remote Sensing, 2023, 15 (18) http://www.keyanzhidian.com/doc/detail?id=2074000763
3
Crocetti L , Schartner M , Soja B . Discontinuity Detection in GNSS Station Coordinate Time Series Using Machine Learning[J]. Remote Sensing, 2021, 13 (19)
4
Gao W Z , Li Z , Chen Q S , et al. Modelling and Prediction of GNSS Time Series Using GBDT, LSTM and SVM Machine Learning Approaches[J]. Journal of Geodesy, 2022, 96 (10)
5
鲁铁定, 李祯. 基于Prophet-XGBoost模型的GNSS高程时间序列预测[J]. 大地测量与地球动力学, 2022, 42 (9): 898- 903
Lu Tieding , Li Zhen . Prediction of GNSS Vertical Coordinate Time Series Based on Prophet-XGBoost Model[J]. Journal of Geodesy and Geodynamics, 2022, 42 (9): 898- 903
6
李威, 鲁铁定, 贺小星, 等. 基于Prophet-RF模型的GNSS高程坐标时间序列预测分析[J]. 大地测量与地球动力学, 2021, 41 (2): 116- 121
Li Wei , Lu Tieding , He Xiaoxing , et al. Prediction and Analysis of GNSS Vertical Coordinate Time Series Based on Prophet-RF Model[J]. Journal of Geodesy and Geodynamics, 2021, 41 (2): 116- 121
7
Jiang W P , Wang J , Li Z , et al. A New Deep Self-Attention Neural Network for GNSS Coordinate Time Series Prediction[J]. GPS Solutions, 2023, 28 (1) http://link.springer.com/doi/10.1007/s10291-023-01544-z
8
Wang J , Jiang W P , Li Z , et al. A New Multi-Scale Sliding Window LSTM Framework(MSSW-LSTM): A Case Study for GNSS Time-Series Prediction[J]. Remote Sensing, 2021, 13 (16)
9
姜卫平, 王锴华, 李昭, 等. GNSS坐标时间序列分析理论与方法及展望[J]. 武汉大学学报: 信息科学版, 2018, 43 (12): 2112- 2123
Jiang Weiping , Wang Kaihua , Li Zhao , et al. Prospect and Theory of GNSS Coordinate Time Series Analysis[J]. Geomatics and Information Science of Wuhan University, 2018, 43 (12): 2112- 2123
10
鲁铁定, 李祯, 贺小星, 等. 融合VMD和XGBoost算法的GNSS高程时间序列预测方法[J]. 测绘学报, 2023, 52 (8): 1235- 1244
Lu Tieding , Li Zhen , He Xiaoxing , et al. GNSS Vertical Time Series Prediction Method Integrating VMD and XGBoost Algorithms[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52 (8): 1235- 1244
11
Chen H K , Lu T D , Huang J H , et al. An Improved VMD-LSTM Model for Time-Varying GNSS Time Series Prediction with Temporally Correlated Noise[J]. Remote Sensing, 2023, 15 (14) http://openurl.ebsco.com/contentitem/doi:10.3390%2Frs15143694?sid=ebsco:plink:crawler&id=ebsco:doi:10.3390%2Frs15143694
12
符娇, 宋圆, 段炬奎, 等. 基于EMD-Prophet模型的GPS坐标时间序列预测[J]. 安徽地质, 2022, 32 (增2): 174- 177
Fu Jiao , Song Yuan , Duan Jukui , et al. Prediction of GPS Coordinate Time Series Based on EMD-Prophet Model[J]. Geology of Anhui, 2022, 32 (S2): 174- 177
13
Zhou H Y , Zhang S H , Peng J Q , et al. Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2021, 35 (12): 11106- 11115 http://www.xueshufan.com/publication/3177318507
14
Bai S J, Zico Kolter J, Koltun V. An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling[EB/OL]. https://arxiv.org/abs/1803.01271, 2018-04-19
15
Torres M E, Colominas M A, Schlotthauer G, et al. A Complete Ensemble Empirical Mode Decomposition with Adaptive Noise[C]. IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP), Prague, 2011
16
Sun X Y , Liu H Z . Multivariate Short-Term Wind Speed Prediction Based on PSO-VMD-SE-ICEEMDAN Two-Stage Decomposition and Att-S2S[J]. Energy, 2024, 305 http://www.sciencedirect.com/science/article/pii/S0360544224020024
17
张文乾, 刘金凤, 江军, 等. 基于CEEMDAN和TCN的变压器油中溶解气体含量预测[J]. 电力工程技术, 2024, 43 (3): 192- 200
Zhang Wenqian , Liu Jinfeng , Jiang Jun , et al. Prediction of Concentration for Dissolved Gas in Oil Based on CEEMDAN and TCN[J]. Electric Power Engineering Technology, 2024, 43 (3): 192- 200
18
Bahdanau D, Cho K, Bengio Y. Neural Machine Translation by Jointly Learning to Align and Translate[EB/OL]. https://arxiv.org/abs/1409.0473, 2014-05-19
19
周建营, 陈国恒, 刘文建. GAMIT/GLOBK多系统GNSS数据处理分析[J]. 导航定位学报, 2020, 8 (5): 85- 91
Zhou Jianying , Chen Guoheng , Liu Wenjian . GAMIT/GLOBK Multi-System GNSS Data Processing Analysis[J]. Journal of Navigation and Positioning, 2020, 8 (5): 85- 91
20
Wu D C , Yan H M , Shen Y C . TSAnalyzer, a GNSS Time Series Analysis Software[J]. GPS Solutions, 2017, 21 (3): 1389- 1394 http://www.keyanzhidian.com/doc/detail?id=2031684016
2025年第45卷第8期
36
14
引用本文
BibTeX
文章信息
doi: 10.14075/j.jgg.2024.10.469
  • 接收时间:2024-10-12
  • 首发时间:2026-10-29
  • 出版时间:2025-08-15
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-10-12
基金
National Natural Science Foundation of China(41861058)
国家自然科学基金(41861058)
作者信息
    1 东华理工大学测绘与空间信息工程学院,南昌市广兰大道418号,330013

通讯作者:

占奥文,硕士生,主要研究方向为测绘数据处理,E-mail:

ZHAN Aowen, postgraduate, majors in surveying and mapping processing, E-mail:

参考文献
分享链接
https://castjournals.cast.org.cn/joweb/ddcl/CN/10.14075/j.jgg.2024.10.469
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
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
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