Article(id=1239165050607423722, tenantId=1146029695717560320, journalId=1238841944844054536, issueId=1239165048212476136, articleNumber=null, orderNo=null, doi=10.12347/j.ycyk.20250616001, pmid=null, cstr=32406.14.ycyk.20250616001, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1750003200000, receivedDateStr=2025-06-16, revisedDate=1753286400000, revisedDateStr=2025-07-24, acceptedDate=null, acceptedDateStr=null, onlineDate=1773369570637, onlineDateStr=2026-03-13, pubDate=1757865600000, pubDateStr=2025-09-15, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1773369570637, onlineIssueDateStr=2026-03-13, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1773369570637, creator=13701087609, updateTime=1773369570637, updator=13701087609, issue=Issue{id=1239165048212476136, tenantId=1146029695717560320, journalId=1238841944844054536, year='2025', volume='46', issue='5', pageStart='1', pageEnd='158', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1773369570067, creator=13701087609, updateTime=1773617788336, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1240206151170707811, tenantId=1146029695717560320, journalId=1238841944844054536, issueId=1239165048212476136, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1240206151170707812, tenantId=1146029695717560320, journalId=1238841944844054536, issueId=1239165048212476136, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=36, endPage=44, ext={EN=ArticleExt(id=1239165052650049779, articleId=1239165050607423722, tenantId=1146029695717560320, journalId=1238841944844054536, language=EN, title=Dual-Branch Neural Network for GNSS Localization Prediction in GNSS-Denied Environments, columnId=1239165052184482030, journalTitle=Journal of Telemetry, Tracking and Command, columnName=Navigation Technology Column, runingTitle=null, highlight=null, articleAbstract=

Inertial/Global Navigation Satellite System (GNSS) integrated navigation has been widely applied in various mobile platforms such as unmanned aerial vehicles (UAVs). However, during GNSS signal outages, INS errors accumulate rapidly, severely degrading navigation accuracy. Existing research primarily focuses on horizontal two-dimensional error modeling while neglecting the dynamic characteristics in the vertical (altitude) direction, limiting its practical application in three-dimensional space. To address this issue, this paper proposes a dual-branch neural network model for three-dimensional navigation, which simultaneously models position increments in the longitude, latitude, and altitude directions to cater to the demands of dynamic navigation in 3D space. The model adopts a decoupled dual-branch structure built with LSTM and GRU networks, designing separate modeling paths for the horizontal and vertical components. A convolutional neural network (CNN) is further incorporated into the main branch to enhance temporal feature extraction. Experimental results demonstrate that the proposed network significantly improves three-dimensional navigation accuracy. Compared with conventional positioning methods, it reduces the root mean square error(RMSE) along the east, north, and up axes by 97.8 %, 97.9 %, and 26.2 %, respectively, demonstrating its strong potential for practical deployment.

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Yini YIN, Xun SUN, Lin JIANG, Zhao HE, Zhiying WANG), CN=ArticleExt(id=1239165065044218337, articleId=1239165050607423722, tenantId=1146029695717560320, journalId=1238841944844054536, language=CN, title=卫星拒止条件下基于双分支神经网络模型的GNSS组合定位算法, columnId=1239165052356448497, journalTitle=遥测遥控, columnName=导航技术专栏, runingTitle=null, highlight=null, articleAbstract=

惯性导航/卫星导航的组合导航广泛应用于无人机等多种移动载体中,然而在全球导航卫星系统信号中断期间,惯性导航系统误差会迅速积累,严重影响导航精度。现有研究多集中于对水平二维的误差建模,忽略了高度方向的动态特性,限制了其实际应用。为解决这一问题,本文提出一种面向三维导航的双分支神经网络模型,同时建模经度、纬度与高度方向的位置增量,以适应三维空间中的动态导航需求。该模型采用LSTM(长短期记忆网络)与GRU(门控循环单元)构建解耦的双分支结构,分别针对经纬面与高度方向设计建模路径,并在主分支中引入卷积神经网络(CNN)以增强时序特征提取能力。实验表明:所提出的网络提升了导航系统在三维空间下的预测精度。相较于传统的定位方法,本文提出的方法在东北天三向的均方根误差分别降低了97.8%、97.9%、26.2%,具备良好的工程应用潜力。

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=n2+zXuNQPK6E0ST470PC8Q==, magXml=31g0J61JzuGuq1VkVaiUtg==, pdfUrl=null, pdf=w0bTJVeo1xiQShZbQVxXSA==, pdfFileSize=3108899, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=a7OnOTMQtKn+HaBP17eGcQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=9v9jooa2dxBqY6rKXurqvA==, mapNumber=null, authorCompany=null, fund=null, authors=

尹旖旎 2000年生,硕士研究生。

孙迅 1974年生,硕士,研究员。

蒋林 1979年生,本科,研究员。

何兆 2000年生,硕士研究生。

王志颖 1998年生,硕士研究生。

, authorsList=尹旖旎, 孙迅, 蒋林, 何兆, 王志颖)}, authors=[Author(id=1239265683469562027, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1239265683591196852, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265683469562027, language=EN, stringName=Yini YIN, firstName=Yini, middleName=null, lastName=YIN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Beijing Research Institute of Telemetry, Beijing 100076, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1239265683700248766, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265683469562027, language=CN, stringName=尹旖旎, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=北京遥测技术研究所 北京 100076, bio={"content":"

尹旖旎 2000年生,硕士研究生。

"}, bioImg=null, bioContent=

尹旖旎 2000年生,硕士研究生。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1239265683331149989, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, xref=null, ext=[AuthorCompanyExt(id=1239265683343732902, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Beijing Research Institute of Telemetry, Beijing 100076, China), AuthorCompanyExt(id=1239265683352121511, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京遥测技术研究所 北京 100076)])]), Author(id=1239265683771551938, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1239265683876409551, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265683771551938, language=EN, stringName=Xun SUN, firstName=Xun, middleName=null, lastName=SUN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Beijing Research Institute of Telemetry, Beijing 100076, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1239265683998044375, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265683771551938, language=CN, stringName=孙迅, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=北京遥测技术研究所 北京 100076, bio={"content":"

孙迅 1974年生,硕士,研究员。

"}, bioImg=null, bioContent=

孙迅 1974年生,硕士,研究员。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1239265683331149989, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, xref=null, ext=[AuthorCompanyExt(id=1239265683343732902, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Beijing Research Institute of Telemetry, Beijing 100076, China), AuthorCompanyExt(id=1239265683352121511, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京遥测技术研究所 北京 100076)])]), Author(id=1239265684123873505, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, 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=1239265684274868463, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265684123873505, language=EN, stringName=Lin JIANG, firstName=Lin, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Beijing Research Institute of Telemetry, Beijing 100076, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1239265684367143158, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265684123873505, language=CN, stringName=蒋林, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=北京遥测技术研究所 北京 100076, bio={"content":"

蒋林 1979年生,本科,研究员。

"}, bioImg=null, bioContent=

蒋林 1979年生,本科,研究员。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1239265683331149989, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, xref=null, ext=[AuthorCompanyExt(id=1239265683343732902, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Beijing Research Institute of Telemetry, Beijing 100076, China), AuthorCompanyExt(id=1239265683352121511, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京遥测技术研究所 北京 100076)])]), Author(id=1239265684497166593, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1239265684581052681, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265684497166593, language=EN, stringName=Zhao HE, firstName=Zhao, middleName=null, lastName=HE, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Beijing Research Institute of Telemetry, Beijing 100076, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1239265684685910288, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265684497166593, language=CN, stringName=何兆, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=北京遥测技术研究所 北京 100076, bio={"content":"

何兆 2000年生,硕士研究生。

"}, bioImg=null, bioContent=

何兆 2000年生,硕士研究生。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1239265683331149989, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, xref=null, ext=[AuthorCompanyExt(id=1239265683343732902, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Beijing Research Institute of Telemetry, Beijing 100076, China), AuthorCompanyExt(id=1239265683352121511, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京遥测技术研究所 北京 100076)])]), Author(id=1239265684794962202, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1239265684887236895, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265684794962202, language=EN, stringName=Zhiying WANG, firstName=Zhiying, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Beijing Research Institute of Telemetry, Beijing 100076, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1239265685004677418, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, authorId=1239265684794962202, language=CN, stringName=王志颖, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=北京遥测技术研究所 北京 100076, bio={"content":"

王志颖 1998年生,硕士研究生。

"}, bioImg=null, bioContent=

王志颖 1998年生,硕士研究生。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1239265683331149989, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, xref=null, ext=[AuthorCompanyExt(id=1239265683343732902, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Beijing Research Institute of Telemetry, Beijing 100076, China), AuthorCompanyExt(id=1239265683352121511, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京遥测技术研究所 北京 100076)])])], keywords=[Keyword(id=1239265685176643895, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, orderNo=1, keyword=Unmanned aerial vehicle), Keyword(id=1239265685277307196, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, orderNo=2, keyword=GNSS denial), Keyword(id=1239265685352804676, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, orderNo=3, keyword=Integrated navigation), Keyword(id=1239265685440885068, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, orderNo=4, keyword=3D positioning), Keyword(id=1239265685579297107, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, orderNo=5, keyword=Dualbranch neural network), Keyword(id=1239265685688349016, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, orderNo=6, keyword=Inertial navigation), Keyword(id=1239265687189909853, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, orderNo=1, keyword=无人机), Keyword(id=1239265687261213028, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, orderNo=2, keyword=全球导航卫星系统), Keyword(id=1239265687340904813, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, orderNo=3, keyword=组合导航), Keyword(id=1239265687424790897, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, orderNo=4, keyword=三维定位), Keyword(id=1239265687554814327, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, orderNo=5, keyword=双分支神经网络), Keyword(id=1239265687663866237, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, orderNo=6, keyword=惯性导航)], refs=[Reference(id=1239265690046231050, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2007, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=HOFMANN-WELLENHOF B, LICHTENEGGER H, WASLE E, journalName=GNSS-global navigation satellite systems: GPS, GLONASS, Galileo, and more, refType=null, unstructuredReference=HOFMANN-WELLENHOF B, LICHTENEGGER H,WASLE E. GNSS-global navigation satellite systems: GPS, GLONASS, Galileo, and more[M]. Vienna:Springer, 2007., articleTitle=null, refAbstract=null), Reference(id=1239265690142700044, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=BROWN R G, HWANG P Y C, journalName=Introduction to random signals and applied Kalman filtering: with MATLAB exercises, refType=null, unstructuredReference=BROWN R G, HWANG P Y C. Introduction to random signals and applied Kalman filtering: with MATLAB exercises[M]. 4th ed. Hoboken: John Wiley & Sons, 2012., articleTitle=null, refAbstract=null), Reference(id=1239265690226586132, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2007, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=GREWAL M S, WEILL L R, ANDREWS A P, journalName=Global positioning systems, inertial navigation, and integration, refType=null, unstructuredReference=GREWAL M S, WEILL L R, ANDREWS A P. Global positioning systems, inertial navigation, and integration[M]. Hoboken: John Wiley & Sons, 2007., articleTitle=null, refAbstract=null), Reference(id=1239265691702981146, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2023, volume=11, issue=null, pageStart=79033, pageEnd=79055, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=HE Y, LI J, LIU J, journalName=IEEE Access, refType=null, unstructuredReference=HE Y, LI J, LIU J. Research on GNSS INS & GNSS/INS integrated navigation method for autonomous vehicles: A survey[J]. IEEE Access, 2023, 11: 79033-79055., articleTitle=Research on GNSS INS & GNSS/INS integrated navigation method for autonomous vehicles: A survey, refAbstract=null), Reference(id=1239265691812033054, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2023, volume=15, issue=17, pageStart=4125, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=4, authorNames=YIN Z, YANG J, MA Y, journalName=Remote Sensing, refType=null, unstructuredReference=YIN Z, YANG J, MA Y, et al. A robust adaptive extended Kalman filter based on an improved measurement noise covariance matrix for the monitoring and isolation of abnormal disturbances in GNSS/INS vehicle navigation[J]. Remote Sensing, 2023, 15(17): 4125., articleTitle=A robust adaptive extended Kalman filter based on an improved measurement noise covariance matrix for the monitoring and isolation of abnormal disturbances in GNSS/INS vehicle navigation, refAbstract=null), Reference(id=1239265691950445091, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2019, volume=69, issue=7, pageStart=5138, pageEnd=5148, url=null, language=null, rfNumber=[6], rfOrder=5, authorNames=WANG D, DONG Y, LI Z, journalName=IEEE Transactions on Instrumentation and Measurement, refType=null, unstructuredReference=WANG D, DONG Y, LI Z, et al. Constrained MEMS-based GNSS/INS tightly coupled system with robust Kalman filter for accurate land vehicular navigation[J]. IEEE Transactions on Instrumentation and Measurement, 2019,69(7): 5138-5148., articleTitle=Constrained MEMS-based GNSS/INS tightly coupled system with robust Kalman filter for accurate land vehicular navigation, refAbstract=null), Reference(id=1239265692042719782, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2021, volume=25, issue=1, pageStart=1, pageEnd=15, url=null, language=null, rfNumber=[7], rfOrder=6, authorNames=JIANG H, SHI C, LI T, journalName=GPS Solutions, refType=null, unstructuredReference=JIANG H, SHI C, LI T, et al. Low-cost GPS/INS integration with accurate measurement modeling using an extended state observer[J]. GPS Solutions, 2021, 25(1):1-15., articleTitle=Low-cost GPS/INS integration with accurate measurement modeling using an extended state observer, refAbstract=null), Reference(id=1239265692164354604, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=68, issue=1, pageStart=499, pageEnd=508, url=null, language=null, rfNumber=[8], rfOrder=7, authorNames=SHEN C, ZHANG Y, GUO X, journalName=IEEE Transactions on Industrial Electronics, refType=null, unstructuredReference=SHEN C, ZHANG Y, GUO X, et al. Seamless GPS/inertial navigation system based on self-learning square-root cubature Kalman filter[J]. IEEE Transactions on Industrial Electronics, 2020, 68(1): 499-508., articleTitle=Seamless GPS/inertial navigation system based on self-learning square-root cubature Kalman filter, refAbstract=null), Reference(id=1239265692273406513, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=8, issue=null, pageStart=53984, pageEnd=53996, url=null, language=null, rfNumber=[9], rfOrder=8, authorNames=LI D, JIA X, ZHAO J, journalName=IEEE Access, refType=null, unstructuredReference=LI D, JIA X, ZHAO J. A novel hybrid fusion algorithm for low-cost GPS/INS integrated navigation system during GPS outages[J]. IEEE Access, 2020, 8: 53984-53996., articleTitle=A novel hybrid fusion algorithm for low-cost GPS/INS integrated navigation system during GPS outages, refAbstract=null), Reference(id=1239265692403429945, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=7, pageStart=4475, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=9, authorNames=JWO D J, BISWAL A, MIR I A, journalName=Applied Sciences, refType=null, unstructuredReference=JWO D J, BISWAL A, MIR I A. Artificial neural networks for navigation systems: A review of recent research[J]. Applied Sciences, 2023, 13(7): 4475., articleTitle=Artificial neural networks for navigation systems: A review of recent research, refAbstract=null), Reference(id=1239265692512481856, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=24, issue=4, pageStart=108, pageEnd=null, url=null, language=null, rfNumber=[11], rfOrder=10, authorNames=ABDOLKARIMI E S, MOSAVI M R, journalName=GPS Solutions, refType=null, unstructuredReference=ABDOLKARIMI E S, MOSAVI M R. A low-cost integrated MEMS-based INS/GPS vehicle navigation system with challenging conditions based on an optimized IT2FNN in occluded environments[J]. GPS Solutions,2020, 24(4): 108., articleTitle=A low-cost integrated MEMS-based INS/GPS vehicle navigation system with challenging conditions based on an optimized IT2FNN in occluded environments, refAbstract=null), Reference(id=1239265692596367938, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=16, issue=2, pageStart=334, pageEnd=340, url=null, language=null, rfNumber=[12], rfOrder=11, authorNames=DAI H, BIAN H, WANG R, journalName=Defence Technology, refType=null, unstructuredReference=DAI H, BIAN H, WANG R, et al. An INS/GNSS integrated navigation in GNSS denied environment using recurrent neural network[J]. Defence Technology, 2020, 16(2): 334-340., articleTitle=An INS/GNSS integrated navigation in GNSS denied environment using recurrent neural network, refAbstract=null), Reference(id=1239265692701225545, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2021, volume=168, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[13], rfOrder=12, authorNames=AL BITAR N, GAVRILOV A, journalName=Measurement, refType=null, unstructuredReference=AL BITAR N, GAVRILOV A. A new method for compensating the errors of integrated navigation systems using artificial neural networks[J]. Measurement, 2021,168: 108391., articleTitle=A new method for compensating the errors of integrated navigation systems using artificial neural networks, refAbstract=null), Reference(id=1239265692772528714, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=9, pageStart=1769, pageEnd=null, url=null, language=null, rfNumber=[14], rfOrder=13, authorNames=YUE D, SHANG S, FENG K, journalName=Agriculture, refType=null, unstructuredReference=YUE D, SHANG S, FENG K, et al. Research on the model of a navigation and positioning algorithm for agricultural machinery based on the IABC-BP network[J]. Agriculture, 2023, 13(9): 1769., articleTitle=Research on the model of a navigation and positioning algorithm for agricultural machinery based on the IABC-BP network, refAbstract=null), Reference(id=1239265692873192017, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2011, volume=12, issue=1, pageStart=48, pageEnd=57, url=null, language=null, rfNumber=[15], rfOrder=14, authorNames=NOURELDIN A, EL-SHAFIE A, BAYOUMI M, journalName=Information Fusion, refType=null, unstructuredReference=NOURELDIN A, EL-SHAFIE A, BAYOUMI M. GPS/INS integration utilizing dynamic neural networks for vehicular navigation[J]. Information Fusion, 2011, 12(1): 48-57., articleTitle=GPS/INS integration utilizing dynamic neural networks for vehicular navigation, refAbstract=null), Reference(id=1239265692999021142, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=70, issue=null, pageStart=1, pageEnd=16, url=null, language=null, rfNumber=[16], rfOrder=15, authorNames=LU S, GONG Y, LUO H, journalName=IEEE Transactions on Instrumentation and Measurement, refType=null, unstructuredReference=LU S, GONG Y, LUO H, et al. Heterogeneous multi-task learning for multiple pseudo-measurement estimation to bridge GPS outages[J]. IEEE Transactions on Instrumentation and Measurement, 2020, 70: 1-16., articleTitle=Heterogeneous multi-task learning for multiple pseudo-measurement estimation to bridge GPS outages, refAbstract=null), Reference(id=1239265693087101529, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=12, issue=2, pageStart=256, pageEnd=null, url=null, language=null, rfNumber=[17], rfOrder=16, authorNames=FANG W, JIANG J, LU S, journalName=Remote Sensing, refType=null, unstructuredReference=FANG W, JIANG J, LU S, et al. A LSTM algorithm estimating pseudo measurements for aiding INS during GNSS signal outages[J]. Remote Sensing, 2020, 12(2): 256., articleTitle=A LSTM algorithm estimating pseudo measurements for aiding INS during GNSS signal outages, refAbstract=null), Reference(id=1239265693196153440, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2022, volume=14, issue=3, pageStart=752, pageEnd=null, url=null, language=null, rfNumber=[18], rfOrder=17, authorNames=TANG Y, JIANG J, LIU J, journalName=Remote Sensing, refType=null, unstructuredReference=TANG Y, JIANG J, LIU J, et al. A GRU and AKF-based hybrid algorithm for improving INS/GNSS navigation accuracy during GNSS outage[J]. Remote Sensing,2022, 14(3): 752., articleTitle=A GRU and AKF-based hybrid algorithm for improving INS/GNSS navigation accuracy during GNSS outage, refAbstract=null), Reference(id=1239265693275845218, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2022, volume=14, issue=18, pageStart=4494, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=18, authorNames=ZHAO S, ZHOU Y, HUANG T, journalName=Remote Sensing, refType=null, unstructuredReference=ZHAO S, ZHOU Y, HUANG T. A novel method for AI-assisted INS/GNSS navigation system based on CNN-GRU and CKF during GNSS outage[J]. Remote Sensing, 2022, 14(18): 4494., articleTitle=A novel method for AI-assisted INS/GNSS navigation system based on CNN-GRU and CKF during GNSS outage, refAbstract=null), Reference(id=1239265693355536997, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2020, volume=12, issue=11, pageStart=1704, pageEnd=null, url=null, language=null, rfNumber=[20], rfOrder=19, authorNames=GAO X, LUO H, NING B, journalName=Remote Sensing, refType=null, unstructuredReference=GAO X, LUO H, NING B, et al. RL-AKF: An adaptive Kalman filter navigation algorithm based on reinforcement learning for ground vehicles[J]. Remote Sensing,2020, 12(11): 1704., articleTitle=RL-AKF: An adaptive Kalman filter navigation algorithm based on reinforcement learning for ground vehicles, refAbstract=null), Reference(id=1239265693431034473, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2024, volume=1, issue=1, pageStart=16, pageEnd=31, url=null, language=null, rfNumber=[21], rfOrder=20, authorNames=金学波, 刘嵩政, journalName=人工智能前沿与应用, refType=null, unstructuredReference=金学波, 刘嵩政. 基于GPS的堆叠串行LSTM组合神经网络目标跟踪方法[J]. 人工智能前沿与应用, 2024,1(1): 16-31., articleTitle=基于GPS的堆叠串行LSTM组合神经网络目标跟踪方法, refAbstract=null), Reference(id=1239265693506531950, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[22], rfOrder=21, authorNames=SHIRI F M, PERUMAL T, MUSTAPHA N, journalName=arXiv preprint arXiv, refType=null, unstructuredReference=SHIRI F M, PERUMAL T, MUSTAPHA N, et al. A comprehensive overview and comparative analysis on deep learning models: CNN, RNN, LSTM, GRU[J].arXiv preprint arXiv: 2023., articleTitle=A comprehensive overview and comparative analysis on deep learning models: CNN, RNN, LSTM, GRU, refAbstract=null), Reference(id=1239265693598806644, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2019, volume=68, issue=3, pageStart=2124, pageEnd=2136, url=null, language=null, rfNumber=[23], rfOrder=22, authorNames=WANG C, WANG J, SHEN Y, journalName=IEEE Transactions on Vehicular Technology, refType=null, unstructuredReference=WANG C, WANG J, SHEN Y, et al. Autonomous navigation of UAVs in large-scale complex environments: A deep reinforcement learning approach[J]. IEEE Transactions on Vehicular Technology, 2019, 68(3): 2124-2136., articleTitle=Autonomous navigation of UAVs in large-scale complex environments: A deep reinforcement learning approach, refAbstract=null), Reference(id=1239265693724635771, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[24], rfOrder=23, authorNames=COHEN N, KLEIN I, journalName=Results in Engineering, refType=null, unstructuredReference=COHEN N, KLEIN I. Inertial navigation meets deep learning: A survey of current trends and future directions[J]. Results in Engineering, 2024: 103565., articleTitle=Inertial navigation meets deep learning: A survey of current trends and future directions, refAbstract=null), Reference(id=1239265693800133248, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[25], rfOrder=24, authorNames=GOLROUDBA RI A A, SABOUR M H, journalName=arXiv preprint arXiv, refType=null, unstructuredReference=GOLROUDBA RI A A, SABOUR M H. Recent advancements in deep learning applications and methods for autonomous navigation: A comprehensive review[J].arXiv preprint arXiv: 2023., articleTitle=Recent advancements in deep learning applications and methods for autonomous navigation: A comprehensive review, refAbstract=null)], funds=[Fund(id=1239265689748435453, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, awardId=2021-JCJQ-ZQ-015, language=CN, fundingSource=国防科技卓越青年科学基金(2021-JCJQ-ZQ-015), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1239265683331149989, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, xref=null, ext=[AuthorCompanyExt(id=1239265683343732902, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Beijing Research Institute of Telemetry, Beijing 100076, China), AuthorCompanyExt(id=1239265683352121511, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, companyId=1239265683331149989, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=北京遥测技术研究所 北京 100076)])], figs=[ArticleFig(id=1239265687923913103, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Fig. 1, caption=Structural diagram of the neural network, figureFileSmall=9Wzy4sxH+Q3J4if3TaNzhA==, figureFileBig=HKTwTBZcP7t7wtwkXB2vpQ==, tableContent=null), ArticleFig(id=1239265688024576404, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=图1, caption=神经网络结构示意图, figureFileSmall=9Wzy4sxH+Q3J4if3TaNzhA==, figureFileBig=HKTwTBZcP7t7wtwkXB2vpQ==, tableContent=null), ArticleFig(id=1239265688150405533, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Fig. 2, caption=Structure of DBNN, figureFileSmall=rQXJBkHS2fWF9Pk/VjWd+w==, figureFileBig=mKt0mCoaqVoi/vWad0zvVw==, tableContent=null), ArticleFig(id=1239265688259457442, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=图2, caption=双分支神经网络架构, figureFileSmall=rQXJBkHS2fWF9Pk/VjWd+w==, figureFileBig=mKt0mCoaqVoi/vWad0zvVw==, tableContent=null), ArticleFig(id=1239265688347537831, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Fig. 3, caption=Fusion structure of NN, figureFileSmall=Bji/Pm6NOD9awvDtJ13TqA==, figureFileBig=dWDirYmoPs6IrP7gRh9hEQ==, tableContent=null), ArticleFig(id=1239265688460784051, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=图3, caption=神经网络预测框架, figureFileSmall=Bji/Pm6NOD9awvDtJ13TqA==, figureFileBig=dWDirYmoPs6IrP7gRh9hEQ==, tableContent=null), ArticleFig(id=1239265688536281527, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Fig. 4, caption=Comparison of 3D trajectory and projection of GNSS rejection position, figureFileSmall=8qXAq7ZcukBHaShiMq2O4w==, figureFileBig=nn7POuVlUlX4qTaE/Mc+dw==, tableContent=null), ArticleFig(id=1239265688649527742, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=图4, caption=三维轨迹与GNSS拒止段位置投影对比图, figureFileSmall=8qXAq7ZcukBHaShiMq2O4w==, figureFileBig=nn7POuVlUlX4qTaE/Mc+dw==, tableContent=null), ArticleFig(id=1239265688745996742, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Fig. 5, caption=3D trajectory figure and projection figure comparison of different model, figureFileSmall=aSTlo9Gs3XjCzz5lAzoyyw==, figureFileBig=+pFC06FfEFrqGUuZrfaEgQ==, tableContent=null), ArticleFig(id=1239265688838271436, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=图5, caption=不同算法模型预测轨迹三维对比图和投影对比图, figureFileSmall=aSTlo9Gs3XjCzz5lAzoyyw==, figureFileBig=+pFC06FfEFrqGUuZrfaEgQ==, tableContent=null), ArticleFig(id=1239265688943129040, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Fig. 6, caption=3D error comparison curve, figureFileSmall=DOplIRQ+3PGbhZC4G7yGMg==, figureFileBig=w0EMafRVJCNVZ5Jb1y5b+A==, tableContent=null), ArticleFig(id=1239265689039598039, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=图6, caption=三维误差对比曲线, figureFileSmall=DOplIRQ+3PGbhZC4G7yGMg==, figureFileBig=w0EMafRVJCNVZ5Jb1y5b+A==, tableContent=null), ArticleFig(id=1239265689161232862, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Table 1, caption=

Parameters of IMU and GNSS error model

, figureFileSmall=null, figureFileBig=null, tableContent=
误差项参数设定值单位
GNSS速度误差[0.1;0.1;0.1]m/s
位置误差[10;10;10]m
INS陀螺零偏3(°)/h
加计零偏100μg
陀螺仪随机游走0.3
加计速度随机游走10
), ArticleFig(id=1239265689249313250, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=表1, caption=

惯性测量单元与GNSS误差模型参数

, figureFileSmall=null, figureFileBig=null, tableContent=
误差项参数设定值单位
GNSS速度误差[0.1;0.1;0.1]m/s
位置误差[10;10;10]m
INS陀螺零偏3(°)/h
加计零偏100μg
陀螺仪随机游走0.3
加计速度随机游走10
), ArticleFig(id=1239265689366753768, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Table 2, caption=

Comparison experiment setting

, figureFileSmall=null, figureFileBig=null, tableContent=
模型编号模型结构特点
A0-纯惯导
A1LSTM统一预测
B1LSTM(主)+GRU分支预测
B2CNN-LSTM(主)+GRU分支预测
), ArticleFig(id=1239265689454834156, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=表2, caption=

对比实验设置

, figureFileSmall=null, figureFileBig=null, tableContent=
模型编号模型结构特点
A0-纯惯导
A1LSTM统一预测
B1LSTM(主)+GRU分支预测
B2CNN-LSTM(主)+GRU分支预测
), ArticleFig(id=1239265689534525936, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=EN, label=Table 3, caption=

Position error result of different model in GNSS rejection area

, figureFileSmall=null, figureFileBig=null, tableContent=
模型编号东向误差/m北向误差/m天向误差/m
RMSEMaxErrorRMSEMaxErrorRMSEMaxError
A080 390.4102 815.916 501.092133.41622 976.488278.242
A19 435.711282.913103 548.640506.7188 002.392203.511
B1984.522146.9033 103.165110.0575 908.970158.934
B2203.5156.0822 187.912105.8645 908.970158.934
), ArticleFig(id=1239265689601634806, tenantId=1146029695717560320, journalId=1238841944844054536, articleId=1239165050607423722, language=CN, label=表3, caption=

不同模型在GNSS拒止段内的定位误差结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型编号东向误差/m北向误差/m天向误差/m
RMSEMaxErrorRMSEMaxErrorRMSEMaxError
A080 390.4102 815.916 501.092133.41622 976.488278.242
A19 435.711282.913103 548.640506.7188 002.392203.511
B1984.522146.9033 103.165110.0575 908.970158.934
B2203.5156.0822 187.912105.8645 908.970158.934
)], attaches=null, journal=Journal(id=1238822691516493831, delFlag=0, nameCn=遥测遥控, nameEn=Journal of Telemetry, Tracking and Command, nameHistory1=null, nameHistory2=null, issn=2095-1000, eissn=null, cn=11-1780/TP, coden=null, periodic=1, 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=uUqWBk5pV4VWPULNoAHkuQ==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Telemetry, Tracking and Command, journalRemark=null, publicationField=null, createdTime=1773287945871, updatedTime=1773299995823, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=uUqWBk5pV4VWPULNoAHkuQ==, picEn=RDDslHHvDBDNnJMjk6BxdA==, jcr=null, cjcr=null, exts=[JournalExt(id=1238873232732909900, language=CN, name=遥测遥控, 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=1773299995835, updatedTime=1773299995835, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://manuscript.spacejournal.cn/ycyk, submissionEditorUrl=https://manuscript.spacejournal.cn/ycyk, submissionReviewUrl=https://manuscript.spacejournal.cn/ycyk, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1238873232779047245, language=EN, name=Journal of Telemetry, Tracking and Command, 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=1773299995846, updatedTime=1773299995846, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://manuscript.spacejournal.cn/ycyk, submissionEditorUrl=https://manuscript.spacejournal.cn/ycyk, submissionReviewUrl=https://manuscript.spacejournal.cn/ycyk, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1238841944844054536, websiteList=[Website(id=1238843830011744952, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1238841944844054536, 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/ycyk/CN, language=CN, createTime=1773292985680, createBy=18614031015, updateTime=1773293577303, updateBy=18614031015, name=遥测遥控-中文, tplId=1146099689490845704, title=遥测遥控, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1238846390617568150, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=articleTextType, value=kx, createTime=1773293596176, updateTime=1773293596176, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390596596627, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=banner, value=null, createTime=1773293596171, updateTime=1773293596171, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390638539673, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=grayFlag, value=0, createTime=1773293596181, updateTime=1773293596181, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390588208018, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=logo, value=https://castjournals.cast.org.cn/joweb/ycyk/CN/file/pic?fileId=g9S2GGWx7iZjf51o9vpmmA==, createTime=1773293596169, updateTime=1773293596169, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390651122587, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=minRunFlag, value=0, createTime=1773293596184, updateTime=1773293596184, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390613373845, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/ycyk/CN/file/pic, createTime=1773293596175, updateTime=1773293596175, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390646928282, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=silenceFlag, value=0, createTime=1773293596183, updateTime=1773293596183, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390604985236, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1773293596173, updateTime=1773293596173, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390625956759, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=themeColor, value=null, createTime=1773293596178, updateTime=1773293596178, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846390634345368, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830011744952, code=themeStyle, value=null, createTime=1773293596180, updateTime=1773293596180, creator=18614031015, updator=18614031015)]), Website(id=1238843830078853819, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1238841944844054536, 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/ycyk/EN, language=EN, createTime=1773292985696, createBy=18614031015, updateTime=1773293566698, updateBy=18614031015, name=遥测遥控-英文, tplId=1146101810881728533, title=Journal of Telemetry, Tracking and Command, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1238846419235304352, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=articleTextType, value=kx, createTime=1773293602999, updateTime=1773293602999, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419214332829, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=banner, value=null, createTime=1773293602994, updateTime=1773293602994, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419256275875, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=grayFlag, value=0, createTime=1773293603004, updateTime=1773293603004, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419205944220, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=logo, value=https://castjournals.cast.org.cn/joweb/ycyk/EN/file/pic?fileId=g9S2GGWx7iZjf51o9vpmmA==, createTime=1773293602992, updateTime=1773293602992, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419268858789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=minRunFlag, value=0, createTime=1773293603007, updateTime=1773293603007, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419231110047, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/ycyk/EN/file/pic, createTime=1773293602998, updateTime=1773293602998, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419264664484, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=silenceFlag, value=0, createTime=1773293603006, updateTime=1773293603006, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419222721438, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1773293602996, updateTime=1773293602996, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419243692961, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=themeColor, value=null, createTime=1773293603001, updateTime=1773293603001, creator=18614031015, updator=18614031015), WebsiteProps(id=1238846419252081570, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1238843830078853819, code=themeStyle, value=null, createTime=1773293603003, updateTime=1773293603003, creator=18614031015, updator=18614031015)])], journalTitle=遥测遥控, weixinUrl=null, journalUrl=https://ycyk.spacejournal.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Telemetry, Tracking and Command, journalPhotoCn=uUqWBk5pV4VWPULNoAHkuQ==, journalPhotoEn=RDDslHHvDBDNnJMjk6BxdA==, journalFirstLetter=J, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/ycyk/CN/10.12347/j.ycyk.20250616001, detailUrlEn=https://castjournals.cast.org.cn/joweb/ycyk/EN/10.12347/j.ycyk.20250616001, pdfUrlCn=https://castjournals.cast.org.cn/joweb/ycyk/CN/PDF/10.12347/j.ycyk.20250616001, pdfUrlEn=https://castjournals.cast.org.cn/joweb/ycyk/EN/PDF/10.12347/j.ycyk.20250616001, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
卫星拒止条件下基于双分支神经网络模型的GNSS组合定位算法
收藏切换
PDF下载
尹旖旎 , 孙迅 , 蒋林 , 何兆 , 王志颖
遥测遥控 | 导航技术专栏 2025,46(5): 36-44
收起
收藏切换
遥测遥控 | 导航技术专栏 2025, 46(5): 36-44
卫星拒止条件下基于双分支神经网络模型的GNSS组合定位算法
全屏
尹旖旎, 孙迅, 蒋林, 何兆, 王志颖
作者信息
  • 北京遥测技术研究所 北京 100076
  • 尹旖旎 2000年生,硕士研究生。

    孙迅 1974年生,硕士,研究员。

    蒋林 1979年生,本科,研究员。

    何兆 2000年生,硕士研究生。

    王志颖 1998年生,硕士研究生。

Dual-Branch Neural Network for GNSS Localization Prediction in GNSS-Denied Environments
Yini YIN, Xun SUN, Lin JIANG, Zhao HE, Zhiying WANG
Affiliations
  • Beijing Research Institute of Telemetry, Beijing 100076, China
出版时间: 2025-09-15 doi: 10.12347/j.ycyk.20250616001
文章导航
收藏切换

惯性导航/卫星导航的组合导航广泛应用于无人机等多种移动载体中,然而在全球导航卫星系统信号中断期间,惯性导航系统误差会迅速积累,严重影响导航精度。现有研究多集中于对水平二维的误差建模,忽略了高度方向的动态特性,限制了其实际应用。为解决这一问题,本文提出一种面向三维导航的双分支神经网络模型,同时建模经度、纬度与高度方向的位置增量,以适应三维空间中的动态导航需求。该模型采用LSTM(长短期记忆网络)与GRU(门控循环单元)构建解耦的双分支结构,分别针对经纬面与高度方向设计建模路径,并在主分支中引入卷积神经网络(CNN)以增强时序特征提取能力。实验表明:所提出的网络提升了导航系统在三维空间下的预测精度。相较于传统的定位方法,本文提出的方法在东北天三向的均方根误差分别降低了97.8%、97.9%、26.2%,具备良好的工程应用潜力。

无人机  /  全球导航卫星系统  /  组合导航  /  三维定位  /  双分支神经网络  /  惯性导航

Inertial/Global Navigation Satellite System (GNSS) integrated navigation has been widely applied in various mobile platforms such as unmanned aerial vehicles (UAVs). However, during GNSS signal outages, INS errors accumulate rapidly, severely degrading navigation accuracy. Existing research primarily focuses on horizontal two-dimensional error modeling while neglecting the dynamic characteristics in the vertical (altitude) direction, limiting its practical application in three-dimensional space. To address this issue, this paper proposes a dual-branch neural network model for three-dimensional navigation, which simultaneously models position increments in the longitude, latitude, and altitude directions to cater to the demands of dynamic navigation in 3D space. The model adopts a decoupled dual-branch structure built with LSTM and GRU networks, designing separate modeling paths for the horizontal and vertical components. A convolutional neural network (CNN) is further incorporated into the main branch to enhance temporal feature extraction. Experimental results demonstrate that the proposed network significantly improves three-dimensional navigation accuracy. Compared with conventional positioning methods, it reduces the root mean square error(RMSE) along the east, north, and up axes by 97.8 %, 97.9 %, and 26.2 %, respectively, demonstrating its strong potential for practical deployment.

Unmanned aerial vehicle  /  GNSS denial  /  Integrated navigation  /  3D positioning  /  Dualbranch neural network  /  Inertial navigation
尹旖旎, 孙迅, 蒋林, 何兆, 王志颖. 卫星拒止条件下基于双分支神经网络模型的GNSS组合定位算法. 遥测遥控, 2025 , 46 (5) : 36 -44 . DOI: 10.12347/j.ycyk.20250616001
Yini YIN, Xun SUN, Lin JIANG, Zhao HE, Zhiying WANG. Dual-Branch Neural Network for GNSS Localization Prediction in GNSS-Denied Environments[J]. Journal of Telemetry, Tracking and Command, 2025 , 46 (5) : 36 -44 . DOI: 10.12347/j.ycyk.20250616001
在过去几十年,全球导航卫星系统(GNSS)与惯性导航系统(INS)已广泛应用于民用与军事领域,并成功集成于多种实际导航系统中[1,2]。INS与GNSS的时空互补特性推动其组合导航系统发展,通过融合GNSS的高精度与INS的连续性,实现更稳健的定位性能。松耦合、紧耦合以及超紧耦合架构的迭代发展,使组合导航在自动驾驶、自助机器人、无人机精准农业、飞行器导航、导弹精确制导等军事与民用领域实现规模化应用[38]。然而,GNSS信号固有的脆弱性严重制约着集成系统的可靠性。在复杂飞行环境下,当飞行器进入城市峡谷、地下隧道或茂密森林等物理遮挡环境时,卫星可见数可能骤降至几何定位阈值以下;在电磁干扰严重的区域、飞行器高机动飞行过程,或受对抗干扰(如干扰器、欺骗信号)影响时,GNSS信号可能被完全屏蔽或失效,使得INS无法通过外部测量进行误差修正,一旦GNSS信号中断超过INS自主导航的临界时间窗口,导航精度将迅速下降[9]。因此,在GNSS不可用条件下保持定位的准确性,对于提升飞行器自主导航能力具有重要意义。
传统解决方案大多基于扩展卡尔曼滤波或粒子滤波构建INS误差模型或载具动力学模型进行约束,以抑制INS系统在无校正情况下的发散。但此类方法存在双重局限性:其一,强依赖于惯性传感器误差的事先标定精度,而温度漂移、刻度因子非线性等时变参数难以在线辨识;其二,在飞行器执行急转弯、大过载机动时,预设的角速度/加速度噪声统计模型易产生失配,导致滤波发散风险显著增加。因此,亟待更为有效的方法解决GNSS信号拒止下的组合导航精度提升问题。
人工智能技术的迅速发展,为突破传统导航算法的理论边界提供了新范式,展现出强大潜力。深度学习与机器学习方法可以通过对海量的高频率INS数据进行挖掘,捕获导航数据中隐含的时空特征,以端对端的方式建立INS数据与导航误差之间的非线性映射,对INS系统进行预测或补偿,以提升INS的抗漂移能力。在GNSS中断期间,AI模型可用于桥接INS与GNSS之间的断层,其研究路径主要包括以下三类[10]:①误差建模类模型[1114]:以INS解算值和IMU(惯性测量单元)原始测量为输入,学习INS与GNSS之间的位置误差,间接估算伪GNSS测量;②状态变量预测类模型[15]:直接输出融合系统的状态向量,如位置、速度等;③GNSS位置增量预测类模型[16,17]:基于历史特征预测相邻时刻的GNSS位置变化量,以补偿导航缺失。相较于前两类,已有研究表明位置增量预测类模型在应对动态环境变化与长时间中断方面表现更优,在鲁棒性与融合适配性方面更具优势。
循环神经网络(RNN)结构(如GRU、LSTM)被广泛用于此类时序建模任务,取得了良好效果。如Tang等人使用GRU预测水平位移并融合KF(卡尔曼滤波)[18];Zhao等人引入CNN-GRU提升经纬预测精度[19];Fang等人验证了LSTM在导航任务中的适用性[20]。但这些研究大多忽略了对高度方向的系统建模,难以满足飞行平台三维导航的需求。
针对上述问题,本文提出一套异构双分支神经网络(Dual-Branch Neural Network, DBNN),分别适配不同导航维度的建模需求。网络采用结构差异化建模策略以适配不同导航维度的数据特性:其中经纬度分支结合GRU与LSTM结构提取惯导数据的长期依赖关系,增强水平平面的状态记忆能力;而高度分支采用轻量CNN与LSTM组合结构,增强局部运动特征提取能力,提高模型对垂直方向突变与缓变的建模适应性。此外,为提升模型对预测偏差的抗扰性,本文引入组合损失函数以联合约束增量精度与累积轨迹误差。通过在仿真数据集上构建包含GNSS失效段的飞行轨迹,本文进一步对比不同结构、损失形式、分支策略下的三维预测精度,并与传统INS、INS+KF等方法进行系统对比。研究结果验证了本文提出的双分支结构在三维导航精度保持与时间鲁棒性方面的显著优势,具有重要的工程推广与理论探索价值。
惯性导航中的数据,如角速度、加速度、速度与姿态信息,具有明显的时间序列特性。当前状态不仅依赖于当前输入,还与历史时刻的状态密切相关。因此,建模此类时序信号的关键在于提取其时间依赖关系,并对状态演化趋势进行预测。传统的卡尔曼滤波器(Kalman Filter)能够对线性动态系统进行递归估计,但在面对强非线性、非高斯误差以及复杂动态环境时,性能受到限制。而神经网络,尤其是针对时间序列设计的结构,如循环神经网络(RNN)、长短期记忆网络(LSTM)、门控循环单元(GRU)等,在建模复杂动态关系方面展现出显著优势[2125]
RNN是处理序列数据的基本结构,其核心思想是利用隐藏状态ht存储前一时刻的信息,并与当前输入xt共同决定当前的状态输出:
其中,W为权重矩阵,b为偏置向量。
RNN结构简洁,但在长时间依赖建模中容易出现梯度消失或爆炸,难以捕捉长距离的状态关联。为克服RNN的记忆瓶颈,LSTM引入了记忆单元ct和门控机制,允许网络在长时间内保留或丢弃信息,其主要计算过程包括:
通过门控机制,LSTM能灵活调控记忆的存取过程,极大提高了对长序列建模的能力。LSTM的具体结构见图1(a)
图1(b)是GRU结构示意图。GRU是LSTM的简化版本,合并了输入门与遗忘门,减少了参数数量,提升了训练效率,其更新公式为:
其中,更新门zt≈1时保留更多旧信息,反之更换为新状态;重置门rt控制“忘掉多少”旧状态的信息;W为输入变量xt的权重矩阵,U为上一状态ht-1的权重矩阵。
GRU在性能上接近甚至优于LSTM,特别适用于数据量有限或对实时性要求高的预测任务[23]
卷积神经网络(CNN)可通过一维卷积在时域上提取局部动态模式(如短时速度变化、姿态突变),与LSTM联合构成CNN-LSTM结构,能实现从局部到全局的多尺度时间建模:
• CNN提取短期动态特征
• LSTM学习长期时序演化关系
该结构在行为识别、轨迹预测与导航状态估计中广泛应用,能够显著提升特征提取与时间关联的能力。
GNSS中断段的轨迹预测问题本质是一个时间序列回归问题,其输入为惯性测量序列,输出为未来的三维导航状态(经纬高)[22]。采用神经网络可有效学习输入与输出之间的非线性时序映射关系,尤其在复杂航迹、变化速度下具有较强的自适应能力[2325]。为此,本文采用基于CNN-LSTM与GRU的双分支结构,分别对水平面与垂直方向的状态演化特性建模,以提升整体导航预测精度。
本节将详细介绍基于双分支神经网络辅助的组合导航方法。第一节介绍系统的设计。随后,构建融合时序建模与特征分离机制的双分支神经网络架构,阐述其在导航误差估计中的建模策略。
在地球的椭球体模型下,载体位置的变化(ΔLλh)由速度vn=(vn,ve,vu)主导,其数学模型为:
其中,RMRN分别表示子午圈曲率半径和卯酉圈曲率半径,Δt为时间间隔,当时间间隔足够短时,式(4)中的等号成立。
在INS中,速度通过IMU测得的比力积分获得,需先进行方向余弦变换并剔除重力项,其公式为:
其中,方向余弦用于坐标变换,将机体坐标系下的比力fb转换到导航系下;gn为重力加速度,为导航系下的地球自转角速度,可视为常量;是导航系相对地球的旋转。
姿态角来源于陀螺仪的观测。综合上述推导可以确定以下函数关系:
进一步得到位置变化的函数关系:
以上公式构成了后续神经网络建模的理论依据,即通过函数h(·)建立输入与位置增量的映射。位置增量的表达为:
其中,表示在时间tktk-1之间估计的位置变化量,分别表示估计的GNSS位置。
在GNSS位置估计中,由真实值PGNSS(tk)和测量误差项δPGNSS(tk)组成。由于GNSS的噪声水平相对稳定,可以认为相邻时间的测量误差属于相同的概率分布,具有相同的方差,所以δPGNSS(tk)-δPGNSS(tk-1)的值较小。因此,采用输入,fb,A,vn作为神经网络的输入参数来预测ΔPGNSS能够有效减少测量噪声的影响,提高预测精度。
导航任务中,经纬度与高度的变化具有明显的时间尺度与幅度差异:高度在多数场景下变化缓慢,但遇地形起伏或飞行高度调整时又可能发生突变。因此,本文设计了一种双分支神经网络架构:其中一支专注于高度方向建模,输入为高度、姿态与惯导观测量;另一支则用于经纬度方向建模的变化趋势。基于此,公式(4)可细化为:
在GNSS中断条件下,近年来多种AI相关研究被用于辅助导航。这些研究大多忽略了对高度方向的系统建模。针对这一问题,本文提出的网络结构如图2所示。
主分支采用CNN-LSTM模块处理经纬度方向的输入序列。该序列长度为200,维度为12,经过卷积层提取局部变化特征,随后池化压缩降维后送入LSTM以捕捉长时依赖特征。LSTM对运动状态变化如转弯、爬升等具有较强建模能力。辅助分支则采用GRU,对高度变化过程建模。GRU在简化网络复杂度的同时,保留了对高度平缓变化的建模能力,具有较高计算效率。
本文提出的在卫星拒止条件下基于双分支神经网络的GNSS定位预测算法,采用GNSS/INS松耦合结构,支持训练模式与预测模式两种运行方式,其整体流程如图3所示。
当GNSS信号可用时,系统处于训练状态,AI模块以IMU观测量(角速度、比力、姿态、速度与高度)为输入,输出预测的位置增量,并与GNSS测得的位置变化值进行对比,通过反向传播优化网络参数。当GNSS信号不可用时,系统切换至预测模式,AI模型基于惯导输入生成伪GNSS测量,作为下一步与INS观测量进行融合的输入。
本章围绕所提出的双分支神经网络预测算法,开展了在GNSS拒止环境下的仿真实验。首先基于PSINS(精确捷联惯性导航系统)工具构建了包含典型飞行阶段的无人机轨迹数据,并设定GNSS丢失时段用于测试。随后设计双重损失函数与指数衰减学习率策略以提升模型训练效果。通过构建多个对比模型,从经纬度和高度两个维度进行性能评估,验证了双分支结构在提升预测精度的有效性。
为评估模型在GNSS拒止环境下的导航性能,本文使用PSINS工具箱模拟了1 500 s的无人机飞行轨迹,包含典型的上升、巡航与降落阶段,如图4所示。
在第700~1 000 s内模拟GNSS信号完全丢失的情形(图3框选段),作为测试数据;第1~700 s的数据则用于训练模型,以保持时间序列的上下文关联性。其中,仿真惯性器件误差和GNSS定位误差参数设定如表1所示。
针对导航任务中对轨迹连贯性与局部预测精度的双重要求,本文设计了结合增量误差与轨迹误差的双重损失函数,以兼顾短期定位精度与长期轨迹约束。
训练过程采用均方误差(MSE)作为基础损失度量,并构建多目标损失函数以约束模型输出的物理一致性与轨迹连续性。总损失函数由三部分组成:
• 增量预测损失Linc:衡量当前时刻预测增量与真实增量之间的差异;
• 累积轨迹损失Ltraj:通过对输出增量进行累加后与真实轨迹进行对比;
• 平滑损失Lsmooth:对相邻时间步之间的预测差值施加二范数约束,限制轨迹剧烈震荡。
损失函数定义如下:
其中,
各权重参数设定为:λ1=1.0, λ2=0.5, λ3=0.1,其中垂直方向损失额外加权λup=2.0,以提升高度方向的建模精度。
模型训练采用Adam(自适应矩估计)优化器,初始学习率设为5×10-4,并引入指数衰减学习率调度器,在训练过程中,采用小批量梯度下降进行参数优化,并将训练轮数设置为固定值。考虑到导航序列具有强时序相关性,且不同阶段的损失下降速率存在差异,为避免学习率过大导致收敛震荡,或学习率过小引起训练停滞,本文采用如下指数衰减更新策略:
其中,ηt表示第t次迭代的学习率,γ∈(0,1)为衰减因子,本文设置为γ=0.999 99。该策略可在训练初期快速收敛,后期缓慢调整,增强模型泛化能力与收敛稳定性。为避免梯度爆炸问题,在每次反向传播后执行梯度裁剪操作,将参数梯度范数限制在1.0以内。此外,训练过程中引入早停机制(Early Stopping),当验证损失在连续10个epoch内未显著下降(小于0.001)时,提前终止训练,防止过拟合并节省计算资源。
此外,综合参考文献[17-18]的实验结论,为平衡性能与计算效率,本文将主分支网络设置为两层LSTM,辅分支网络设置为两层GRU,且隐藏单元数均设定为128。
为验证CNN-LSTM+GRU预测算法的有效性,本文设计了三组对比实验,各实验模型的核心结构及差异设置见表2
表2中,A1表示使用单分支神经网络对无人机东北天增量进行统一预测;B1与B2表示引入双分支结构后的网络变体。A1为单分支LSTM模型,B1为LSTM+GRU双分支模型,B2为引入CNN模块的CNN-LSTM+GRU模型。所有模型均采用相同的训练数据与样本划分策略,并在多个导航精度指标上进行了综合评估,包括轨迹重建误差、各方向误差均值与方差,以及经纬度与高度方向的RMSE等。
本文首先对各方法在700~1 000 s期间的预测轨迹进行了三维轨迹重建,并绘制了其在东-北(平面)与东-高(高度)方向的投影轨迹图。图5(a)展示了各方法在GNSS拒止环境下的三维轨迹对比图,图5(b)为其对应的平面与高度投影图。
图5可以看出,相较于纯惯性导航(INS)轨迹,各类融合神经网络模型在GNSS拒止段内的预测轨迹均表现出一定程度的收敛趋势,显著提升了轨迹重建的稳定性。其中,采用CNN–LSTM+GRU双分支结构的B2模型,其预测轨迹在东、北、天三个方向上均与真实轨迹更贴合,表现出最优的重建精度。相比之下,A1模型虽同样引入了LSTM结构,但由于采用同一网络同时对三维位移进行回归预测,导致模型难以有效学习各方向间的独立动态特征,最终出现较明显的漂移误差,特别是高度直接发散。这一现象进一步验证了双分支结构在建模空间方向异构性方面的优势,有助于提升神经网络在复杂导航场景中的泛化能力。
为了更直观地对比各模型在不同方向上的导航精度,本文进一步绘制了各模型东(East)、北(North)及高程(Up)三个方向上的误差随时间变化的曲线图,如图6所示。
由于实验设计的B1和B2实验采用的是同一分支网络预测高度,其误差是一致的,所以图6(c)中,只需对比IMU、单分支网络和双分支网络的定位误差结果。从图中可以看出,传统的纯惯性导航系统误差在缺失段内快速发散,而融合神经网的模型能够显著抑制误差增长。其中,采用CNN-LSTM+GRU双分支神经网络模型(B2)在各方向上均表现出较低的误差波动,验证了其对系统状态变化有更强的适应能力。值得注意的是,在高程方向上,误差曲线仍存在一定程度的振荡与发散趋势,分析其原因主要包括:一是高度方向的惯导误差受加速度双积分影响,易发生系统性偏移;二是高程数据本身动态特征较弱,模型难以有效捕捉时序模式;三是目前模型中高度方向分支未引入CNN等增强机制,导致其对长时间段预测中的趋势变化适应性有限。后续可通过引入更强的特征提取模块或高度方向的结构专门设计,以进一步提升三维导航建模的鲁棒性与精度。
为全面评估各模型在GNSS信号不可用场景下的导航精度,本文统计了不同模型在东、北、天三个方向上的定位误差,包括均方根误差(RMSE)与最大误差(Max Error)两项指标,实验结果如表3所示。
在东向,相较于基准模型A1(RMSE为9 435.711 m),引入GRU辅助分支的B1模型将RMSE显著降低至984.522 m,降幅达89.6%;进一步引入CNN模块以构建CNN-LSTM高度分支的B2模型,将RMSE进一步压缩至203.510 m,较A1降低约97.8%,相较B1亦下降约79.3%。在最大误差指标上,B1和B2分别为146.903 m与56.082 m,均远优于A1的282.913 m。
在北向,B1模型将RMSE从A1的103 548.640m降至3 103.165 m,降幅高达97.0%;B2模型进一步优化至2 187.912 m,总降幅达97.9%,相较B1仍提升约 29.5%。对应的最大误差方面,B1为110.057 m,B2为105.864 m,变化较小,显示两者在北向误差极值上表现相近。
在高程方向(天向),A1的RMSE为8 002.392 m,由于B1和B2模型用的是同一分支网络,其RMSE均为5 908.970 m,相较单分支模型A1降幅为26.2%,说明双分支网络结构在高度预测中具有更稳定的误差控制能力。
整体来看,在神经网络结构相同的前提下,对比B1(LSTM+GRU)与B2(CNN-LSTM+GRU),可观察到后者在水平方向上的误差指标均有所下降,这验证了CNN模块对于特征提取与建模的增强能力。同时,双分支神经网络结构能更好地分担不同方向的建模压力,提升多维度导航状态的协同预测性能,展现出更强的精度优势与系统鲁棒性。
本文针对GNSS信号中断条件下定位精度下降的问题,提出一种基于CNN–LSTM与GRU结构的双分支神经网络导航预测方法。该方法通过结构解耦与特征建模,提升了导航系统在三维空间下的预测精度。实验结果表明:相较于传统单分支模型,该结构在多个导航误差指标上具有更优表现。
本研究的主要工作如下:
①设计双分支融合网络结构:构建由CNN-LSTM主分支与GRU辅助分支组成的并行神经网络模型,提高了模型对不同空间维度动态特性的学习能力。
②基于已有仿真平台进行实验验证:依托PSINS工具箱所提供的UAV飞行仿真功能,选取各类典型飞行状态的轨迹数据,设置GNSS中断场景,构建训练与测试数据集,对提出模型的预测性能进行对比验证。
未来可进一步优化天向建模路径,引入轻量CNN或注意力机制,以提升整体三维导航精度。同时,可探索将增量学习机制与当前网络结构结合,以实现模型在动态场景中的持续适应与性能保持。
  • 国防科技卓越青年科学基金(2021-JCJQ-ZQ-015)
参考文献 引证文献
排序方式:
[1]
HOFMANN-WELLENHOF B, LICHTENEGGER H,WASLE E. GNSS-global navigation satellite systems: GPS, GLONASS, Galileo, and more[M]. Vienna:Springer, 2007.
[2]
BROWN R G, HWANG P Y C. Introduction to random signals and applied Kalman filtering: with MATLAB exercises[M]. 4th ed. Hoboken: John Wiley & Sons, 2012.
[3]
GREWAL M S, WEILL L R, ANDREWS A P. Global positioning systems, inertial navigation, and integration[M]. Hoboken: John Wiley & Sons, 2007.
[4]
HE Y, LI J, LIU J. Research on GNSS INS & GNSS/INS integrated navigation method for autonomous vehicles: A survey[J]. IEEE Access, 2023, 11: 79033-79055.
[5]
YIN Z, YANG J, MA Y, et al. A robust adaptive extended Kalman filter based on an improved measurement noise covariance matrix for the monitoring and isolation of abnormal disturbances in GNSS/INS vehicle navigation[J]. Remote Sensing, 2023, 15(17): 4125.
[6]
WANG D, DONG Y, LI Z, et al. Constrained MEMS-based GNSS/INS tightly coupled system with robust Kalman filter for accurate land vehicular navigation[J]. IEEE Transactions on Instrumentation and Measurement, 2019,69(7): 5138-5148.
[7]
JIANG H, SHI C, LI T, et al. Low-cost GPS/INS integration with accurate measurement modeling using an extended state observer[J]. GPS Solutions, 2021, 25(1):1-15.
[8]
SHEN C, ZHANG Y, GUO X, et al. Seamless GPS/inertial navigation system based on self-learning square-root cubature Kalman filter[J]. IEEE Transactions on Industrial Electronics, 2020, 68(1): 499-508.
[9]
LI D, JIA X, ZHAO J. A novel hybrid fusion algorithm for low-cost GPS/INS integrated navigation system during GPS outages[J]. IEEE Access, 2020, 8: 53984-53996.
[10]
JWO D J, BISWAL A, MIR I A. Artificial neural networks for navigation systems: A review of recent research[J]. Applied Sciences, 2023, 13(7): 4475.
[11]
ABDOLKARIMI E S, MOSAVI M R. A low-cost integrated MEMS-based INS/GPS vehicle navigation system with challenging conditions based on an optimized IT2FNN in occluded environments[J]. GPS Solutions,2020, 24(4): 108.
[12]
DAI H, BIAN H, WANG R, et al. An INS/GNSS integrated navigation in GNSS denied environment using recurrent neural network[J]. Defence Technology, 2020, 16(2): 334-340.
[13]
AL BITAR N, GAVRILOV A. A new method for compensating the errors of integrated navigation systems using artificial neural networks[J]. Measurement, 2021,168: 108391.
[14]
YUE D, SHANG S, FENG K, et al. Research on the model of a navigation and positioning algorithm for agricultural machinery based on the IABC-BP network[J]. Agriculture, 2023, 13(9): 1769.
[15]
NOURELDIN A, EL-SHAFIE A, BAYOUMI M. GPS/INS integration utilizing dynamic neural networks for vehicular navigation[J]. Information Fusion, 2011, 12(1): 48-57.
[16]
LU S, GONG Y, LUO H, et al. Heterogeneous multi-task learning for multiple pseudo-measurement estimation to bridge GPS outages[J]. IEEE Transactions on Instrumentation and Measurement, 2020, 70: 1-16.
[17]
FANG W, JIANG J, LU S, et al. A LSTM algorithm estimating pseudo measurements for aiding INS during GNSS signal outages[J]. Remote Sensing, 2020, 12(2): 256.
[18]
TANG Y, JIANG J, LIU J, et al. A GRU and AKF-based hybrid algorithm for improving INS/GNSS navigation accuracy during GNSS outage[J]. Remote Sensing,2022, 14(3): 752.
[19]
ZHAO S, ZHOU Y, HUANG T. A novel method for AI-assisted INS/GNSS navigation system based on CNN-GRU and CKF during GNSS outage[J]. Remote Sensing, 2022, 14(18): 4494.
[20]
GAO X, LUO H, NING B, et al. RL-AKF: An adaptive Kalman filter navigation algorithm based on reinforcement learning for ground vehicles[J]. Remote Sensing,2020, 12(11): 1704.
[21]
金学波, 刘嵩政. 基于GPS的堆叠串行LSTM组合神经网络目标跟踪方法[J]. 人工智能前沿与应用, 2024,1(1): 16-31.
[22]
SHIRI F M, PERUMAL T, MUSTAPHA N, et al. A comprehensive overview and comparative analysis on deep learning models: CNN, RNN, LSTM, GRU[J].arXiv preprint arXiv: 2023.
[23]
WANG C, WANG J, SHEN Y, et al. Autonomous navigation of UAVs in large-scale complex environments: A deep reinforcement learning approach[J]. IEEE Transactions on Vehicular Technology, 2019, 68(3): 2124-2136.
[24]
COHEN N, KLEIN I. Inertial navigation meets deep learning: A survey of current trends and future directions[J]. Results in Engineering, 2024: 103565.
[25]
GOLROUDBA RI A A, SABOUR M H. Recent advancements in deep learning applications and methods for autonomous navigation: A comprehensive review[J].arXiv preprint arXiv: 2023.
2025年第46卷第5期
PDF下载
133
58
引用本文
BibTeX
文章信息
doi: 10.12347/j.ycyk.20250616001
  • 接收时间:2025-06-16
  • 首发时间:2026-03-13
  • 出版时间:2025-09-15
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-06-16
  • 修回日期:2025-07-24
基金
国防科技卓越青年科学基金(2021-JCJQ-ZQ-015)
作者信息
    北京遥测技术研究所 北京 100076
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/ycyk/CN/10.12347/j.ycyk.20250616001
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