Article(id=1149743084524712289, tenantId=1146029695717560320, journalId=1146031787341344770, issueId=1149743083069288795, articleNumber=1003-3033(2024)06-0164-09, orderNo=null, doi=10.16265/j.cnki.issn1003-3033.2024.06.0720, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1702483200000, receivedDateStr=2023-12-14, revisedDate=1710777600000, revisedDateStr=2024-03-19, acceptedDate=null, acceptedDateStr=null, onlineDate=1752049712543, onlineDateStr=2025-07-09, pubDate=1719504000000, pubDateStr=2024-06-28, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1752049712543, onlineIssueDateStr=2025-07-09, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1752049712543, creator=13701087609, updateTime=1752049712543, updator=13701087609, issue=Issue{id=1149743083069288795, tenantId=1146029695717560320, journalId=1146031787341344770, year='2024', volume='34', issue='6', pageStart='1', pageEnd='252', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1752049712197, creator=13701087609, updateTime=1756468919644, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1168278582599098697, tenantId=1146029695717560320, journalId=1146031787341344770, issueId=1149743083069288795, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1168278582599098698, tenantId=1146029695717560320, journalId=1146031787341344770, issueId=1149743083069288795, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=164, endPage=172, ext={EN=ArticleExt(id=1149743086110159209, articleId=1149743084524712289, tenantId=1146029695717560320, journalId=1146031787341344770, language=EN, title=Vehicle trajectory prediction based on EKF-GRU, columnId=1149733269173878863, journalTitle=China Safety Science Journal, columnName=Safety engineering technology, runingTitle=null, highlight=null, articleAbstract=

To enhance the driving safety and achieve correct decision planning for autonomous vehicles,a safe driving trajectory prediction method based on EKF-GRU was proposed. By combining learning-based methods with physics-based approaches,the prediction accuracy was improved and the rationality of the predicted trajectories was enhanced. In the first step of this method,a prediction network was constructed based on GRU to predict the longitudinal acceleration and yaw angular velocity of vehicles by extracting historical trajectory features. In the second step,an EKF state estimator was built based on the nonlinear vehicle kinematics to generate the vehicle's future limited-time trajectory,incorporating the observations obtained previously. The trajectory prediction method was validated on the NGSIM I-80 and US-101 multi-vehicle trajectory datasets. Experimental results demonstrate that the final distance errors (FDE),root mean square errors (RMSE),and average distance errors (ADE) of the predicted trajectories generated by traditional physics-based methods are 6.48,7.69 and 3.03 meters,respectively. In contrast,trajectories predicted using EKF-GRU exhibit higher accuracy,and the corresponding values are 5.45,6.67 and 2.56 meters,respectively. This represents improvements of 15.90%,13.26% and 15.51%.

, 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=Chuanying ZHANG, Guoyan XU, Zhifa CHEN, Bin ZHOU, Liwei CHEN, Wei HONG), CN=ArticleExt(id=1149743104707703787, articleId=1149743084524712289, tenantId=1146029695717560320, journalId=1146031787341344770, language=CN, title=基于EKF-GRU的车辆轨迹预测, columnId=1149733269727526997, journalTitle=中国安全科学学报, columnName=安全工程技术, runingTitle=null, highlight=null, articleAbstract=

为提升行车安全,实现自动驾驶车辆正确的决策规划,提出基于扩展卡尔曼滤波(EKF)-门控循环单元(GRU)的车辆轨迹预测方法,结合学习方法与物理模型,在提升预测精度的同时,提高轨迹预测的合理性。首先,基于GRU构建预测网络,通过提取车辆的历史轨迹特征预测车辆的纵向加速度及横摆角速度;其次,基于车辆非线性运动学构建EKF状态估计器,结合观测值生成车辆未来有限时域的行驶轨迹;最后,在高速公路多车轨迹数据集NGSIM I-80和US-101上进行轨迹预测方法验证。结果表明:采用传统的物理模型生成预测轨迹,其最终距离误差(FDE)、均方根误差(RMSE)、平均距离误差(ADE)值分别为6.48、7.69和3.03 m。相比之下,利用EKF-GRU生成的预测轨迹表现出更高的准确性,对应的数值分别为5.45、6.67和2.56 m,分别提升15.90%、13.26%和15.51%。

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=3voP7EKm8Mlh6WHuoKJCwA==, magXml=1v+I+9D8DAPsURUZc0FIig==, pdfUrl=null, pdf=5Pg36SK6RK3h3yrbkkmM1g==, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=XHobppg/UFRBi7MmOg0DLg==, mapNumber=null, authorCompany=null, fund=null, authors=

张传莹 (1997—),男,山东潍坊人,硕士研究生,主要研究方向为轨迹预测和碰撞风险评估。E-mail:

徐国艳 副教授

, authorsList=张传莹, 徐国艳, 陈志发, 周彬, 陈立伟, 洪玮)}, authors=[Author(id=1168181787835510976, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=zf2113106@buaa.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1168181787957145794, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181787835510976, language=EN, stringName=Chuanying ZHANG, firstName=Chuanying, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1168181788045226179, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181787835510976, language=CN, stringName=张传莹, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 北京航空航天大学 交通科学与工程学院,北京 100191, bio={"img":"m5CTbsJZyp6TrGOgN+0ihQ==","content":"

张传莹 (1997—),男,山东潍坊人,硕士研究生,主要研究方向为轨迹预测和碰撞风险评估。E-mail:

"}, bioImg=m5CTbsJZyp6TrGOgN+0ihQ==, bioContent=

张传莹 (1997—),男,山东潍坊人,硕士研究生,主要研究方向为轨迹预测和碰撞风险评估。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1168181787638378681, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=1, ext=[AuthorCompanyExt(id=1168181787663544506, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China), AuthorCompanyExt(id=1168181787671933115, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 北京航空航天大学 交通科学与工程学院,北京 100191)])]), Author(id=1168181788103946437, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, 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=1168181788158472391, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788103946437, language=EN, stringName=Guoyan XU, firstName=Guoyan, middleName=null, lastName=XU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1168181788208804040, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788103946437, language=CN, stringName=徐国艳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 北京航空航天大学 交通科学与工程学院,北京 100191, bio={"content":"

徐国艳 副教授

"}, bioImg=null, bioContent=

徐国艳 副教授

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1168181787638378681, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=1, ext=[AuthorCompanyExt(id=1168181787663544506, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China), AuthorCompanyExt(id=1168181787671933115, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 北京航空航天大学 交通科学与工程学院,北京 100191)])]), Author(id=1168181788267524298, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, 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=1168181788347216076, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788267524298, language=EN, stringName=Zhifa CHEN, firstName=Zhifa, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1168181788397547725, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788267524298, language=CN, stringName=陈志发, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 北京航空航天大学 交通科学与工程学院,北京 100191, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1168181787638378681, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=1, ext=[AuthorCompanyExt(id=1168181787663544506, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China), AuthorCompanyExt(id=1168181787671933115, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 北京航空航天大学 交通科学与工程学院,北京 100191)])]), Author(id=1168181788447879375, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, 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=1168181788502405329, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788447879375, language=EN, stringName=Bin ZHOU, firstName=Bin, middleName=null, lastName=ZHOU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1168181788569514194, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788447879375, language=CN, stringName=周彬, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 北京航空航天大学 交通科学与工程学院,北京 100191, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1168181787638378681, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=1, ext=[AuthorCompanyExt(id=1168181787663544506, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China), AuthorCompanyExt(id=1168181787671933115, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 北京航空航天大学 交通科学与工程学院,北京 100191)])]), Author(id=1168181788619845844, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, 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=1168181788691149014, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788619845844, language=EN, stringName=Liwei CHEN, firstName=Liwei, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 Guizhou Kaiyuan Explosive Engineering Co.,Ltd.,Guiyang Guizhou 551400,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1168181788766646487, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788619845844, language=CN, stringName=陈立伟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 贵州开源爆破工程有限公司,贵州 贵阳 551400, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1168181787739041980, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=2, ext=[AuthorCompanyExt(id=1168181787747430589, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787739041980, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Guizhou Kaiyuan Explosive Engineering Co.,Ltd.,Guiyang Guizhou 551400,China), AuthorCompanyExt(id=1168181787755819198, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787739041980, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 贵州开源爆破工程有限公司,贵州 贵阳 551400)])]), Author(id=1168181788850532569, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, orderNo=5, 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=1168181788934418651, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788850532569, language=EN, stringName=Wei HONG, firstName=Wei, middleName=null, lastName=HONG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 Guizhou Kaiyuan Explosive Engineering Co.,Ltd.,Guiyang Guizhou 551400,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1168181789001527516, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, authorId=1168181788850532569, language=CN, stringName=洪玮, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 贵州开源爆破工程有限公司,贵州 贵阳 551400, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1168181787739041980, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=2, ext=[AuthorCompanyExt(id=1168181787747430589, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787739041980, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Guizhou Kaiyuan Explosive Engineering Co.,Ltd.,Guiyang Guizhou 551400,China), AuthorCompanyExt(id=1168181787755819198, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787739041980, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 贵州开源爆破工程有限公司,贵州 贵阳 551400)])])], keywords=[Keyword(id=1168181789177688285, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, orderNo=1, keyword=extended Kalman filtering (EKF)), Keyword(id=1168181789232214238, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, orderNo=2, keyword=gate recurrent unit (GRU)), Keyword(id=1168181789282545887, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, orderNo=3, keyword=vehicle trajectory), Keyword(id=1168181789349654752, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, orderNo=4, keyword=trajectory prediction), Keyword(id=1168181789404180705, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, orderNo=5, keyword=next generation simulation (NGSIM) dataset), Keyword(id=1168181789454512354, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, orderNo=6, keyword=neural network), Keyword(id=1168181789500649699, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, orderNo=1, keyword=扩展卡尔曼滤波(EKF)), Keyword(id=1168181789546787044, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, orderNo=2, keyword=门控循环单元(GRU)), Keyword(id=1168181789597118693, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, orderNo=3, keyword=车辆轨迹), Keyword(id=1168181789643256038, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, orderNo=4, keyword=轨迹预测), Keyword(id=1168181789710364903, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, orderNo=5, keyword=NGSIM数据集), Keyword(id=1168181789785862376, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, orderNo=6, keyword=神经网络)], refs=[Reference(id=1168181791975289092, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2021, volume=31, issue=10, pageStart=136, pageEnd=143, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=姚志洪, 金玉婷, 王思琛, journalName=中国安全科学学报, refType=null, unstructuredReference=姚志洪, 金玉婷, 王思琛, 等. 混入智能网联汽车的交通流稳定性与安全性分析[J]. 中国安全科学学报, 2021, 31(10):136-143., articleTitle=混入智能网联汽车的交通流稳定性与安全性分析, refAbstract=null), Reference(id=1168181792059175173, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2021, volume=31, issue=10, pageStart=136, pageEnd=143, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=YAO Zhihong, JIN Yuting, WANG Sichen, journalName=China Safety Science Journal, refType=null, unstructuredReference=YAO Zhihong, JIN Yuting, WANG Sichen, et al. Stability and safety analysis on traffic flow mixed with ICV[J]. China Safety Science Journal, 2021, 31(10):136-143., articleTitle=Stability and safety analysis on traffic flow mixed with ICV, refAbstract=null), Reference(id=1168181792117895430, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=11, pageStart=1, pageEnd=9, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=高振海, 鲍明喜, 高菲, journalName=汽车技术, refType=null, unstructuredReference=高振海, 鲍明喜, 高菲, 等. 智能驾驶系统对周边交通车辆预期轨迹的单模态网络预测方法[J]. 汽车技术, 2022(11):1-9., articleTitle=智能驾驶系统对周边交通车辆预期轨迹的单模态网络预测方法, refAbstract=null), Reference(id=1168181792168227079, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=11, pageStart=1, pageEnd=9, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=GAO Zhenhai, BAO Mingxi, GAO Fei, journalName=Automobile Technology, refType=null, unstructuredReference=GAO Zhenhai, BAO Mingxi, GAO Fei, et al. A uni-modal network prediction method for surrounding vehicle expected trajectory in intelligent driving system[J]. Automobile Technology, 2022(11):1-9., articleTitle=A uni-modal network prediction method for surrounding vehicle expected trajectory in intelligent driving system, refAbstract=null), Reference(id=1168181792226947336, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2020, volume=39, issue=8, pageStart=895, pageEnd=935, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=RUDENKO A, PALMIERI L, HERMAN M, journalName=The International Journal of Robotics Research, refType=null, unstructuredReference=RUDENKO A, PALMIERI L, HERMAN M, et al. Human motion trajectory prediction: a survey[J]. The International Journal of Robotics Research, 2020, 39(8):895-935., articleTitle=Human motion trajectory prediction: a survey, refAbstract=null), Reference(id=1168181792319222025, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2019, volume=21, issue=11, pageStart=78, pageEnd=83, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=谢枫, 李永乐, 苏致远, journalName=军事交通学院学报, refType=null, unstructuredReference=谢枫, 李永乐, 苏致远, 等. 一种城市交叉路口转弯车辆轨迹预测方法[J]. 军事交通学院学报, 2019, 21(11):78-83., articleTitle=一种城市交叉路口转弯车辆轨迹预测方法, refAbstract=null), Reference(id=1168181792394719498, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2019, volume=21, issue=11, pageStart=78, pageEnd=83, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=XIE Feng, LI Yongle, SU Zhiyuan, journalName=Journal of Military Transportation University, refType=null, unstructuredReference=XIE Feng, LI Yongle, SU Zhiyuan, et al. A method for predicting turning vehicle trajectory in urban intersection[J]. Journal of Military Transportation University, 2019, 21(11):78-83., articleTitle=A method for predicting turning vehicle trajectory in urban intersection, refAbstract=null), Reference(id=1168181792470216971, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2018, volume=46, issue=2, pageStart=418, pageEnd=423, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=乔少杰, 韩楠, 朱新文, journalName=电子学报, refType=null, unstructuredReference=乔少杰, 韩楠, 朱新文, 等. 基于卡尔曼滤波的动态轨迹预测算法[J]. 电子学报, 2018, 46(2):418-423., articleTitle=基于卡尔曼滤波的动态轨迹预测算法, refAbstract=null), Reference(id=1168181792575074572, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2018, volume=46, issue=2, pageStart=418, pageEnd=423, url=null, language=null, rfNumber=[5], rfOrder=8, authorNames=QIAO Shaojie, HAN Nan, ZHU Xinwen, journalName=Acta Electronica Sinica, refType=null, unstructuredReference=QIAO Shaojie, HAN Nan, ZHU Xinwen, et al. A dynamic trajectory prediction algorithm based on kalman filter[J]. Acta Electronica Sinica, 2018, 46(2):418-423., articleTitle=A dynamic trajectory prediction algorithm based on kalman filter, refAbstract=null), Reference(id=1168181792629600525, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=25, pageEnd=30, url=null, language=null, rfNumber=[6], rfOrder=9, authorNames=QUN L, KRITIKA J, U-XUAN T, journalName=TENCON 2009-2019 IEEE Region 10 Conference (TENCON), refType=null, unstructuredReference=QUN L, KRITIKA J, U-XUAN T. Gaussian process auto regression for vehicle center coordinates trajectory prediction[C]. TENCON 2009-2019 IEEE Region 10 Conference (TENCON), 2019:25-30., articleTitle=Gaussian process auto regression for vehicle center coordinates trajectory prediction, refAbstract=null), Reference(id=1168181792696709390, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=2090, pageEnd=2096, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=CUI Henggang, VLADAN R, FANG-CHIEH C, journalName=International Conference on Robotics and Automation (ICRA), refType=null, unstructuredReference=CUI Henggang, VLADAN R, FANG-CHIEH C, et al. Multimodal trajectory predictions for autonomous driving using deep convolutional networks[C]. International Conference on Robotics and Automation (ICRA), 2019:2090-2096., articleTitle=Multimodal trajectory predictions for autonomous driving using deep convolutional networks, refAbstract=null), Reference(id=1168181792772206863, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=9299, pageEnd=9305, url=null, language=null, rfNumber=[8], rfOrder=11, authorNames=MARCEL S, STEFAN H, KLAUS D, journalName=International Conference on Robotics and Automation (ICRA), refType=null, unstructuredReference=MARCEL S, STEFAN H, KLAUS D. Long-term occupancy grid prediction using recurrent neural networks[C]. International Conference on Robotics and Automation (ICRA), 2019:9299-9305., articleTitle=Long-term occupancy grid prediction using recurrent neural networks, refAbstract=null), Reference(id=1168181792822538512, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2020, volume=63, issue=11, pageStart=1, pageEnd=21, url=null, language=null, rfNumber=[9], rfOrder=12, authorNames=XIE Guo, SHANGGUAN Anqi, FEI Rong, journalName=Science China Information Sciences, refType=null, unstructuredReference=XIE Guo, SHANGGUAN Anqi, FEI Rong, et al. Motion trajectory prediction based on a CNN-LSTM sequential model[J]. Science China Information Sciences, 2020, 63(11):1-21., articleTitle=Motion trajectory prediction based on a CNN-LSTM sequential model, refAbstract=null), Reference(id=1168181792889647377, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=10 563, pageEnd=10 569, url=null, language=null, rfNumber=[10], rfOrder=13, authorNames=CUI Henggang, THI N, FANG-CHIEH C, journalName=IEEE International Conference on Robotics and Automation (ICRA), refType=null, unstructuredReference=CUI Henggang, THI N, FANG-CHIEH C, et al. Deep kinematic models for kinematically feasible vehicle trajectory predictions[C]. IEEE International Conference on Robotics and Automation (ICRA), 2020:10 563-10 569., articleTitle=Deep kinematic models for kinematically feasible vehicle trajectory predictions, refAbstract=null), Reference(id=1168181792944173330, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2018, volume=56, issue=5, pageStart=77, pageEnd=80, url=null, language=null, rfNumber=[11], rfOrder=14, authorNames=吕梦男, 孙涛, 李洁, journalName=农业装备与车辆工程, refType=null, unstructuredReference=吕梦男, 孙涛, 李洁. 基于扩展卡尔曼滤波的车辆状态参数估计[J]. 农业装备与车辆工程, 2018, 56(5):77-80., articleTitle=基于扩展卡尔曼滤波的车辆状态参数估计, refAbstract=null), Reference(id=1168181792998699283, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2018, volume=56, issue=5, pageStart=77, pageEnd=80, url=null, language=null, rfNumber=[11], rfOrder=15, authorNames=LYU Mengnan, SUN Tao, LI Jie, journalName=Agricultural Equipment & Vehicle Engineering, refType=null, unstructuredReference=LYU Mengnan, SUN Tao, LI Jie. Vehicle state parameter estimation based on extended kalman filter[J]. Agricultural Equipment & Vehicle Engineering, 2018, 56(5):77-80., articleTitle=Vehicle state parameter estimation based on extended kalman filter, refAbstract=null), Reference(id=1168181793074196756, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2022, volume=32, issue=5, pageStart=170, pageEnd=176, url=null, language=null, rfNumber=[12], rfOrder=16, authorNames=熊晓夏, 刘擎超, 沈钰杰, journalName=中国安全科学学报, refType=null, unstructuredReference=熊晓夏, 刘擎超, 沈钰杰, 等. 基于LSTM-BF的高速公路交通事故风险模型[J]. 中国安全科学学报, 2022, 32(5):170-176., articleTitle=基于LSTM-BF的高速公路交通事故风险模型, refAbstract=null), Reference(id=1168181793124528405, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=2022, volume=32, issue=5, pageStart=170, pageEnd=176, url=null, language=null, rfNumber=[12], rfOrder=17, authorNames=XIONG Xiaoxia, LIU Qingchao, SHEN Yujie, journalName=China Safety Science Journal, refType=null, unstructuredReference=XIONG Xiaoxia, LIU Qingchao, SHEN Yujie, et al. Study on risk model of highway traffic accidents based on LSTM-BF[J]. China Safety Science Journal, 2022, 32(5):170-176., articleTitle=Study on risk model of highway traffic accidents based on LSTM-BF, refAbstract=null), Reference(id=1168181793183248662, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=https://www.fhwa.dot.gov/publications/research/operations/06137/, language=null, rfNumber=[13], rfOrder=18, authorNames=COLYAR J, journalName=null, refType=null, unstructuredReference=COLYAR J. Us highway i-80 dataset[EB/0L].[2023-09-15]. https://www.fhwa.dot.gov/publications/research/operations/06137/., articleTitle=Us highway i-80 dataset, refAbstract=null), Reference(id=1168181793246163223, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=https://www.fhwa.dot.gov/publications/research/operations/07030/, language=null, rfNumber=[14], rfOrder=19, authorNames=COLYAR J, journalName=null, refType=null, unstructuredReference=COLYAR J. Us highway 101 dataset[EB/0L].[2023-09-15]. https://www.fhwa.dot.gov/publications/research/operations/07030/., articleTitle=Us highway 101 dataset, refAbstract=null)], funds=[Fund(id=1168181791811711235, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, awardId=2022YFB4703702, language=CN, fundingSource=国家重点研发计划课题项目(2022YFB4703702), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1168181787638378681, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=1, ext=[AuthorCompanyExt(id=1168181787663544506, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China), AuthorCompanyExt(id=1168181787671933115, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787638378681, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 北京航空航天大学 交通科学与工程学院,北京 100191)]), AuthorCompany(id=1168181787739041980, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, xref=2, ext=[AuthorCompanyExt(id=1168181787747430589, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787739041980, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Guizhou Kaiyuan Explosive Engineering Co.,Ltd.,Guiyang Guizhou 551400,China), AuthorCompanyExt(id=1168181787755819198, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, companyId=1168181787739041980, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 贵州开源爆破工程有限公司,贵州 贵阳 551400)])], figs=[ArticleFig(id=1168181789949440233, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.1, caption=Bicycle model, figureFileSmall=cPvFxnHyPdEnroGNDqxHCQ==, figureFileBig=mEOXOnLURUqawh0b9Js/UQ==, tableContent=null), ArticleFig(id=1168181790003966186, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图1, caption=自行车模型, figureFileSmall=cPvFxnHyPdEnroGNDqxHCQ==, figureFileBig=mEOXOnLURUqawh0b9Js/UQ==, tableContent=null), ArticleFig(id=1168181790050103531, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.2, caption=Process of EKF-CTRV trajectory prediction method, figureFileSmall=eUlAo6Wa+xw25uENN2gOnA==, figureFileBig=NkgnU+JMmaY5it7dW3uQPQ==, tableContent=null), ArticleFig(id=1168181790092046572, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图2, caption=基于EKF-CTRV轨迹预测方法流程, figureFileSmall=eUlAo6Wa+xw25uENN2gOnA==, figureFileBig=NkgnU+JMmaY5it7dW3uQPQ==, tableContent=null), ArticleFig(id=1168181790142378221, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.3, caption=GRU, figureFileSmall=tPE3jBT6cfOVrz2rAqtsaQ==, figureFileBig=n2JvdJ/d+G3Ob9UEpKjbvw==, tableContent=null), ArticleFig(id=1168181790201098478, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图3, caption=GRU

注:hk为当前时刻的隐状态;Xk+1为下一时刻的输入;Rk+1为下一时刻的重置门状态;Zk+1为下一时刻的更新门状态; h ˙ k + 1为下一时刻的候选隐状态;hk+1为下一时刻的隐状态;σ为sigmoid激活函数;tanh为双曲正切函数激活函数;☉为点乘;+为点加。

, figureFileSmall=tPE3jBT6cfOVrz2rAqtsaQ==, figureFileBig=n2JvdJ/d+G3Ob9UEpKjbvw==, tableContent=null), ArticleFig(id=1168181790276595951, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.4, caption=Architecture of deep kinematic-based GRU for trajectory prediction, figureFileSmall=OjBnPSbinZZCAg7C5FLzyA==, figureFileBig=5gfOtRSlsJgrwmz1DlGsFA==, tableContent=null), ArticleFig(id=1168181790347899120, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图4, caption=基于深层运动学的GRU轨迹预测方法架构, figureFileSmall=OjBnPSbinZZCAg7C5FLzyA==, figureFileBig=5gfOtRSlsJgrwmz1DlGsFA==, tableContent=null), ArticleFig(id=1168181790398230769, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.5, caption=I-80 collection environment, figureFileSmall=VOsF/2hVkketBsvXIjBvFw==, figureFileBig=G4LGTmNWO1dthwsz7IFwRw==, tableContent=null), ArticleFig(id=1168181790448562418, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图5, caption=I-80采集环境, figureFileSmall=VOsF/2hVkketBsvXIjBvFw==, figureFileBig=G4LGTmNWO1dthwsz7IFwRw==, tableContent=null), ArticleFig(id=1168181790549225715, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.6, caption=US-101 collection environment, figureFileSmall=ONiJY2w8HtdJS1AW4UVWQw==, figureFileBig=AvNocOMsV64aIMnK2RPeug==, tableContent=null), ArticleFig(id=1168181790624723188, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图6, caption=US-101采集环境, figureFileSmall=ONiJY2w8HtdJS1AW4UVWQw==, figureFileBig=AvNocOMsV64aIMnK2RPeug==, tableContent=null), ArticleFig(id=1168181790691832053, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.7, caption=Comparison of velocity and acceleration processing results, figureFileSmall=wUW43xC62bHzpGvR6bVpbw==, figureFileBig=s9xB9gP4TXpudWFy1uBiIg==, tableContent=null), ArticleFig(id=1168181790809272566, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图7, caption=速度、加速度处理结果对比, figureFileSmall=wUW43xC62bHzpGvR6bVpbw==, figureFileBig=s9xB9gP4TXpudWFy1uBiIg==, tableContent=null), ArticleFig(id=1168181790888964343, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.8, caption=Calculation and smoothing filter effect of heading angle and roll angle velocity, figureFileSmall=NNadCP8lDVFNV+Ubh0swtA==, figureFileBig=XJsKTP5rWfMpgoh2UbrVXw==, tableContent=null), ArticleFig(id=1168181790939295992, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图8, caption=航向角、横摆角速度计算及平滑过滤效果, figureFileSmall=NNadCP8lDVFNV+Ubh0swtA==, figureFileBig=XJsKTP5rWfMpgoh2UbrVXw==, tableContent=null), ArticleFig(id=1168181791002210553, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Fig.9, caption=EKF-GRU trajectory prediction results, figureFileSmall=2X+Xn1uBnnlIncC6/j07IA==, figureFileBig=XKHiS7nSlYy3fHsqgda5JQ==, tableContent=null), ArticleFig(id=1168181791056736506, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=图9, caption=EKF-GRU轨迹预测结果, figureFileSmall=2X+Xn1uBnnlIncC6/j07IA==, figureFileBig=XKHiS7nSlYy3fHsqgda5JQ==, tableContent=null), ArticleFig(id=1168181791128039675, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Table 1, caption=

Encoder-Decoder network architecture

, figureFileSmall=null, figureFileBig=null, tableContent=
网络架构:
层级: 4×1
输入层: 1-D, X R 128 × 30
GRU层: 80Units h k R 128 × 256
W x r W x z W x h R 30 × 256 W h r W h z W h h R 256 × 256 b r b z b h R 1 × 256
全连接层: W F C R 50 × 50
输出层: Regression
), ArticleFig(id=1168181791199342844, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=表1, caption=

Encoder-Decoder网络架构

, figureFileSmall=null, figureFileBig=null, tableContent=
网络架构:
层级: 4×1
输入层: 1-D, X R 128 × 30
GRU层: 80Units h k R 128 × 256
W x r W x z W x h R 30 × 256 W h r W h z W h h R 256 × 256 b r b z b h R 1 × 256
全连接层: W F C R 50 × 50
输出层: Regression
), ArticleFig(id=1168181791258063101, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
算法1 EKF- GRU轨迹预测
需求: Q R
for k from 0 to 49 step 1 do
Z ( a ω ) k + 1 i R ( σ a 2 σ ω 2 ) k + 1 i G R U ( X ^ k = ( k - 29 ) ~ k i )
X ^ k + 1 i P ^ k + 1 i E K F ( Z ( a ω ) k + 1 i X ^ k i P ^ k i Q R ( σ a 2 σ ω 2 ) k + 1 i )
end for
return ( X ^ k + 1 i P ^ k + 1 i ) k = 0 ~ 49 i = 1 ~ N
), ArticleFig(id=1168181791333560574, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=, caption=

, figureFileSmall=null, figureFileBig=null, tableContent=
算法1 EKF- GRU轨迹预测
需求: Q R
for k from 0 to 49 step 1 do
Z ( a ω ) k + 1 i R ( σ a 2 σ ω 2 ) k + 1 i G R U ( X ^ k = ( k - 29 ) ~ k i )
X ^ k + 1 i P ^ k + 1 i E K F ( Z ( a ω ) k + 1 i X ^ k i P ^ k i Q R ( σ a 2 σ ω 2 ) k + 1 i )
end for
return ( X ^ k + 1 i P ^ k + 1 i ) k = 0 ~ 49 i = 1 ~ N
), ArticleFig(id=1168181791446806783, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
算法2 数据平滑过滤过程
需求: Q R
设置: X ^ - 30 P ^ - 30
for k from-30 to -1 step 1 do
X ^ k + 1 P ^ k + 1 K F ( Z k + 1 X ^ k P ^ k Q R )
end for
X = M A ( X ^ )
), ArticleFig(id=1168181791513915648, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=, caption=

, figureFileSmall=null, figureFileBig=null, tableContent=
算法2 数据平滑过滤过程
需求: Q R
设置: X ^ - 30 P ^ - 30
for k from-30 to -1 step 1 do
X ^ k + 1 P ^ k + 1 K F ( Z k + 1 X ^ k P ^ k Q R )
end for
X = M A ( X ^ )
), ArticleFig(id=1168181791585218817, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=EN, label=Table 2, caption=

Comparison of prediction results between baseline method EKF-CTRV and extended method EKF-GRU

, figureFileSmall=null, figureFileBig=null, tableContent=
预测步长/s 1 2 3 4 5
FDE EKF-CTRV 0.61 1.76 2.98 4.65 6.48
EKF-GRU* 0.61 1.51 2.52 4.09 5.45
RMSE EKF-CTRV 0.81 2.03 3.62 5.84 7.69
EKF-GRU* 0.76 1.78 3.18 4.73 6.67
ADE EKF-CTRV 3.03
EKF-GRU* 2.56
), ArticleFig(id=1168181791664910594, tenantId=1146029695717560320, journalId=1146031787341344770, articleId=1149743084524712289, language=CN, label=表2, caption=

基线方法EKF-CTRV和扩展方法EKF-GRU的预测结果对比

, figureFileSmall=null, figureFileBig=null, tableContent=
预测步长/s 1 2 3 4 5
FDE EKF-CTRV 0.61 1.76 2.98 4.65 6.48
EKF-GRU* 0.61 1.51 2.52 4.09 5.45
RMSE EKF-CTRV 0.81 2.03 3.62 5.84 7.69
EKF-GRU* 0.76 1.78 3.18 4.73 6.67
ADE EKF-CTRV 3.03
EKF-GRU* 2.56
)], attaches=null, journal=Journal(id=1123942128916217864, delFlag=0, nameCn=中国安全科学学报, nameEn=China Safety Science Journal, nameHistory1=null, nameHistory2=null, issn=1003-3033, eissn=, cn=11-2865/X, coden=null, periodic=0, language=CN, oaType=0, 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=fkqsFM6VKlHC4gCtS5XqTw==, journalPrice=null, startedYear=null, abbrevIsoEn=Chin Saf Sci J, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1754269350027, createdBy=null, updatedBy=13701087609, firstLetterCn=C, firstLetterEn=C, subjectCode=Engineering, subjectName=工程, subjectCodeEn=Engineering, subjectNameEn=null, picCn=fkqsFM6VKlHC4gCtS5XqTw==, picEn=SHn9HgqSxtJrOcAxqD++4Q==, jcr=null, cjcr=null, exts=[JournalExt(id=1159052918994595848, language=CN, name=中国安全科学学报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.cssjj.com.cn/, createdTime=1754269350050, updatedTime=1754269350050, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.cssjj.com.cn/CN/column/item15.shtml, submissionAuthorUrl=https://zgaqkxxbauthor.manuscriptcloud.com/login, submissionEditorUrl=https://zgaqkxxbeditor.manuscriptcloud.com/login, submissionReviewUrl=https://zgaqkxxbauthor.manuscriptcloud.com/login, submissionCeEditorUrl=https://zgaqkxxbeditor.manuscriptcloud.com/login, submissionAeEditorUrl=https://zgaqkxxbeditor.manuscriptcloud.com/login, option={"copyright":""}), JournalExt(id=1159052919040733193, language=EN, name=China Safety Science Journal, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.cssjj.com.cn/EN/1003-3033/home.shtml, createdTime=1754269350061, updatedTime=1754269350061, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=https://synbioj.cip.com.cn/EN/column/column3.shtml, submissionAuthorUrl=https://zgaqkxxbauthor.manuscriptcloud.com/login, submissionEditorUrl=https://zgaqkxxbeditor.manuscriptcloud.com/login, submissionReviewUrl=https://zgaqkxxbauthor.manuscriptcloud.com/login, submissionCeEditorUrl=https://zgaqkxxbeditor.manuscriptcloud.com/login, submissionAeEditorUrl=https://zgaqkxxbeditor.manuscriptcloud.com/login, option={"copyright":""})], databaseList=null, tenantJournalId=1146031787341344770, websiteList=[Website(id=1148243202345263519, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146031787341344770, 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/zgaqkxxb/CN, language=CN, createTime=1751692112766, createBy=18614031015, updateTime=1753502583634, updateBy=18614031015, name=《中国安全科学学报》中文站点, tplId=1146099689490845704, title=中国安全科学学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148618794941046792, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202345263519, code=articleTextType, value=kx, createTime=1751781661020, updateTime=1751781661020, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618794911686661, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202345263519, code=banner, value=null, createTime=1751781661012, updateTime=1751781661012, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618794894909444, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202345263519, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=tui0IVO9FMwB61HHtX5scg==, createTime=1751781661008, updateTime=1751781661008, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618794932658183, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202345263519, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751781661017, updateTime=1751781661017, creator=18614031015, updator=18614031015), WebsiteProps(id=1148618794924269574, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202345263519, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751781661015, updateTime=1751781661015, creator=18614031015, updator=18614031015)]), Website(id=1155836763751993353, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146031787341344770, 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/zgaqkxxb/EN, language=EN, createTime=1753502558893, createBy=18614031015, updateTime=1753524450387, updateBy=18614031015, name=《中国安全科学学报》英文站点, tplId=1146101810881728533, title=China Safety Science Journal, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155895925743669425, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155836763751993353, code=articleTextType, value=kx, createTime=1753516664205, updateTime=1753516664205, creator=18614031015, updator=18614031015), WebsiteProps(id=1155895925722697902, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155836763751993353, code=banner, value=null, createTime=1753516664200, updateTime=1753516664200, creator=18614031015, updator=18614031015), WebsiteProps(id=1155895925714309293, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155836763751993353, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=tui0IVO9FMwB61HHtX5scg==, createTime=1753516664198, updateTime=1753516664198, creator=18614031015, updator=18614031015), WebsiteProps(id=1155895925735280816, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155836763751993353, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753516664203, updateTime=1753516664203, creator=18614031015, updator=18614031015), WebsiteProps(id=1155895925731086511, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155836763751993353, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753516664202, updateTime=1753516664202, creator=18614031015, updator=18614031015)])], journalTitle=中国安全科学学报, weixinUrl=null, journalUrl=null, iacademicId=null, status=0, seqNo=null, journalTitleEn=China Safety Science Journal, journalPhotoCn=fkqsFM6VKlHC4gCtS5XqTw==, journalPhotoEn=SHn9HgqSxtJrOcAxqD++4Q==, journalFirstLetter=C, journalRecommend=null, journalNew=null, journalCollection=1, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=null, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/zgaqkxxb/CN/10.16265/j.cnki.issn1003-3033.2024.06.0720, detailUrlEn=https://castjournals.cast.org.cn/joweb/zgaqkxxb/EN/10.16265/j.cnki.issn1003-3033.2024.06.0720, pdfUrlCn=https://castjournals.cast.org.cn/joweb/zgaqkxxb/CN/PDF/10.16265/j.cnki.issn1003-3033.2024.06.0720, pdfUrlEn=https://castjournals.cast.org.cn/joweb/zgaqkxxb/EN/PDF/10.16265/j.cnki.issn1003-3033.2024.06.0720, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于EKF-GRU的车辆轨迹预测
收藏切换
PDF下载
张传莹 1 , 徐国艳 1 , 陈志发 1 , 周彬 1 , 陈立伟 2 , 洪玮 2
中国安全科学学报 | 安全工程技术 2024,34(6): 164-172
收起
收藏切换
中国安全科学学报 | 安全工程技术 2024, 34(6): 164-172
基于EKF-GRU的车辆轨迹预测
全屏
张传莹1 , 徐国艳1, 陈志发1, 周彬1, 陈立伟2, 洪玮2
作者信息
  • 1 北京航空航天大学 交通科学与工程学院,北京 100191
  • 2 贵州开源爆破工程有限公司,贵州 贵阳 551400
  • 张传莹 (1997—),男,山东潍坊人,硕士研究生,主要研究方向为轨迹预测和碰撞风险评估。E-mail:

    徐国艳 副教授

Vehicle trajectory prediction based on EKF-GRU
Chuanying ZHANG1 , Guoyan XU1, Zhifa CHEN1, Bin ZHOU1, Liwei CHEN2, Wei HONG2
Affiliations
  • 1 School of Transportation Science and Engineering,Beihang University,Beijing 100191,China
  • 2 Guizhou Kaiyuan Explosive Engineering Co.,Ltd.,Guiyang Guizhou 551400,China
出版时间: 2024-06-28 doi: 10.16265/j.cnki.issn1003-3033.2024.06.0720
文章导航
收藏切换

为提升行车安全,实现自动驾驶车辆正确的决策规划,提出基于扩展卡尔曼滤波(EKF)-门控循环单元(GRU)的车辆轨迹预测方法,结合学习方法与物理模型,在提升预测精度的同时,提高轨迹预测的合理性。首先,基于GRU构建预测网络,通过提取车辆的历史轨迹特征预测车辆的纵向加速度及横摆角速度;其次,基于车辆非线性运动学构建EKF状态估计器,结合观测值生成车辆未来有限时域的行驶轨迹;最后,在高速公路多车轨迹数据集NGSIM I-80和US-101上进行轨迹预测方法验证。结果表明:采用传统的物理模型生成预测轨迹,其最终距离误差(FDE)、均方根误差(RMSE)、平均距离误差(ADE)值分别为6.48、7.69和3.03 m。相比之下,利用EKF-GRU生成的预测轨迹表现出更高的准确性,对应的数值分别为5.45、6.67和2.56 m,分别提升15.90%、13.26%和15.51%。

扩展卡尔曼滤波(EKF)  /  门控循环单元(GRU)  /  车辆轨迹  /  轨迹预测  /  NGSIM数据集  /  神经网络

To enhance the driving safety and achieve correct decision planning for autonomous vehicles,a safe driving trajectory prediction method based on EKF-GRU was proposed. By combining learning-based methods with physics-based approaches,the prediction accuracy was improved and the rationality of the predicted trajectories was enhanced. In the first step of this method,a prediction network was constructed based on GRU to predict the longitudinal acceleration and yaw angular velocity of vehicles by extracting historical trajectory features. In the second step,an EKF state estimator was built based on the nonlinear vehicle kinematics to generate the vehicle's future limited-time trajectory,incorporating the observations obtained previously. The trajectory prediction method was validated on the NGSIM I-80 and US-101 multi-vehicle trajectory datasets. Experimental results demonstrate that the final distance errors (FDE),root mean square errors (RMSE),and average distance errors (ADE) of the predicted trajectories generated by traditional physics-based methods are 6.48,7.69 and 3.03 meters,respectively. In contrast,trajectories predicted using EKF-GRU exhibit higher accuracy,and the corresponding values are 5.45,6.67 and 2.56 meters,respectively. This represents improvements of 15.90%,13.26% and 15.51%.

extended Kalman filtering (EKF)  /  gate recurrent unit (GRU)  /  vehicle trajectory  /  trajectory prediction  /  next generation simulation (NGSIM) dataset  /  neural network
张传莹, 徐国艳, 陈志发, 周彬, 陈立伟, 洪玮. 基于EKF-GRU的车辆轨迹预测. 中国安全科学学报, 2024 , 34 (6) : 164 -172 . DOI: 10.16265/j.cnki.issn1003-3033.2024.06.0720
Chuanying ZHANG, Guoyan XU, Zhifa CHEN, Bin ZHOU, Liwei CHEN, Wei HONG. Vehicle trajectory prediction based on EKF-GRU[J]. China Safety Science Journal, 2024 , 34 (6) : 164 -172 . DOI: 10.16265/j.cnki.issn1003-3033.2024.06.0720
当前自动驾驶技术快速发展,受多种因素约束,自动驾驶车辆会长期与有人驾驶车辆混合行驶[1]。为保障复杂混行交通环境下行车安全,自动驾驶车辆必须准确理解周围动态交通环境并作出合理驾驶决策[2],而准确预测周围其他有人驾驶车辆行驶轨迹是其关键技术之一。自动驾驶车辆根据预测轨迹预见未来可能的冲突风险,从而作出更加合理的决策,规划安全路径,减少碰撞风险。
目前,国内外学者深入研究了车辆轨迹预测问题,主要分为基于物理模型和学习的方法[3]。基于物理模型的轨迹预测方法以车辆动力学或运动学模型为基础,利用对车辆部分状态的历史观测推演车辆未来的轨迹。谢枫等[4]提出一种基于三阶贝塞尔曲线和运动学模型的转弯车辆轨迹预测方法,相较于恒加速度恒横摆角速度模型提高了算法的时效性和准确性;乔少杰等[5]使用扩展卡尔曼滤波(Extended Kalman Filter,EKF)预测移动对象的轨迹,得到较高的预测精度;LIM等[6]结合车辆中心坐标历史数据,建立基于高斯过程回归的轨迹预测方法,相较于传统的恒速度(Constant Velocity,CV)模型及卡尔曼滤波(Kalman Filtering,KF)模型能够获得更好的结果。基于物理模型的方法不依赖数据,可解释性强,但仅依据车辆历史状态信息的轨迹预测方法对于车辆较长时间轨迹的预测能力较差。近年来,越来越多机器学习技术被应用到车辆轨迹预测领域,主要包括基于卷积神经网络(Convolutional Neural Network,CNN)和循环神经网络(Recurrent Neural Network,RNN)2类方法。CUI Henggang等[7]将车辆周围环境编码成光栅图像,作为深度卷积网络的输入,使用CNN输出最大似然轨迹簇及其概率分布。上述仅基于CNN的轨迹预测方法未考虑目标车辆状态信息在时间维度的关联性,因此,有学者设计RNN预测车辆轨迹。MARCEL等[8]提出一种基于激光雷达网格融合和RNN的预测架构,利用网格单元之间的时空相关性提高轨迹预测精度。XIE Guo等[9]结合CNN空间扩展和长短时记忆网络(Long Short-Term Memory,LSTM)时间扩展的特点,设计基于时间序列的预测模型计算周围车辆轨迹,结果表明:相较于CNN、LSTM和门控循环单元(Gated Recurrent Unit,GRU),CNN-LSTM均方根误差更低,时间成本更短。基于深度学习的轨迹预测方法直接从大量动态场景数据中提取可能影响目标车辆轨迹的特征,可能导致预测轨迹不符合车辆运动学约束[10]。然而,基于学习的方法大多依赖高精地图且网络模型较为复杂,使得同时预测周围多个车辆轨迹时耗费大量计算资源,影响实时性。此外,基于数据驱动的轨迹预测方法易产生车辆无法跟随的次优轨迹,从而影响车辆的正确决策,产生安全风险。
鉴于此,笔者拟综合物理模型与学习类算法优点,设计基于GRU与运动学模型相结合的车辆轨迹预测方法,在避免产生次优轨迹的同时,提高预测轨迹准确性。以期使生成轨迹更符合车辆的运动学特性,保障车辆决策安全性。
车辆轨迹预测是根据车辆自身历史运行状态估计未来行驶轨迹。在当前时间步内,获取目标车辆历史观测时域th的状态信息,设计轨迹预测方法,预测车辆未来预测时域tp的行驶轨迹。为使生成的轨迹更符合实际,预测算法的输入数据能准确表示车辆的横纵向位置及其变化趋势。文中所设计GRU的输入特征包括车辆位置 ( x y )、航向角 φ以及纵向速度 v,特征向量如下式:
X = [ x y φ v ]
基于车辆在短时间内的行为具有连续性的特点,获取历史3s的车辆状态作为输入数据。从任务需求角度,考虑到自动驾驶车辆需要及时应对突发情况和交通状况的变化,决策规划任务应在短时间内作出安全决策,规划更合理的车辆行驶路径;从安全角度考虑,预留4s的紧急制动响应时间,因此,生成车辆未来5s的轨迹。即th = 3s,tp = 5s。
基于车辆运动学模型的轨迹预测方法,对车辆未来预测时域的速度及横摆角速度进行恒速度恒横摆角速度(Constant Turn Rate and Velocity,CTRV)假设,使用EKF作为非线性车辆运动学模型的状态估计器,生成最终轨迹。
使用自行车模型作为车辆运动学模型,自行车模型如图1所示。车辆的位置以车辆中心点 C ( x y )表示,左右轮等效为单个轮子,前轮中心点为A点,后轮等效后的中心点为BO1为车辆瞬时旋转中心。假设车辆平动,车辆运动状态可用 ( x y φ )描述,其中,r为车辆转弯半径,δ为前轮转角。
自行车模型是一个简单的运动学模型,但可有效描述车辆的运动状态转移。在控制领域,常采用自行车模型中的位置、航向角、速度、加速度以及横摆角速度作为状态向量 X = [ x y φ v a ω ] T,近似描述车辆运动学。
基于自行车模型构建离散形式的车辆非线性状态转移方程,如下式:
X k + 1 = f X k + q k = [ f 1 f 2 f 3 f 4 f 5 f 6 ] + q k
f 1 : x k + 1 = x k + ( v k c o s φ k ) T + ( a k c o s φ k + ω k v k s i n φ k ) T 2 / 2
f 2 : y k + 1 = y k + ( v k s i n φ k ) T + ( a k s i n φ k + ω k v k c o s φ k ) T 2 / 2
f 3 : φ k + 1 = φ k + ω k T
f 4 : v k + 1 = v k + a k T
f 5 : ω k + 1 = ω k
f 6 : a k + 1 = a k
式中: q k为状态转移方程过程噪声, q k = N ( 0 D k ) D k为协方差矩阵, D k = d i a g { 0,0 q d φ q d v q d ω q d a },d为微分;T为计算周期;k为当前时刻。
KF是一种用于描述线性高斯系统的模型,而车辆运动学模型为非线性系统。当前大多采用EKF和无迹卡尔曼滤波(Unscented Kalman Filter,UKF)等解决非线性系统的KF线性化表达问题。其中,EKF采用一阶泰勒展开对非线性系统的函数线性化逼近[11],随后通过KF算法进行状态估计,该方法更适用于低维状态空间的问题。UKF使用一组sigma点,通过非线性变换得到其均值和协方差来估计非线性系统,该方法适用于处理高维状态空间问题。采用上述状态向量X,并选择EKF处理该非线性系统。
EKF包括预测和更新2步,具体过程如下:
预测:
X ^ - k + 1 = f X ^ k
P ^ k + 1 - = F k P ^ k F k T + Q k
更新:
K k + 1 = P ^ k + 1 - H T ( H P ^ - k + 1 H T + R k + 1 ) - 1
X ^ k + 1 = X ^ k + 1 - + K k + 1 ( Z k + 1 - H X ^ k + 1 - )
P ^ k + 1 = ( I - K k + 1 H ) P ^ k + 1 -
式中: X ^ k + 1 -为下一时刻的先验估计状态; X ^ k为当前时刻的后验估计状态; f ( · )为非线性状态转移方程; P ^ k + 1 -为下一时刻的先验协方差矩阵; F k为当前时刻的雅可比矩阵; P ^ k + 1为下一时刻的后验协方差矩阵; Q k为当前时刻的过程噪声; R k + 1为下一时刻的观测噪声; K k + 1为下一时刻的卡尔曼增益; H为观测矩阵。
基于CTRV假设,以当前时刻速度 v k = 0及横摆角速度 ω k = 0作为未来5s内的速度 v k = 1 ~ 50与横摆角速度 ω k = 1 ~ 50。由此构建观测方程为:
Z k = H X k + r k = 0   0   0   1   0   0   0 0   0   0   0   1   0   0 X k + r k = ( v k ω k ) T k = 1 ~ 50
式中: v k ω k为观测状态; r k ~ N ( 0 R k )为虚拟传感器的测量噪声,用于评估速度和横摆角速度观测值的准确度,其中, R k = d i a g { σ v 2 σ ω 2 }。考虑观测误差的问题,使用随时间线性增加的速度和横摆角速度的方差,作为噪声模型。
将上述观测方程代入EKF的预测和更新过程,即可预测车辆未来5s的轨迹。基于EKF-CTRV轨迹预测方法流程如图2所示。
基于数据驱动的轨迹预测方式能够充分提取车辆的历史状态信息,生成更加符合行驶“意图”的车辆轨迹,应用较广的方法包括RNN、LSTM以及GRU。它们都能够处理时间序列信息,但RNN会受到短时记忆的影响,难以将较早的信息传递到后面的时间步。LSTM和GRU都通过指示网络解决了历史信息的遗忘时间及新信息的更新时间RNN的“梯度消失”问题[12],其中,GRU采用张量运算更少的“门控”单元,训练速度较LSTM更快。综上,文中使用GRU提取车辆历史状态特征。
GRU包含更新门和重置门2个部分,使用隐藏状态进行信息传递,具体结构如图3所示。
GRU的工作过程如下。
重置门:
R k + 1 = σ ( X k + 1 W x r + h k W h r + b r )
更新门:
Z k + 1 = σ ( X k + 1 W x z + h k W h z + b z )
候选隐状态:
h ~ k + 1 = t a n h ( X k + 1 W x h + ( R k + 1 h k ) W h h + b h )
隐状态:
h k + 1 = Z k + 1 h k + ( 1 - Z k + 1 ) h ~ k + 1
其中,假设输入一个批量为n,序列长度为d X k + 1 R n × d,假设隐藏单元个数为h,则 W x r W x z W x h R d × h W h r W h z W h h R h × h是权重参数, b r b z b h R 1 × h是偏置参数。
在时序预测中,GRU对输入的扰动非常敏感,需要引入EKF来改善GRU的表现。因此,提出一种基于EKF-GRU的轨迹预测方法。该方法将构建的非线性运动学模型作为网络的最后一层集成到解码器中,基于EKF完成轨迹预测任务,从而实现融合车辆的数据特征与运动学特征。具体来讲,该方法使用GRU提取车辆长期状态特征,并改进车辆运动学模型的不确定性;同时,使用EKF跟踪系统状态并处理非线性误差。EKF-GRU网络能够同时处理车辆长期的意图依赖性和短期的状态依赖性,因此,在轨迹预测方面具有良好的性能。
基于GRU构建一种Encoder-Decoder网络架构,该网络通过GRU将车辆历史状态编码作为网络隐状态,再通过GRU与全连接层(Fully Connected Layer,FC)将隐状态解码为车辆中间状态向量,同时产生中间状态向量的标准差以描述其准确性。将车辆中间状态向量作为观测值输入至最后一层的运动学模型以生成车辆轨迹。结合EKF与GRU,将基于循环神经网络的轨迹预测问题转换为基于EKF的状态估计问题,从而解决预测轨迹长期预测性差的问题。Encoder-Decoder网络架构见表1。网络整体架构如图4所示。
采用GRU基于目标车辆历史状态信息预测车辆的加速度及横摆角速度 [ a ω ],并使用加速度及横摆角速度的标准差 [ σ a σ ω ]描述两者准确性。利用EKF-GRU方法,状态 X迭代过程如下:
步骤1:在每个时间 t 0,坐标原点即为当前车辆位置 ( x 0 y 0 ),方向固定。将历史3s(采样频率10Hz)观测序列状态向量 X k = - 29 ~ 0作为网络输入。
步骤2:利用GRU处理历史3s轨迹的时间序列特征,得到未来5s的加速度 a k、横摆角速度 ω k以及二者的标准差 σ a k σ ω k,将加速度及横摆角速度作为观测值 Z k = ( a k ω k ) T,同时将二者的方差作为观测噪声 R k = d i a g { σ a k 2 σ ω k 2 }
步骤3:将 a k ω k σ a k以及 σ ω k代入到EKF中得到下一时刻的后验位置 ( x k y k )及协方差矩阵,其中,k=1~50。
整理步骤2—步骤3的伪代码:
GRU函数具有车辆状态 Z = ( a ω ) T和观测噪声 R = d i a g { σ a 2 σ ω 2 }这2个输出。车辆物理模型与GRU结合的轨迹预测网络将ZR代入到更新步骤中,作为观测值。
为验证所提出方法的有效性,使用NGSIM数据集中I-80和US-101训练和验证模型。在高速公路数据集,如图5的NGSIM I-80[13]图6的US-101[14]上进行验证。
KF根据系统模型和观测数据动态估计系统的状态,因此,能够预测系统的未来状态,并且在存在噪声或不确定性时,提供更为准确的估计结果。然而,KF并不能完全消除噪声或不确定性,采用移动平均滤波(Moving Average Filter,MA),通过平均信号来减小随机噪声的影响,可进一步平滑KF的输出,从而减小估计的噪声和抖动,提高估计的稳定性。同时,MA可减少KF的计算量,因此,提高滤波器的实时性能。但MA会引入一定的延迟,需要合理选择MA参数。采用KF+MA平滑过滤历史3s的状态量的数据,整个数据平滑过滤过程的伪代码如下:
EKF-CTRV以及EKF-GRU方法基于车辆状态转移方程进行轨迹预测。NGSIM数据集提供了位置、速度以及加速度信息,其中,位置数据已滤波处理。考虑到数据集采集过程中可能因感知精度等的影响而存在误差,需要平滑过滤处理速度和加速度,而NGSIM数据集缺少关键的航向角、横摆角速度信息。基于车辆位置计算出车辆的航向角,通过航向角微分计算横摆角速度,同样使用算法2平滑过滤计算结果的数据。数据平滑过滤结果如图7图8所示。原始数据中速度与加速度在短时间内的波动较大,相较于KF方法,KF+MA的处理方案将数据处理的更加平滑,如9~10s内,速度无明显的变化,加速度也保持一个稳定且趋近于0的状态。KF+MA消除了航向角计算中的突变问题,结合图8b所示横摆角速度,原始数据以及KF方法处理的横摆角速度变化范围较大,经过KF+MA处理后的数据处于[-40,30]范围内,符合车辆真实的横摆角速度变化。
EKF-GRU方法的目标使预测输出的轨迹尽可能的符合真实轨迹,而不是盲目的追求最终的位置误差最小。因此,采用3个轨迹预测中常用的评价指标,其中,第m点的预测轨迹和真实轨迹的位移误差 e m = ( x - x ^ ) 2 + ( y - y ^ ) 2 ( x y )为真实位置, ( x ^ y ^ )为预测位置。
最终距离误差(Final Displacement Error,FDE)测量预测时域内预测轨迹的最终位置和真实轨迹最终位置之间的距离L,而不考虑预测范围中其他时间步长中发生的预测误差。FDE计算为:
F D E = | L f |
平均距离误差(Average Displacement Error,ADE)评价预测轨迹与真实轨迹在每个时刻的平均位置误差,通常以m为单位。ADE计算为:
A D E = 1 N m = 1 N | e m |
均方根误差(Root Mean Square Error,RMSE)主要用于评价预测值与真实值之间的差异,该指标相较于ADE对大误差更加敏感。RMSE计算为
R M S E = 1 N m = 1 N e m 2
式中N为预测时间序列的长度。
EKF-CTRV基于CTRV假设预测未来轨迹的速度及横摆角速度,经过EKF生成未来轨迹。EKF-GRU则使用GRU网络处理具有时间序列信息的车辆历史轨迹,并生成加速度及横摆角速度,它们被建模为0均值高斯随机变量,标准差由GRU网络输出。GRU的输出结果并不是车辆的最终位置,因为此时完全基于数据驱动的信息缺少车辆的运动学约束,生成的预测轨迹可能不符合车辆实际行驶特性。基于此,使用EKF预测车辆的位置,生成带有运动学约束的预测轨迹。2种方法的定量预测结果见表2。典型场景下轨迹预测的定性结果如图9所示。
表2可知:EKF-GRU在3项指标上均优于EKF-CTRV,证明所提出方法的有效性。在第1s预测时间内,EKF-CTRV与EKF-GRU具有相似的指标,这是因为在短时间运动中,车辆受到运动学约束的影响更大,因此,基于CTRV假设的方法能够较为准确的预测出车辆的轨迹。在2~5s的预测时间内,基线方法EKF-CTRV产生更高的FDE值,而EKF-GRU在此时域内具有明显优势,这是因为随着时间的推移,基线方法因CTRV假设导致在未来2~5s内观测值出现严重偏差,而扩展方法则由GRU产生更符合车辆未来意图的加速度与横摆角速度,从而提升长时间轨迹预测的准确性。这表明:在长时间的轨迹预测中,行驶意图对车辆的影响要大于车辆运动学约束的影响。
表2定量结果中,EKF-GRU在RMSE方面远小于EKF-CTRV。这显示出EKF-GRU生成的轨迹不仅在预测时域最终位置更精确,而且在整个预测时域内更加符合车辆的真实轨迹。
需注意的是,EKF-CTRV和EKF-GRU在第4s及第5s时的FDE和RMSE都超过4m,由于周围环境车辆以及道路交通约束等对车辆的行驶意图产生交互影响。这表明:只考虑车辆自身历史状态信息及运动学约束的轨迹预测方法仍存在局限性。
图9a中可以看出,EKF-GRU方法能够比较准确地预测车辆轨迹。而对于图9b,EKF-GRU方法预测了其前2s向右换道的意图,但是该车辆在2s之后产生了向左换道的意图,此时EKF-GRU产生的轨迹与实际轨迹具有较大的偏差。这一现象的原因在于随着预测时间的延长,人类驾驶员的意图会出现更大的不确定性,因此,其预测也变得更加困难。
总体来看,在1~2s的时间范围内,预测轨迹与实际轨迹高度重合。这归因于基于EKF-GRU的预测方法会根据车辆状态转移方程进行预测,并且GRU在短时间内的输出方面表现出出色的预测效果。然而,在3~5s的时间段内,预测轨迹与实际轨迹出现明显偏差。这是因为长时间内的车辆状态预测需要综合考虑道路状况、周围车辆的意图等周围环境的交互信息,才能实现较为准确的轨迹预测。
1) 基于EKF-GRU的车辆轨迹预测方法充分考虑车辆的运动学约束,通过引入EKF状态转移方程进行预测,有助于提升短时轨迹预测的精度,并解决基于学习方法产生预测轨迹随意性的问题。
2) EKF-GRU方法引入循环神经网络,通过学习车辆历史轨迹信息,有效提高长时间轨迹预测的精度。基于NGSIM数据集进行的试验表明:EKF-GRU方法相较于传统基于运动学模型的轨迹预测方法,在FDE、RMSE和ADE上性能分别提升15.90%、13.26%和15.51%。
3) 尽管相较于基于运动学模型的轨迹预测方法,EKF-GRU方法在长时预测方面表现更为精准,但与结合周围车辆交互意图及高精度地图的预测方法相比,仍存在一定差距。在未来,可基于车辆动力学模型与数据驱动相结合的方法,结合环境的交互信息,作出更准确的轨迹预测;同时,采用注意力机制对周围环境分级处理,降低计算资源的消耗。
  • 国家重点研发计划课题项目(2022YFB4703702)
参考文献 引证文献
排序方式:
[1]
姚志洪, 金玉婷, 王思琛, 等. 混入智能网联汽车的交通流稳定性与安全性分析[J]. 中国安全科学学报, 2021, 31(10):136-143.
YAO Zhihong, JIN Yuting, WANG Sichen, et al. Stability and safety analysis on traffic flow mixed with ICV[J]. China Safety Science Journal, 2021, 31(10):136-143.
[2]
高振海, 鲍明喜, 高菲, 等. 智能驾驶系统对周边交通车辆预期轨迹的单模态网络预测方法[J]. 汽车技术, 2022(11):1-9.
GAO Zhenhai, BAO Mingxi, GAO Fei, et al. A uni-modal network prediction method for surrounding vehicle expected trajectory in intelligent driving system[J]. Automobile Technology, 2022(11):1-9.
[3]
RUDENKO A, PALMIERI L, HERMAN M, et al. Human motion trajectory prediction: a survey[J]. The International Journal of Robotics Research, 2020, 39(8):895-935.
[4]
谢枫, 李永乐, 苏致远, 等. 一种城市交叉路口转弯车辆轨迹预测方法[J]. 军事交通学院学报, 2019, 21(11):78-83.
XIE Feng, LI Yongle, SU Zhiyuan, et al. A method for predicting turning vehicle trajectory in urban intersection[J]. Journal of Military Transportation University, 2019, 21(11):78-83.
[5]
乔少杰, 韩楠, 朱新文, 等. 基于卡尔曼滤波的动态轨迹预测算法[J]. 电子学报, 2018, 46(2):418-423.
QIAO Shaojie, HAN Nan, ZHU Xinwen, et al. A dynamic trajectory prediction algorithm based on kalman filter[J]. Acta Electronica Sinica, 2018, 46(2):418-423.
[6]
QUN L, KRITIKA J, U-XUAN T. Gaussian process auto regression for vehicle center coordinates trajectory prediction[C]. TENCON 2009-2019 IEEE Region 10 Conference (TENCON), 2019:25-30.
[7]
CUI Henggang, VLADAN R, FANG-CHIEH C, et al. Multimodal trajectory predictions for autonomous driving using deep convolutional networks[C]. International Conference on Robotics and Automation (ICRA), 2019:2090-2096.
[8]
MARCEL S, STEFAN H, KLAUS D. Long-term occupancy grid prediction using recurrent neural networks[C]. International Conference on Robotics and Automation (ICRA), 2019:9299-9305.
[9]
XIE Guo, SHANGGUAN Anqi, FEI Rong, et al. Motion trajectory prediction based on a CNN-LSTM sequential model[J]. Science China Information Sciences, 2020, 63(11):1-21.
[10]
CUI Henggang, THI N, FANG-CHIEH C, et al. Deep kinematic models for kinematically feasible vehicle trajectory predictions[C]. IEEE International Conference on Robotics and Automation (ICRA), 2020:10 563-10 569.
[11]
吕梦男, 孙涛, 李洁. 基于扩展卡尔曼滤波的车辆状态参数估计[J]. 农业装备与车辆工程, 2018, 56(5):77-80.
LYU Mengnan, SUN Tao, LI Jie. Vehicle state parameter estimation based on extended kalman filter[J]. Agricultural Equipment & Vehicle Engineering, 2018, 56(5):77-80.
[12]
熊晓夏, 刘擎超, 沈钰杰, 等. 基于LSTM-BF的高速公路交通事故风险模型[J]. 中国安全科学学报, 2022, 32(5):170-176.
XIONG Xiaoxia, LIU Qingchao, SHEN Yujie, et al. Study on risk model of highway traffic accidents based on LSTM-BF[J]. China Safety Science Journal, 2022, 32(5):170-176.
[13]
COLYAR J. Us highway i-80 dataset[EB/0L].[2023-09-15]. https://www.fhwa.dot.gov/publications/research/operations/06137/. https://www.fhwa.dot.gov/publications/research/operations/06137/
[14]
COLYAR J. Us highway 101 dataset[EB/0L].[2023-09-15]. https://www.fhwa.dot.gov/publications/research/operations/07030/. https://www.fhwa.dot.gov/publications/research/operations/07030/
2024年第34卷第6期
PDF下载
317
121
引用本文
BibTeX
文章信息
doi: 10.16265/j.cnki.issn1003-3033.2024.06.0720
  • 接收时间:2023-12-14
  • 首发时间:2025-07-09
  • 出版时间:2024-06-28
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2023-12-14
  • 修回日期:2024-03-19
基金
国家重点研发计划课题项目(2022YFB4703702)
作者信息
    1 北京航空航天大学 交通科学与工程学院,北京 100191
    2 贵州开源爆破工程有限公司,贵州 贵阳 551400
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/zgaqkxxb/CN/10.16265/j.cnki.issn1003-3033.2024.06.0720
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