Article(id=1263819611133304925, tenantId=1146029695717560320, journalId=1263530845441638439, issueId=1263818962224165389, articleNumber=null, orderNo=null, doi=10.19693/j.issn.1673-3185.04271, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1731427200000, receivedDateStr=2024-11-13, revisedDate=1741017600000, revisedDateStr=2025-03-04, acceptedDate=null, acceptedDateStr=null, onlineDate=1779247675927, onlineDateStr=2026-05-20, pubDate=1777478400000, pubDateStr=2026-04-30, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1779247675927, onlineIssueDateStr=2026-05-20, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1779247675927, creator=13041195026, updateTime=1779247675927, updator=13041195026, issue=Issue{id=1263818962224165389, tenantId=1146029695717560320, journalId=1263530845441638439, year='2026', volume='21', issue='2', pageStart='1', pageEnd='444', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1779247521215, creator=13041195026, updateTime=1779247861438, updator=13041195026, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1263820389638070544, tenantId=1146029695717560320, journalId=1263530845441638439, issueId=1263818962224165389, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1263820389638070545, tenantId=1146029695717560320, journalId=1263530845441638439, issueId=1263818962224165389, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=404, endPage=414, ext={EN=ArticleExt(id=1263819614107066494, articleId=1263819611133304925, tenantId=1146029695717560320, journalId=1263530845441638439, language=EN, title=Generalization generation of ship overtaking scenarios for autonomous collision avoidance testing, columnId=1263819608264458931, journalTitle=Chinese Journal of Ship Research, columnName=Weapon, Electronic and Information System, runingTitle=null, highlight=null, articleAbstract=
Objective

This paper introduces a novel data-driven approach for generating realistic and hazardous overtaking scenarios. These scenarios are crucial for rigorously evaluating the autonomous collision avoidance capabilities of autonomous ships. Existing methods often struggle to balance scenario diversity, realism, and the representation of hazardous situations. To overcome this limitation, our method leverages the rich information embedded in automatic identification system (AIS) data to generate diverse and realistic overtaking encounters.

Method

Specifically, we propose a hybrid model that integrates a sequence generative adversarial network (SeqGAN) with a self-attention mechanism (SAM). The SeqGAN captures the complex patterns and dynamics in AIS-based ship trajectories, enabling the generation of novel, yet plausible, overtaking maneuvers. The incorporation of a SAM further enhances the model's ability to capture long-range dependencies in ship trajectories, resulting in more realistic and nuanced simulations. To ensure that the generated scenarios accurately reflect hazardous situations, we have developed a constraint model based on longitudinal and lateral safety distances between vessels to define realistic initial conditions. This model dynamically adjusts the initial positions and velocities of both the target vessel and the autonomous ship under test, ensuring that each generated scenario presents a genuine collision risk.

Results

The results show that the effectiveness of our approach is validated through extensive simulations. A total of 500 hazardous overtaking scenarios were generated, significantly improving the coverage of test scenarios. Notably, 97.3% of these generated trajectories fall within a predefined buffer zone that encompasses real-world trajectories, demonstrating the high fidelity of our model. Furthermore, the speed distributions of the generated target vessels closely match those observed in real-world AIS data, further validating the realism of our approach.

Conclusion

The enhanced realism and diversity of scenarios generated by this method significantly improve the efficiency of autonomous collision avoidance testing. This allows for a more precise definition of safety performance boundaries and accelerates the development and optimization of autonomous collision avoidance algorithms. Ultimately, this work contributes to the development of safer and more reliable autonomous maritime systems capable of navigating the complexities of modern maritime environments.

, correspAuthors=Xinyu ZHANG, authorNote=null, correspAuthorsNote=null, copyrightStatement=Copyright © 2026 Chinese Journal of Ship Research. All rights reserved., 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=Chenxing JIANG, Xinyu ZHANG, Kangjie ZHENG, Wenqiang GUO, Fangyuan TANG), CN=ArticleExt(id=1263819698144141441, articleId=1263819611133304925, tenantId=1146029695717560320, journalId=1263530845441638439, language=CN, title=面向自主避碰测试的船舶追越场景泛化生成方法, columnId=1263819609539527351, journalTitle=中国舰船研究, columnName=武器与电子信息系统, runingTitle=null, highlight=null, articleAbstract=
目的

为测试智能船舶自主避碰系统在追越场景下的安全性能,提出一种数据−模型驱动的船舶追越场景生成方法。

方法

基于自动识别系统(AIS)数据中的船舶追越轨迹,结合序列生成对抗网络与自注意力机制,提出高风险追越轨迹生成方法。构建以两船纵、横向安全距离为基础的追越初始状态约束模型,计算和调整追越场景下被测智能船舶与目标船的初始状态,并设计自动生成船舶追越场景的泛化算法。

结果

结果显示,基于真实数据生成的500条高风险追越轨迹能够提高测试场景覆盖度,其中97.3%的生成轨迹位于真实轨迹缓冲区内,与真实轨迹位置分布相符;目标追越船的航速概率密度与实际航速的分布一致,验证了所生成船舶追越场景的真实性与有效性。

结论

所提方法能够为智能船舶自主避碰系统的性能优化和测试评估提供符合航海实际的船舶追越场景,利用此类场景进行测试,能够有效评估自主避碰系统在追越场景下的安全性能,提高该场景下的测试效率,加速自主避碰技术的研发进程,最终提升智能船舶在追越场景下的安全性和可靠性。

, correspAuthors=张新宇, authorNote=null, correspAuthorsNote=
* 张新宇
, copyrightStatement=版权所有 © 《中国舰船研究》编辑部 2026, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=B4gDjPQmEROmzO8q2IS4JA==, magXml=V6o98TOQ4J73ZyWN5OcZUg==, pdfUrl=null, pdf=ysJIV+PlV614l3EQcwc8Ow==, pdfFileSize=8126951, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=k2JxQ0S7sTVIw63InL+yFQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=Yo283PxrVOk52JvsjVhhlA==, mapNumber=null, authorCompany=null, fund=null, authors=

蒋晨星,男,2000年生,硕士生。研究方向:智能船舶测试验证技术。E-mail:

张新宇,男,1978年生,博士,教授,博士生导师。研究方向:智能船舶自主航行关键技术,水运交通运输与系统工程学,船舶交通组织调度,海事大数据。E-mail:

, authorsList=蒋晨星, 张新宇, 郑康洁, 郭文强, 唐方圆)}, authors=[Author(id=1263819699012362394, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=2226281226@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1263819699184328864, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819699012362394, language=EN, stringName=Chenxing JIANG, firstName=Chenxing, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1Navigation College, Dalian Maritime University, Dalian 116026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1263819699654090915, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819699012362394, language=CN, stringName=蒋晨星, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1大连海事大学 航海学院,辽宁 大连 116026, bio={"content":"

蒋晨星,男,2000年生,硕士生。研究方向:智能船舶测试验证技术。E-mail:

"}, bioImg=null, bioContent=

蒋晨星,男,2000年生,硕士生。研究方向:智能船舶测试验证技术。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1263819698517434508, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=1, ext=[AuthorCompanyExt(id=1263819698555183245, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Navigation College, Dalian Maritime University, Dalian 116026, China), AuthorCompanyExt(id=1263819698588737678, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1大连海事大学 航海学院,辽宁 大连 116026)])]), Author(id=1263819700216127656, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=zhangxy@dlmu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1263819700392288428, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819700216127656, language=EN, stringName=Xinyu ZHANG, firstName=Xinyu, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, 1, address=1Navigation College, Dalian Maritime University, Dalian 116026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1263819700769775792, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819700216127656, language=CN, stringName=张新宇, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, 1, address=1大连海事大学 航海学院,辽宁 大连 116026, bio={"content":"

张新宇,男,1978年生,博士,教授,博士生导师。研究方向:智能船舶自主航行关键技术,水运交通运输与系统工程学,船舶交通组织调度,海事大数据。E-mail:

"}, bioImg=null, bioContent=

张新宇,男,1978年生,博士,教授,博士生导师。研究方向:智能船舶自主航行关键技术,水运交通运输与系统工程学,船舶交通组织调度,海事大数据。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1263819698517434508, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=1, ext=[AuthorCompanyExt(id=1263819698555183245, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Navigation College, Dalian Maritime University, Dalian 116026, China), AuthorCompanyExt(id=1263819698588737678, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1大连海事大学 航海学院,辽宁 大连 116026)])]), Author(id=1263819702527189171, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, 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=1263819702741098679, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819702527189171, language=EN, stringName=Kangjie ZHENG, firstName=Kangjie, middleName=null, lastName=ZHENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1Navigation College, Dalian Maritime University, Dalian 116026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1263819702934036667, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819702527189171, language=CN, stringName=郑康洁, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1大连海事大学 航海学院,辽宁 大连 116026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1263819698517434508, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=1, ext=[AuthorCompanyExt(id=1263819698555183245, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Navigation College, Dalian Maritime University, Dalian 116026, China), AuthorCompanyExt(id=1263819698588737678, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1大连海事大学 航海学院,辽宁 大连 116026)])]), Author(id=1263819703294746814, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, 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=1263819703835812037, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819703294746814, language=EN, stringName=Wenqiang GUO, firstName=Wenqiang, middleName=null, lastName=GUO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1Navigation College, Dalian Maritime University, Dalian 116026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1263819704070693066, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819703294746814, language=CN, stringName=郭文强, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1大连海事大学 航海学院,辽宁 大连 116026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1263819698517434508, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=1, ext=[AuthorCompanyExt(id=1263819698555183245, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Navigation College, Dalian Maritime University, Dalian 116026, China), AuthorCompanyExt(id=1263819698588737678, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1大连海事大学 航海学院,辽宁 大连 116026)])]), Author(id=1263819704351711439, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, 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=1263819704620146899, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819704351711439, language=EN, stringName=Fangyuan TANG, firstName=Fangyuan, middleName=null, lastName=TANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Dalian Maritime Safety Administration, Dalian 116001, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1263819705052160217, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, authorId=1263819704351711439, language=CN, stringName=唐方圆, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2大连海事局,辽宁 大连 116001, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1263819698806841491, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=2, ext=[AuthorCompanyExt(id=1263819698815230099, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698806841491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Dalian Maritime Safety Administration, Dalian 116001, China), AuthorCompanyExt(id=1263819698823618709, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698806841491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大连海事局,辽宁 大连 116001)])])], keywords=[Keyword(id=1263819705643557086, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, orderNo=1, keyword=intelligent ships), Keyword(id=1263819707098980576, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, orderNo=2, keyword=collision avoidance), Keyword(id=1263819707384193251, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, orderNo=3, keyword=ship overtaking), Keyword(id=1263819707673600231, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, orderNo=4, keyword=test scenario), Keyword(id=1263819708210471146, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, orderNo=5, keyword=scenario generation), Keyword(id=1263819708575375596, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, orderNo=1, keyword=智能船舶), Keyword(id=1263819708860588271, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, orderNo=2, keyword=避碰), Keyword(id=1263819709271630066, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, orderNo=3, keyword=船舶追越), Keyword(id=1263819709422625013, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, orderNo=4, keyword=测试场景), Keyword(id=1263819709703643383, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, orderNo=5, keyword=场景生成)], refs=[Reference(id=1263819718822060364, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHANG X Y, WANG C B, JIANG L L, et al. Collision-avoidance navigation systems for maritime autonomous surface ships: a state of the art survey[J]. Ocean Engineering, 2021, 235: 109380., articleTitle=null, refAbstract=null), Reference(id=1263819720558502224, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=中国船级社. 智能船舶规范[S]. 北京: 人民交通出版社, 2024., articleTitle=null, refAbstract=null), Reference(id=1263819720667554130, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=China Classification Society. Rules for intelligent ships[S]. Beijing: China Communications Press, 2024 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819720847909206, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=郑茂, 丁世淦, 兰加芬, 等. 内河航道避碰测试会遇场景建模方法研究[J]. 中国舰船研究, 2023, 18(5): 121–132., articleTitle=null, refAbstract=null), Reference(id=1263819721128927577, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHENG M, DING S G, LAN G F, et al. Study on scenario modeling method for collision avoidance test in inland waterway[J]. Chinese Journal of Ship Research, 2023, 18(5): 121–132 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819721326059868, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=张笛, 赵银祥, 崔一帆, 等. 智能船舶的研究现状可视化分析与发展趋势[J]. 交通信息与安全, 2021, 39(1): 7–16, 34., articleTitle=null, refAbstract=null), Reference(id=1263819721506414943, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHANG D, ZHAO Y X, CUI Y F, et al. A visualization analysis and development trend of intelligent ship studies[J]. Journal of Transport Information and Safety, 2021, 39(1): 7–16, 34 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819721657409891, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=GUO W Q, ZHANG X Y, LIU W, et al. Mixed traffic conditions of autonomous and human-driven ships: assessing channel traffic capacity bounds and optimizing channel management[J]. Ocean Engineering, 2024, 314: 119734., articleTitle=null, refAbstract=null), Reference(id=1263819721800016230, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=张浩杰, 苏治宝, 杨甜甜. 基于USARSim和ROS的无人平台编队仿真系统[J]. 自动化学报, 2021, 47(6): 1390–1400., articleTitle=null, refAbstract=null), Reference(id=1263819722097811818, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHANG H J, SU Z B, YANG T T. Design of team formation simulation system for unmanned ground vehicles based on USARSim and ROS[J]. Acta Automatica Sinica, 2021, 47(6): 1390–1400 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819722278166893, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=张英俊, 翟鹏宇. 海运船舶自主避碰技术研究进展与趋势[J]. 大连海事大学学报, 2022, 48(3): 1–11., articleTitle=null, refAbstract=null), Reference(id=1263819722450133361, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHANG Y J, ZHAI P Y. Research progress and trend of autonomous collision avoidance technology for marine ships[J]. Journal of Dalian Maritime University, 2022, 48(3): 1–11 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819722538213748, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=ISO. Road vehicles—safety of the intended functionality: ISO/PAS 21448: 2019[S]. Vernier: ISO, 2019., articleTitle=null, refAbstract=null), Reference(id=1263819722668237175, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=翟强, 程洪, 黄瑞, 等. 智能汽车中人工智能算法应用及其安全综述[J]. 电子科技大学学报, 2020, 49(4): 490–498, 510., articleTitle=null, refAbstract=null), Reference(id=1263819722785677689, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHAI Q, CHENG H, HUANG R, et al. A survey: artificial intelligence and its security in intelligent vehicle[J]. Journal of University of Electronic Science and Technology of China, 2020, 49(4): 490–498, 510 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819722869563772, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=陈立家, 刘岩植, 许毅, 等. AIS数据驱动的船舶智能避碰测试场景构建方法[J]. 武汉理工大学学报(交通科学与工程版), 2025, 49(1): 195–202, 209., articleTitle=null, refAbstract=null), Reference(id=1263819723033141632, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=CHEN L J, LIU Y Z, XU Y, et al. Construction method of intelligent collision avoidance test scene for ships based on AIS data-driven[J]. Journal of Wuhan University of Technology (Transportation Science & Engineering), 2025, 49(1): 195–202, 209 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819723213496708, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=STANKIEWICZ P G, MULLINS G E. Improving evaluation methodology for autonomous surface vessel COLREGs compliance[C]//OCEANS 2019-Marseille. Marseille, France: IEEE, 2019: 1−7., articleTitle=null, refAbstract=null), Reference(id=1263819723330937223, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=PORRES I, AZIMI S, LILIUS J. Scenario-based testing of a ship collision avoidance system[C]//2020 46th Euromicro Conference on Software Engineering and Advanced Applications (SEAA). Portoroz, Slovenia: IEEE, 2020: 545−552., articleTitle=null, refAbstract=null), Reference(id=1263819725235151242, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=WANG X Y, WANG Y H, MA I, et al. Research on the process-based generation method of scenarios database for maritime autonomous surface ship test[C]//ICETIS 2022; 7th International Conference on Electronic Technology and Information Science. Harbin, China: VDE, 2022: 1−8., articleTitle=null, refAbstract=null), Reference(id=1263819725491003790, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=陈立家, 王凯, 黄立文, 等. 船舶自主航行虚拟测试场景研究进展[J]. 中国舰船研究, 2025, 20(1): 25–37., articleTitle=null, refAbstract=null), Reference(id=1263819725755244946, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=CHEN L J, WANG K, HUANG L W, et al. Research progress on test scenario of ship autonomous navigation[J]. Chinese Journal of Ship Research, 2025, 20(1): 25–37 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819726061429141, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=SNIJDERS R, ELROFAI H. Scenario identification for safety assessment of autonomous shipping using AIS data[C]//15th International Naval Engineering Conference & Exhibition. Delft, 2020., articleTitle=null, refAbstract=null), Reference(id=1263819726199841174, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=BOLBOT V, GKEREKOS C, THEOTOKATOS G, et al. Automatic traffic scenarios generation for autonomous ships collision avoidance system testing[J]. Ocean engineering, 2022, 254: 111309., articleTitle=null, refAbstract=null), Reference(id=1263819726283727258, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=丁志国, 张新宇, 王程博, 等. 基于驾驶实践的无人船智能避碰决策方法[J]. 中国舰船研究, 2021, 16(1): 96–104, 113., articleTitle=null, refAbstract=null), Reference(id=1263819726518608284, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=DING Z G, ZHANG X Y, WANG C B, et al. Intelligent collision avoidance decision-making method for unmanned ships based on driving practice[J]. Chinese Journal of Ship Research, 2021, 16(1): 96–104, 113 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819726686380447, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=CHAUVIN C, LARDJANE S. Decision making and strategies in an interaction situation: collision avoidance at sea[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2008, 11(4): 259–269., articleTitle=null, refAbstract=null), Reference(id=1263819727076450722, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=赵祥模, 赵玉钰, 景首才, 等. 面向自动驾驶测试的危险变道场景泛化生成[J]. 自动化学报, 2023, 49(10): 2211–2223., articleTitle=null, refAbstract=null), Reference(id=1263819727172919717, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHAO X M, ZHAO Y Y, JING S C, et al. Generalization generation of hazardous lane-changing scenarios for automated vehicle testing[J]. Acta Automatica Sinica, 2023, 49(10): 2211–2223 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819727391023530, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems. Montreal: MIT Press, 2014: 2672−2680., articleTitle=null, refAbstract=null), Reference(id=1263819727466521005, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=JING S C, ZHAO Y Y, ZHAO X M, et al. An efficient high-risk lane-changing scenario edge cases generation method for autonomous vehicle safety testing[J/OL]. IEEE Transactions on Intelligent Vehicles, 2024. https://dx.doi.org/10.1109/TIV.2024.3414828., articleTitle=null, refAbstract=null), Reference(id=1263819727571378606, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHANG H, GOODFELLOW I, METAXAS D, et al. Self-attention generative adversarial networks[C]//Proceedings of the 36th International Conference on Machine Learning. Long Beach, CA, USA: PMLR, 2019: 7354−7363., articleTitle=null, refAbstract=null), Reference(id=1263819727663653298, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=刘钊, 谢金鹏, 张明阳, 等. 考虑避让紧迫度的数字化船舶领域模型[J]. 中国航海, 2021, 44(2): 15–20., articleTitle=null, refAbstract=null), Reference(id=1263819727772705204, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=LIU Z, XIE J P, ZHANG M Y, et al. A numerical model of ship domain reflecting collision urgency ditribution[J]. Navigation of China, 2021, 44(2): 15–20 (in Chinese)., articleTitle=null, refAbstract=null), Reference(id=1263819727860785591, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=KIJIMA K, FURUKAWA Y. Automatic collision avoidance system using the concept of blocking area[J]. IFAC Proceedings Volumes, 2003, 36(21): 223–228., articleTitle=null, refAbstract=null), Reference(id=1263819729706279355, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=李元奎, 彭飞, 贾梓淇, 等. 基于行船视距的船舶会遇安全距离研究[J]. 安全与环境学报, 2024, 24(4): 1442–1449., articleTitle=null, refAbstract=null), Reference(id=1263819729848885695, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=LI Y K, PENG F, JIA Z Q, et al. Study on safe distances of ships encounter situations based on sight distance[J]. Journal of Safety and Environment, 2024, 24(4): 1442–1449 (in Chinese)., articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1263819698517434508, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=1, ext=[AuthorCompanyExt(id=1263819698555183245, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Navigation College, Dalian Maritime University, Dalian 116026, China), AuthorCompanyExt(id=1263819698588737678, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698517434508, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1大连海事大学 航海学院,辽宁 大连 116026)]), AuthorCompany(id=1263819698806841491, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, xref=2, ext=[AuthorCompanyExt(id=1263819698815230099, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698806841491, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Dalian Maritime Safety Administration, Dalian 116001, China), AuthorCompanyExt(id=1263819698823618709, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, companyId=1263819698806841491, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大连海事局,辽宁 大连 116001)])], figs=[ArticleFig(id=1263819711763046652, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.1, caption=Stage division of ship overtaking process, figureFileSmall=2JFr+rI4/MS6EbQzsMpqzw==, figureFileBig=k2JxQ0S7sTVIw63InL+yFQ==, tableContent=null), ArticleFig(id=1263819711914041599, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图1, caption=船舶追越过程划分, figureFileSmall=2JFr+rI4/MS6EbQzsMpqzw==, figureFileBig=k2JxQ0S7sTVIw63InL+yFQ==, tableContent=null), ArticleFig(id=1263819712593518852, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.2, caption=Coordinate system of ship overtaking scenario, figureFileSmall=aoUd71B2LLe3sHfy117I8g==, figureFileBig=ZsHf2QX1cG9AZbsTUF+2JQ==, tableContent=null), ArticleFig(id=1263819713021337864, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图2, caption=追越场景坐标系, figureFileSmall=aoUd71B2LLe3sHfy117I8g==, figureFileBig=ZsHf2QX1cG9AZbsTUF+2JQ==, tableContent=null), ArticleFig(id=1263819713264607499, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.3, caption=Generation process for hazardous ship overtaking test scenarios, figureFileSmall=yTpnt41kIWyCPNMCO7Gq4A==, figureFileBig=FVZWrBd/qPJ4j1RKtfg6Ew==, tableContent=null), ArticleFig(id=1263819713457545486, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图3, caption=高风险追越测试场景生成流程, figureFileSmall=yTpnt41kIWyCPNMCO7Gq4A==, figureFileBig=FVZWrBd/qPJ4j1RKtfg6Ew==, tableContent=null), ArticleFig(id=1263819713671454993, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.4, caption=Architecture diagram of SAM-SeqGAN model, figureFileSmall=Fm6e4LREDuLBrmsl5IYZQw==, figureFileBig=Z+1t3OkNbaN/VFBRt+NdtA==, tableContent=null), ArticleFig(id=1263819713902141717, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图4, caption=SAM-SeqGAN模型架构, figureFileSmall=Fm6e4LREDuLBrmsl5IYZQw==, figureFileBig=Z+1t3OkNbaN/VFBRt+NdtA==, tableContent=null), ArticleFig(id=1263819714095079705, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.5, caption=The network structure of the generator in the SAM-SeqGAN model, figureFileSmall=aQo1q44agPWH1mNGb4L8xA==, figureFileBig=sZ1YPH+EP7Mc2zchRvO9bw==, tableContent=null), ArticleFig(id=1263819714342543642, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图5, caption=SAM-SeqGAN模型生成器的网络结构, figureFileSmall=aQo1q44agPWH1mNGb4L8xA==, figureFileBig=sZ1YPH+EP7Mc2zchRvO9bw==, tableContent=null), ArticleFig(id=1263819716028653853, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.6, caption=The network structure of the discriminator in the SAM-SeqGAN model, figureFileSmall=G/TGR75HN3XYOkJ/OZAdzQ==, figureFileBig=pVkPsmo912kpt0HzHQONOg==, tableContent=null), ArticleFig(id=1263819716234174753, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图6, caption=SAM-SeqGAN模型判别器的网络结构, figureFileSmall=G/TGR75HN3XYOkJ/OZAdzQ==, figureFileBig=pVkPsmo912kpt0HzHQONOg==, tableContent=null), ArticleFig(id=1263819716632633637, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.7, caption=Trajectory buffer area of the target overtake vessel (OS), figureFileSmall=HT1WjXREyKJ7YsIQEoe44Q==, figureFileBig=g6U1LTzsflcDFtwaX9fDwQ==, tableContent=null), ArticleFig(id=1263819716880097576, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图7, caption=目标追越船真实轨迹缓冲区, figureFileSmall=HT1WjXREyKJ7YsIQEoe44Q==, figureFileBig=g6U1LTzsflcDFtwaX9fDwQ==, tableContent=null), ArticleFig(id=1263819717089812780, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.8, caption=Speed distribution of the target overtaking vessel (OS) at the start of the entry phase, figureFileSmall=eZTAQHscHqyWIbpnXhgSEg==, figureFileBig=KmmyCMC2nWukosoLTY3MUA==, tableContent=null), ArticleFig(id=1263819717215641904, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图8, caption=目标追越船切入阶段起始时刻速度分布, figureFileSmall=eZTAQHscHqyWIbpnXhgSEg==, figureFileBig=KmmyCMC2nWukosoLTY3MUA==, tableContent=null), ArticleFig(id=1263819717467300149, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.9, caption=Training loss values of three generative adversarial networks, figureFileSmall=Uv9fV4IuwtYibUcf1lwEFw==, figureFileBig=8U74n7Crn3gAXolb5+KFsA==, tableContent=null), ArticleFig(id=1263819717618295097, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图9, caption=3种生成对抗网络的损失值, figureFileSmall=Uv9fV4IuwtYibUcf1lwEFw==, figureFileBig=8U74n7Crn3gAXolb5+KFsA==, tableContent=null), ArticleFig(id=1263819717807038780, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.10, caption=Hazardous ship overtaking scenarios, figureFileSmall=xek6bn169Az93w6DNdT1YA==, figureFileBig=HFLR4jIz3C8IlqEPrl/73w==, tableContent=null), ArticleFig(id=1263819717970616637, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图10, caption=高风险追越测试场景, figureFileSmall=xek6bn169Az93w6DNdT1YA==, figureFileBig=HFLR4jIz3C8IlqEPrl/73w==, tableContent=null), ArticleFig(id=1263819718142583104, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Fig.11, caption=Virtual ship overtaking test scenario, figureFileSmall=0RRbXvez2tmfGSLJkumPqA==, figureFileBig=5hI6mMNzs2Uvq+inLZmplg==, tableContent=null), ArticleFig(id=1263819718440378693, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=图11, caption=虚拟追越测试场景, figureFileSmall=0RRbXvez2tmfGSLJkumPqA==, figureFileBig=5hI6mMNzs2Uvq+inLZmplg==, tableContent=null), ArticleFig(id=1263819718608150854, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=EN, label=Tab.1, caption=

Parameter settings in the experiment

, figureFileSmall=null, figureFileBig=null, tableContent=
参数数值
船长L/m100
TS的船速/kn7
OS的船速/kn10
真实追越轨迹数量104
真实追越轨迹序列长度N200
生成器嵌入维数64
生成器初始学习率0.002
判别器初始学习率0.000 5
生成器预训练次数200
判别器预训练次数100
生成器训练步数gstep1
判别器训练步数dstep1
生成与真实轨迹数量比ratio3
), ArticleFig(id=1263819718704619849, tenantId=1146029695717560320, journalId=1263530845441638439, articleId=1263819611133304925, language=CN, label=表1, caption=

实验参数设置

, figureFileSmall=null, figureFileBig=null, tableContent=
参数数值
船长L/m100
TS的船速/kn7
OS的船速/kn10
真实追越轨迹数量104
真实追越轨迹序列长度N200
生成器嵌入维数64
生成器初始学习率0.002
判别器初始学习率0.000 5
生成器预训练次数200
判别器预训练次数100
生成器训练步数gstep1
判别器训练步数dstep1
生成与真实轨迹数量比ratio3
)], attaches=null, journal=Journal(id=1263530641632018469, delFlag=0, nameCn=中国舰船研究, nameEn=Chinese Journal of Ship Research, nameHistory1=null, nameHistory2=null, issn=1673-3185, eissn=null, cn=42-1755/TJ, 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=uuiC2KHI0RbgmeHEYieSVQ==, journalPrice=null, startedYear=null, abbrevIsoEn=Chinese Journal of Ship Research, journalRemark=null, publicationField=null, createdTime=1779178780231, updatedTime=1779179141739, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=C, firstLetterEn=C, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=uuiC2KHI0RbgmeHEYieSVQ==, picEn=742/gRTuoSZweF2ujSJArQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1263532158132564178, language=CN, name=中国舰船研究, 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=1779179141791, updatedTime=1779179141791, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://zgjcyjauthor.manuscriptcloud.com/, submissionEditorUrl=https://zgjcyjeditor.manuscriptcloud.com/, submissionReviewUrl=https://zgjcyjauthor.manuscriptcloud.com/, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1263532158254198995, language=EN, name=Chinese Journal of Ship Research, 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=1779179141820, updatedTime=1779179141820, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://zgjcyjauthor.manuscriptcloud.com/, submissionEditorUrl=https://zgjcyjeditor.manuscriptcloud.com/, submissionReviewUrl=https://zgjcyjauthor.manuscriptcloud.com/, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1263530845441638439, websiteList=[Website(id=1263532309169451247, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1263530845441638439, 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/zgjcyj/CN, language=CN, createTime=1779179177801, createBy=18614031015, updateTime=1779180752761, updateBy=18614031015, name=中国舰船研究-中文, tplId=1146099689490845704, title=中国舰船研究, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1263552162215375681, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=articleTextType, value=kx, createTime=1779183911136, updateTime=1779183911136, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162181821246, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=banner, value=null, createTime=1779183911128, updateTime=1779183911128, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162253124420, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=grayFlag, value=0, createTime=1779183911145, updateTime=1779183911145, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162173432637, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=logo, value=https://castjournals.cast.org.cn/joweb/zgjcyj/CN/file/pic?fileId=nJodoAVDNU0dVNGTgYrzsA==, createTime=1779183911126, updateTime=1779183911126, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162299261766, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=minRunFlag, value=0, createTime=1779183911156, updateTime=1779183911156, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162198598464, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zgjcyj/CN/file/pic, createTime=1779183911132, updateTime=1779183911132, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162286678853, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=silenceFlag, value=0, createTime=1779183911153, updateTime=1779183911153, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162190209855, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1779183911130, updateTime=1779183911130, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162232152898, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=themeColor, value=null, createTime=1779183911140, updateTime=1779183911140, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552162240541507, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309169451247, code=themeStyle, value=null, createTime=1779183911142, updateTime=1779183911142, creator=18614031015, updator=18614031015)]), Website(id=1263532309249143025, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1263530845441638439, 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/zgjcyj/EN, language=EN, createTime=1779179177820, createBy=18614031015, updateTime=1779180748021, updateBy=18614031015, name=中国舰船研究-英文, tplId=1146101810881728533, title=Chinese Journal of Ship Research, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1263552187725132620, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=articleTextType, value=kx, createTime=1779183917218, updateTime=1779183917218, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187704161097, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=banner, value=null, createTime=1779183917213, updateTime=1779183917213, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187754492751, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=grayFlag, value=0, createTime=1779183917225, updateTime=1779183917225, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187695772488, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=logo, value=https://castjournals.cast.org.cn/joweb/zgjcyj/EN/file/pic?fileId=nJodoAVDNU0dVNGTgYrzsA==, createTime=1779183917211, updateTime=1779183917211, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187779658577, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=minRunFlag, value=0, createTime=1779183917231, updateTime=1779183917231, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187720938315, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zgjcyj/EN/file/pic, createTime=1779183917217, updateTime=1779183917217, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187762881360, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=silenceFlag, value=0, createTime=1779183917228, updateTime=1779183917228, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187712549706, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1779183917215, updateTime=1779183917215, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187733521229, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=themeColor, value=null, createTime=1779183917220, updateTime=1779183917220, creator=18614031015, updator=18614031015), WebsiteProps(id=1263552187737715534, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1263532309249143025, code=themeStyle, value=null, createTime=1779183917222, updateTime=1779183917222, creator=18614031015, updator=18614031015)])], journalTitle=中国舰船研究, weixinUrl=null, journalUrl=https://www.ship-research.com/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Chinese Journal of Ship Research, journalPhotoCn=uuiC2KHI0RbgmeHEYieSVQ==, journalPhotoEn=742/gRTuoSZweF2ujSJArQ==, journalFirstLetter=C, 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/zgjcyj/CN/10.19693/j.issn.1673-3185.04271, detailUrlEn=https://castjournals.cast.org.cn/joweb/zgjcyj/EN/10.19693/j.issn.1673-3185.04271, pdfUrlCn=https://castjournals.cast.org.cn/joweb/zgjcyj/CN/PDF/10.19693/j.issn.1673-3185.04271, pdfUrlEn=https://castjournals.cast.org.cn/joweb/zgjcyj/EN/PDF/10.19693/j.issn.1673-3185.04271, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
面向自主避碰测试的船舶追越场景泛化生成方法
收藏切换
PDF下载
蒋晨星 1 , 张新宇 *, 1 , 郑康洁 1 , 郭文强 1 , 唐方圆 2
中国舰船研究 | 武器与电子信息系统 2026,21(2): 404-414
收起
收藏切换
中国舰船研究 | 武器与电子信息系统 2026, 21(2): 404-414
面向自主避碰测试的船舶追越场景泛化生成方法
全屏
蒋晨星1 , 张新宇*, 1 , 郑康洁1, 郭文强1, 唐方圆2
作者信息
  • 1大连海事大学 航海学院,辽宁 大连 116026
  • 2大连海事局,辽宁 大连 116001
  • 蒋晨星,男,2000年生,硕士生。研究方向:智能船舶测试验证技术。E-mail:

    张新宇,男,1978年生,博士,教授,博士生导师。研究方向:智能船舶自主航行关键技术,水运交通运输与系统工程学,船舶交通组织调度,海事大数据。E-mail:

通讯作者:

* 张新宇
Generalization generation of ship overtaking scenarios for autonomous collision avoidance testing
Chenxing JIANG1 , Xinyu ZHANG*, 1 , Kangjie ZHENG1, Wenqiang GUO1, Fangyuan TANG2
Affiliations
  • 1Navigation College, Dalian Maritime University, Dalian 116026, China
  • 2Dalian Maritime Safety Administration, Dalian 116001, China
出版时间: 2026-04-30 doi: 10.19693/j.issn.1673-3185.04271
文章导航
收藏切换
目的

为测试智能船舶自主避碰系统在追越场景下的安全性能,提出一种数据−模型驱动的船舶追越场景生成方法。

方法

基于自动识别系统(AIS)数据中的船舶追越轨迹,结合序列生成对抗网络与自注意力机制,提出高风险追越轨迹生成方法。构建以两船纵、横向安全距离为基础的追越初始状态约束模型,计算和调整追越场景下被测智能船舶与目标船的初始状态,并设计自动生成船舶追越场景的泛化算法。

结果

结果显示,基于真实数据生成的500条高风险追越轨迹能够提高测试场景覆盖度,其中97.3%的生成轨迹位于真实轨迹缓冲区内,与真实轨迹位置分布相符;目标追越船的航速概率密度与实际航速的分布一致,验证了所生成船舶追越场景的真实性与有效性。

结论

所提方法能够为智能船舶自主避碰系统的性能优化和测试评估提供符合航海实际的船舶追越场景,利用此类场景进行测试,能够有效评估自主避碰系统在追越场景下的安全性能,提高该场景下的测试效率,加速自主避碰技术的研发进程,最终提升智能船舶在追越场景下的安全性和可靠性。

智能船舶  /  避碰  /  船舶追越  /  测试场景  /  场景生成
Objective

This paper introduces a novel data-driven approach for generating realistic and hazardous overtaking scenarios. These scenarios are crucial for rigorously evaluating the autonomous collision avoidance capabilities of autonomous ships. Existing methods often struggle to balance scenario diversity, realism, and the representation of hazardous situations. To overcome this limitation, our method leverages the rich information embedded in automatic identification system (AIS) data to generate diverse and realistic overtaking encounters.

Method

Specifically, we propose a hybrid model that integrates a sequence generative adversarial network (SeqGAN) with a self-attention mechanism (SAM). The SeqGAN captures the complex patterns and dynamics in AIS-based ship trajectories, enabling the generation of novel, yet plausible, overtaking maneuvers. The incorporation of a SAM further enhances the model's ability to capture long-range dependencies in ship trajectories, resulting in more realistic and nuanced simulations. To ensure that the generated scenarios accurately reflect hazardous situations, we have developed a constraint model based on longitudinal and lateral safety distances between vessels to define realistic initial conditions. This model dynamically adjusts the initial positions and velocities of both the target vessel and the autonomous ship under test, ensuring that each generated scenario presents a genuine collision risk.

Results

The results show that the effectiveness of our approach is validated through extensive simulations. A total of 500 hazardous overtaking scenarios were generated, significantly improving the coverage of test scenarios. Notably, 97.3% of these generated trajectories fall within a predefined buffer zone that encompasses real-world trajectories, demonstrating the high fidelity of our model. Furthermore, the speed distributions of the generated target vessels closely match those observed in real-world AIS data, further validating the realism of our approach.

Conclusion

The enhanced realism and diversity of scenarios generated by this method significantly improve the efficiency of autonomous collision avoidance testing. This allows for a more precise definition of safety performance boundaries and accelerates the development and optimization of autonomous collision avoidance algorithms. Ultimately, this work contributes to the development of safer and more reliable autonomous maritime systems capable of navigating the complexities of modern maritime environments.

intelligent ships  /  collision avoidance  /  ship overtaking  /  test scenario  /  scenario generation
蒋晨星, 张新宇, 郑康洁, 郭文强, 唐方圆. 面向自主避碰测试的船舶追越场景泛化生成方法. 中国舰船研究, 2026 , 21 (2) : 404 -414 . DOI: 10.19693/j.issn.1673-3185.04271
Chenxing JIANG, Xinyu ZHANG, Kangjie ZHENG, Wenqiang GUO, Fangyuan TANG. Generalization generation of ship overtaking scenarios for autonomous collision avoidance testing[J]. Chinese Journal of Ship Research, 2026 , 21 (2) : 404 -414 . DOI: 10.19693/j.issn.1673-3185.04271
智能船舶凭借其在提升航行安全性、降低人力依赖与优化通航效能方面的显著优势[1],正逐步从理论构想走向实际应用。在日益复杂的航行环境下,智能船舶需要遵守《1972年国际海上避碰规则》(以下简称《规则》),并实施高效的避碰决策与操纵,以避免碰撞事故的发生[2]。智能船舶需应对追越、对遇以及交叉相遇三类会遇态势。其中,船舶追越行为多发于进出港航道、海峡等水域,是导致船舶碰撞的高频因素[3],尤其是在低能见度或航道宽度受限的情况下,由于船吸效应、操纵空间不足以及决策失误等因素,更易引发碰撞事故。因此,针对智能船舶在追越场景中的自主避碰能力开展系统性的测试,是确保智能船舶从“实验室可靠”向“全场景可靠”升级的核心环节[4]
自主避碰系统通过实时感知周围环境、及时识别潜在碰撞危险并自动采取避碰操纵,不仅能够减轻船员的工作负担,降低人为失误的可能性,还能提供数据支持智能船舶进行安全分析和风险评估,从而改进航运安全管理并降低整体事故率。鉴于智能船舶的测试范畴已从传统的“人−船”二元独立体系深化为“人−船−环境−任务”高度耦合的复杂系统[5],传统的实船和模型测试方法受测试效率、成本及安全风险的限制,已难以满足自主避碰测试的多元化需求。与之相比,虚拟测试凭借其零风险、高效率、低成本的优势,已成为一种重要的智能船舶自主航行功能测试验证手段[6]
场景作为虚拟测试的基础,是对航行过程中人、船、环境等要素的综合反映与抽象表达。由于当前船舶测试领域缺乏统一的指标体系[7],借鉴道路交通相关标准[8],船舶自主避碰测试场景可以从功能安全的角度划分为安全场景、危险场景和边缘场景三类。其中,危险场景在实际航行过程中较为罕见[9],如何有效提高场景的危险程度,以全面评估智能船舶的自主避碰能力,成为亟待解决的问题。
现有测试场景生成方法可以分为模型驱动和数据驱动两类[10]。在模型驱动的测试场景生成方面,研究人员依靠专业知识理解相关规则和测试需求,通过构建模型来推导出测试场景。Stankiewicz等[11]基于《规则》,人为创建了一系列测试场景组合,并对每个场景的风险程度进行了评分。Porres等[12]将测试功能与具体场景相联系,使用神经网络模型生成了大量自主避碰测试场景。Wang等[13]针对内河水域高风险通航场景,进行了场景解构与分析,建立了自主航行测试场景数据库。采用模型驱动方法能够快速生成大量测试场景,可有效提升场景覆盖率。然而,模型假设的局限性和置信度不足有可能导致生成大量低风险的无意义场景或偏离航海实际的极端场景,从而影响测试效率与准确性。
数据驱动的场景生成方法侧重于对自动识别系统(automatic identification system,AIS)、雷达等历史航行数据进行特征提取和规律总结,能够增加测试场景的真实性与合理性[14]。Snijders等[15]通过分析历史AIS数据,成功识别出避碰规则中描述的各类会遇局面并进行分类,可为后续测试场景的构建提供数据基础。Bolbot等[16]定义了自主避碰测试场景的关键参数,使用Sobol采样确定了参数的具体数值,随后又通过参数重组生成测试场景,可在保持测试覆盖范围的同时减少所需场景数量。尽管真实的数据能够反映船舶航行规律,但其主要来源于安全航行过程,高风险场景的天然稀缺性会导致场景覆盖度不足。此外,实际数据的参数范围受限于历史航行条件,难以灵活调整从而生成可控风险等级的测试场景。当前,数据驱动方法仅关注输入数据与输出场景的对应关系,缺乏对内部逻辑的解释,难以深入理解自主避碰系统的功能和行为。
现有的方法大多假设目标船不具备操纵能力,仅按照《规则》将保向保速行驶作为唯一的场景演变基准[17]。然而,船舶驾驶员的避碰操纵行为存在高度的差异与不确定性,有时甚至会采取违背《规则》要求的行动以减少偏航损失[18]。因此,若要构建符合实际航海情境的追越测试场景,需解决以下2个关键问题:
1) 生成的目标船追越轨迹数量需满足测试覆盖度要求,轨迹的真实性符合船舶在驾驶员操纵下的航行特征[19]
2) 在追越测试场景构建过程中,应根据船舶操纵性限制构建边界场景模型,剔除背离航海实际的极端场景,以确保场景的有效性与可实施性。
为解决上述问题,本文拟提出一种数据−模型驱动的船舶追越场景泛化生成方法,即基于船舶AIS数据中的实际追越数据,提出序列生成对抗网络(sequence generative adversarial nets with policy gradient,SeqGAN)与自注意力机制(self-attention mecha-nism,SAM)相融合的高风险追越轨迹生成方法,同时,构建基于两船纵、横向安全距离的追越初始状态约束模型,计算并调整追越场景中被测智能船与目标船的初始状态,设计船舶追越场景泛化生成算法,自动生成能够用于自主避碰系统安全性能测试的追越测试场景。
根据《规则》的定义,追越是指一船从另一船正横后大于22.5°的某一方向接近并超越该船的行为。结合航海实践,本文将追越场景定义为:两船同向航行,距离小于3 n mile且后船船速大于前船船速的会遇过程。在这一过程中,后船为追越船(overtake ship,OS),前船为被追越船(test ship,TS)。本文在测试过程中将OS设置为目标追越船,TS设置为被测智能船,以此评估TS在追越场景下的自主避碰能力。
图1所示,根据船舶的操纵行为及两船的相对运动状态,追越过程可以划分为以下4个阶段:
1) 起始阶段。OS以大于TS的航行速度与TS同向行驶,当两船距离达到追越起始距离时,OS做出开始追越TS的决策。
2) 转向阶段。OS根据与TS的相对方位及相对距离,选择向左或右转向,此时,两船产生横向距离。
3) 并行阶段。OS保持一定的横向安全距离 与TS并排行驶。
4) 切入阶段。OS的船尾超过TS的船首,在留有纵向安全距离的情形下切入TS艏向,并在驶过让清后恢复原航向继续行驶。
本文通过船位、船速、航向等时空特征,从大连港某水域2019年的船舶AIS数据中识别出了104条真实追越轨迹。根据OS的航向航速变化以及两船的相对位置,将真实追越轨迹按阶段进行了划分。为在统一尺度下分析和比较不同追越轨迹的危险程度,并提升模型训练效果,本文对数据进行了归一化处理,建立了以OS的初始位置为原点的追越场景坐标系,如图2所示。
在该坐标系中,真实的追越轨迹可以表示为集合:${R_{{\text{OT}}}} = \left\{ {{R_1}, \cdots ,{R_i}, \cdots ,{R_{104}}} \right\}$,其中第i条追越轨迹为${R_{i}} = \left\{ {{P_1^i}, \cdots ,{P_t^i}, \cdots ,{P_{n}^i}} \right\}$$P_t^i$为OS在t时刻的轨迹点,满足如下关系:
$ \left\{ \begin{gathered} {R_i} = \sum\limits_{t = 1}^T {P_t^i} \\ P_t^i = \left( {{x_t},{y_t},{v_t},{c_t}} \right) \\ \end{gathered} \right. $
式中:T为追越过程总时长;${x_t}$${y_t}$为OS在t时刻的位置坐标;${v_t}$${{{c}}_t}$分别为OS在t时刻的对地航速及对地航向。
本文所述船舶追越场景以起始阶段OS的船位作为场景的坐标原点,以OS的位置($x_t^{\rm{o}}$,$y_t^{\rm{o}}$)、OS的航行速度$v_t^{\rm{o}}$,以及横向安全距离${S_{\rm{c}}}$和纵向安全距离${S_{\rm{l}}}$作为场景特征参数。在这些参数中,横向位置和纵向速度变化最能表征OS的追越轨迹特征,而纵向速度是位置对时间t的导数。因此,在生成轨迹时,主要考虑OS在X轴方向的速度和Y轴方向的位置。
构建船舶追越场景需要模拟实际追越过程,通过建模提升轨迹数量及风险程度,从而生成新的高风险追越场景。该场景主要包括OS的航行轨迹生成和TS的初始状态设定这2个要素。尽管船舶的AIS数据反映了船舶的实际航行过程,但其数量有限且危险程度不足,难以全面测试智能船舶在追越场景下的自主避碰能力。
图3所示,本文基于真实追越轨迹进行高风险追越轨迹生成,并构建基于两船纵、横向安全距离的追越初始状态约束模型,通过调整高风险追越场景中TS与OS的初始状态,设计出危险场景泛化生成算法,从而大规模生成高风险追越测试场景。
为了确保生成轨迹的真实性,需要学习真实追越轨迹的数据特征分布。真实追越轨迹由离散轨迹点构成,由于离散取值操作不可微,传统的生成对抗网络无法直接生成此类数据[20]。为此,本文引入SeqGAN,通过融合强化学习的策略梯度机制,有效解决离散序列衍生问题。该方法已在道路交通领域的车辆轨迹生成中展现出良好的效果[21]
然而,船舶追越过程具有持续时间长、航速航向渐变性强等特点,传统的SeqGAN在处理此类长时序数据时易出现长程依赖关系捕捉能力不足的问题。SAM[22]通过学习轨迹中船位、航速等特征的动态关联关系,可有效捕捉序列中的长程依赖特征。因此,本文在SeqGAN架构中融入了SAM,以提升生成轨迹的准确性和对高风险特征的表达能力。
图4所示,SAM-SeqGAN由生成器G和判别器D通过动态的对抗博弈过程不断优化,以生成符合真实追越行为的船舶轨迹。生成器G以真实追越轨迹${R_{\rm{OT}}}$作为输入,通过调整参数θ生成新的轨迹;判别器D负责鉴别输入数据为真实追越轨迹${R_{\rm{OT}}}$或生成轨迹${L_{\rm{OT}}}$,参数为$\phi $
SAM-SeqGAN通过长短时记忆(long short-term memory,LSTM)网络与SAM的协同作用学习真实追越轨迹的时空动态特征,用以确保生成的轨迹不仅满足静态分布相似性,还能复现真实追越过程中船舶的操纵时序逻辑。此外,还将通过蒙特卡洛搜索对生成轨迹进行物理可行性校验,剔除违反船舶动力学规律的异常轨迹。判别器D则利用卷积神经网络(convolutional neural network,CNN)的特征提取能力区分输入数据的真伪。二者有各自的损失函数$J_{}^{(G)}(\theta ,\phi )$$J_{}^{(D)}(\theta ,\phi )$,在训练过程中寻找一对最优参数组合$(\theta ,\phi )$,以使两者的损失函数均达到局部最小值。
具体而言,生成器G生成的追越轨迹可以表示为${L_{{\text{OT}}}} = \left\{ {{L_1},...,{L_m},...,{L_M}} \right\}$,其中M为生成船舶追越轨迹的总数量,生成的第m条轨迹为${L_m} = \left\{ {L_1^m,...,L_n^m,...,L_N^m} \right\}$。此时,生成器以当前已生成的轨迹点${Z_{1:n - 1}}$为当前状态,预测并输出下一个轨迹点$L_n^m$,以此逐步构建完整的追越轨迹。为提升轨迹的完整性和真实性,生成器生成每条轨迹后都会使用蒙特卡洛搜索(MC search)对轨迹进行补全,然后再输入判别器。
判别器D同时接收真实追越轨迹和生成器的生成轨迹作为输入,通过评估输出策略梯度,指导生成器G进行策略优化。针对第m条生成轨迹${L_m}$的具体训练方式如式(2)所示,通过计算生成轨迹与真实轨迹在判别器视角下的价值差异,实现生成器的精准优化。
$ \begin{split} &\mathop {\min }\limits_G \mathop {\max }\limits_D V\left( {G,D} \right) = - \frac{1}{2}{E_{{R_i}:D}}[\log D({R_i})] - \\&\qquad {E_{{L_{{\text{OT}}}}:G}}\{ \log [1 - D[G({L_{{\text{OT}}}})]\} \end{split} $
式中:V为生成器和判别器的收益;$D({R_i})$为判别器判断第i条真实追越轨迹是否真实的概率;$D[G({L_m})]$为判别器判断生成器生成的第m条轨迹${L_m}$是否真实的概率;$ {E_{{R_i}:D}}[\log D({R_i})] $为判别器D的参数下对输入数据${R_i}$的期望;$ {E_{{L_{{\text{OT}}}}:G}} $为生成器G输入$ {L_{{\text{OT}}}} $的期望,其分布依赖于生成器G的参数。
在训练过程中,生成器采用策略梯度中的对数损失函数(式(3))作为优化目标,通过最大化判别器将生成数据误判为真实数据的概率,不断提升生成轨迹的质量与多样性。其中,$L_{n - 1}^m$表示生成的第m条追越轨迹的第n−1个值。
$ \begin{split} &\qquad\qquad\qquad\qquad {\rm{log}}\;Loss = \\&\sum\limits_{n - 1}^N {E_{L_{n - 1}^m~G}}\left[\sum\nolimits_{L_n^m \in {L_m}} {G(L_1^m|L_{n - 1}^m) \cdot } Q_D^G(L_{n - 1}^m,L_1^m)\right] \end{split} $
式中:$ G(L_1^m|L_{n - 1}^m) $为生成器在给定历史状态$ L_{n - 1}^m $下生成新状态$ L_1^m $的条件概率;$ Q_D^G $为该生成步骤的评分,评分越高,表示生成样本越逼真。
判别器D采用交叉熵损失函数作为损失函数,其目的是最大化判别器对真实数据和生成数据的分类准确度。损失函数${f_D}$可以描述为
$ {f_D} = \mathop {\min }\limits_D ( - {E_{{R_i}}}[\ln D(L)] - {E_{{L_{\rm{OT}}}}}[\ln (1 - D(L))]) $
传统的SeqGAN受限于LSTM网络的记忆能力,难以有效捕捉船舶追越轨迹中特征间的全局依赖关系。为克服此缺陷,本文在LSTM网络中引入了SAM。通过分析真实追越轨迹中的航速变化率、转向速率以及坐标系中两船相对位置分布等关键特征,构建船舶追越行为的多维特征矩阵,然后利用SAM捕捉特征间的动态关联,以确保生成更贴近实船行为的轨迹。
生成器网络通过遗忘门${f_t}$、输入门${i_t}$和输出门${O_t}$控制真实追越轨迹的流动和更新,具体的结构如图5所示。
1) 遗忘门${f_t}$
假设当前的输入数据为${R_i}$,优化后的生成器中遗忘门${f_t}$可以表示为
$ {f}_{t}=\sigma ({\boldsymbol{W}}_{f}\cdot[{H}_{t-1},{R}_{i}^{}]+{\boldsymbol{b}}_{f}) $
式中:${{\boldsymbol{W}}_f}$为遗忘门的权重矩阵;${H_{t - 1}}$为前一时间步的隐藏状态;${{\boldsymbol{b}}_f}$为遗忘门的偏置向量。
2) 输入门${i_t}$
输入门决定哪些新的信息需要被添加到细胞状态中,其由2个模块组成:sigmoid层和tanh层。其中,sigmoid层筛选数据信息保存到记忆细胞中并决定哪些值将被更新,如式(6)所示。
$ {i}_{t}=\sigma ({\boldsymbol{W}}_{i}\cdot[{H}_{t-1},{L}_{n}^{m}]+{\boldsymbol{b}}_{i}) $
式中:${{\boldsymbol{W}}_i}$为输入权重;${{\boldsymbol{b}}_i}$为输入门的偏置项。
tanh层生成一个新的候选细胞状态$ \tilde C_t $,把当前时刻传递的数据信息保存到记忆细胞中,如式(7)所示。
$ \tilde{C}_t=\mathrm{tanh}(\boldsymbol{W}_{\rm{c}}\cdot[H_{t-1},L_n^m]+\boldsymbol{b}_{\rm{c}}) $
式中:tanh为双曲线正切激活函数,其将实值压缩至区间(−1,1)内;${W_{\rm{c}}}$${b_{\rm{c}}}$分别为记忆细胞权重和记忆细胞偏置项。
新的记忆细胞${C_t}$的更新由输入门${i_t}$、遗忘门${f_t}$与候选细胞状态$\tilde C_t $共同决定,如式(8)所示。
$ {C}_{t}={f}_{t}\cdot{C}_{t-1}+{i}_{t}\cdot \tilde C_t $
3) 输出门${O_t}$
输出门${O_t}$计算需要输出的信息,传递隐藏状态${h_t}$
$ O_t=\sigma(\boldsymbol{W}_o\cdot[H_{t-1},L_n^m]+\boldsymbol{b}_o] $
$ {h}_{t}={O}_{t}\cdot\mathrm{tanh}({C}_{t}) $
式中:${{\boldsymbol{W}}_o}$${{\boldsymbol{b}}_o}$分别为输出门权重和记忆细胞偏置项;${h_{t - 1}}$为上一时刻的隐藏状态。
最后,优化后的生成器对输出状态${O_t}$进行自注意力计算,自注意力层通过计算轨迹序列中任意两个位置之间的相关性生成一个注意力权重矩阵,然后将该矩阵应用于序列表示上,以生成新的、包含全局信息的序列表示。这使得生成器可以根据船舶加、减速或转向的动作来预测其后期的航行路径,从而避免生成不合理的轨迹。
传统SeqGAN框架的判别器采用单一CNN结构进行特征提取,在处理船舶追越轨迹的复杂时空动态时,易出现梯度爆炸及特征表征能力不足的问题。为此,本文提出了如图6所示的改进判别器结构,通过在卷积层前、后分别嵌入SAM,显著提升轨迹特征的提取能力。
输入序列长度为N的追越轨迹后,首先进入第1层注意力机制,通过放缩参数H对序列各节点进行动态加权,生成大小为$N \times H$的加权轨迹矩阵。其中,高权重节点聚焦于航行状态变化特征,低权重节点对应相对稳定的航行阶段,从而实现关键特征的自适应强化。
随后,将加权矩阵输入卷积层,利用多尺度卷积核进行局部特征提取,然后通过卷积运算量化生成轨迹与真实数据的分布差异。为进一步增强全局特征捕获能力,在卷积层后引入第2层SAM(缩放参数m = 64),通过计算查询Q、键K和值V矩阵的相似度(式(11)),建模轨迹节点间的长程依赖关系。该机制能有效捕捉船舶轨迹中复杂的时空关联特性,例如当前转向动作与历史航向调整的逻辑耦合。
$ {\text{Attention}}\left( {{\boldsymbol{Q}},{\boldsymbol{K}},{\boldsymbol{V}}} \right) = {\text{soft}}\max \left( {\frac{{{\boldsymbol{Q}}{{\boldsymbol{K}}^N}}}{{\sqrt m }}} \right){\boldsymbol{V}} $
最终,池化层对特征向量降维并保留显著信息,分类层通过Softmax函数输出判别概率,反馈至生成器驱动参数优化。双注意力机制的协同作用不仅缓解了传统模型的梯度异常问题,更通过多层次特征融合提升了轨迹生成的逼真度与判别精度,可为后续高风险场景泛化奠定基础。
这种双层注意力机制的引入有效解决了传统SeqGAN在处理船舶轨迹数据时存在的梯度爆炸和特征提取能力不足的问题,提升了模型的鲁棒性和判别精度。
为了构建完整的测试场景,在高风险追越轨迹生成的基础上,需设置TS的初始状态。本文聚焦于两船之间的安全距离约束,避免生成背离航海实际的极端场景。通过分析船舶追越过程,综合考虑船舶尺寸、航速和操纵性能等因素,将以并行阶段两船最小横向距离${S_{\rm{c}}}$和切入阶段两船最小纵向距离${S_{\rm l}}$来确定TS的初始状态。
受船间效应及船舶操纵性的影响,追越过程中两船需保持足够的横向安全距离。当船舶改变航向进行追越时,会产生显著的横向位移[23]。为确保安全边界条件,排除极端情况,本文假设TS采取向OS相反方向的最大舵角转向作为避让策略。TS的横向位移ST与OS的横向位移SO
$ \left\{\begin{gathered}S_{\text{O}}=D_{\text{O}} \\ S_{\text{T}}=0.514\ 4v_{\text{T}}t_{90}\sin u \\ \end{gathered}\right. $
式中:DO为OS旋回横距;vT为TS航速;u为两船对地航向的差值;t90为TS航向改变90°所需时间[24]
进一步地,t90可由进距AT和旋回直径DT估算得出:
$ {t_{90}} = \frac{{0.67}}{{{v_{\text{T}}}}}\sqrt {A_{\text{T}}^2 + {{\left( {\frac{{{D_{\text{T}}}}}{2}} \right)}^2}} $
鉴于ATDT通常无法通过船载设备直接获取,本文使用TS船速vT和船长L作为估算依据[25],模型中使用
$ \left\{ \begin{gathered} {A_{\text{T}}} = {10^{0.36{v_{\text{T}}} + 0.095}}L \\ {D_{\text{T}}} = {10^{0.54{v_{\text{T}}} - 0.079}}L \\ \end{gathered} \right. $
此外,船间效应要求两船在近距离驶过时至少保持0.5L的横向间距,以避免相互间的不良影响。因此,并行阶段的两船最小横向安全距离设定为
$ {S_{\text{c}}} = \max \left( {\left| {{S_{\text{O}}} - {S_{\text{T}}}} \right|,\frac{L}{2}} \right) $
在船舶追越的切入阶段,OS应在其船尾超越TS船首后,保留足够的纵向安全距离,然后再切入TS船首方向,并于驶过让清后恢复原航向。然而,当遇到对向来船或航道内障碍物等紧急情况时,OS可能会被迫在尚未完全驶过TS船首的情况下快速改变航向以完成追越。这种紧急操作通常伴随着显著的航向变化,可能使OS直接逼近TS的航行路径,从而导致两船之间的最近会遇距离(distance to closest point of approach,DCPA)急剧减小。同时,即使OS理论上已完成切入,最近会遇时间(time to closest point of approach,TCPA)趋近于0,但因大幅度转向造成的严重降速仍可能形成暂时的反追越情况,存在碰撞风险。在此情况下,TS的自主避碰系统需要在较短的决策时间内执行转向或降速措施,确保两船之间的安全距离以避免碰撞。
不同于传统研究中以紧急倒车停船距离作为纵向安全距离的做法,本文考虑到倒车停船可能引发的横移以及低速失控风险,采用船舶转至新航向所需的最短航行距离作为切入阶段的纵向安全距离Sl。此时,两船不发生碰撞的临界条件是当OS完成切入时,TS立即向相反方向转向,经过一段时间后与OS保持相同的航向,并维持两船之间的最小距离。此时,TS转向避碰所需时间tl与纵向安全距离Sl分别为
$ {t_{\text{l}}} = {T'}\frac{{{L_{}}}}{{{v_{\text{T}}}}} + \frac{{{t_0}}}{2} + \frac{L}{{{K'}{v_{\text{T}}}}}\tan \frac{\theta }{2} $
$ {S_{\text{l}}} = 0.514\;4{v_{\text{T}}}{t_{\text{l}}} $
式中:${T'}$为量纲一的船舶追随性指数;${K'}$为量纲一的船舶旋回性指数;${t_0}$为TS船舶主机和驾驶员的反应时间,根据实际操船经验,确定为60 s;$\theta $为TS的航向变化角。
本文将OS的初始位置设置为坐标原点,因此,TS的纵向位置表示为${x_{{\text{TS}}}} = - {S_{\text{l}}}$,横向位置表示为${y_{{\text{TS}}}} = {S_{\text{c}}}$,从而得到TS的初始状态$ \left( {{x_{\rm{TS}}},{y_{\rm{TS}}},{v_{\rm{TS}}}} \right) $
为了高效且大规模生成船舶追越场景,本文结合OS的高风险追越轨迹生成和TS的初始状态约束,提出追越测试场景泛化生成方法。
首先,对船舶AIS数据进行预处理,包括清洗异常数据和填补缺失值,确保数据完整性。然后,通过位置特征和速度变化筛选出符合定义的104条真实追越轨迹,并将每条追越轨迹的起始点坐标进行归一化处理。在此基础上,利用高风险追越轨迹生成方法生成并筛选出一系列新的高风险追越轨迹。最后,引入船间安全距离作为约束条件,计算每条生成轨迹对应的TS的初始状态,确保测试场景的合理性。追越轨迹与对应的TS初始状态构成高风险追越测试场景,测试场景泛化生成的详细流程如下。
1) 输入:船舶AIS数据。
2) 数据预处理。
根据船舶相对位置筛选船舶AIS数据,得到104条真实追越轨迹${R_{\rm{OT}}} = \left\{ {{R_1}, \cdots ,{R_i}, \cdots ,{R_{104}}} \right\}$
提取真实追越轨迹每个坐标点y轴方向的位置坐标和x轴方向的速度,对每条轨迹的坐标值进行归一化处理,以使OS的轨迹起始点为原点,同时调整数据长度以保证一致性。
3) 追越轨迹生成。
以预处理后的真实追越轨迹作为输入,运用基于SAM-SeqGAN的高风险追越轨迹生成方法生成新的追越轨迹数据。
4) 初始状态计算。
使用初始状态约束模型分别计算每条高风险追越轨迹对应的船舶初始状态。
5) 高风险追越场景泛化生成。
将生成的追越轨迹与其对应的船舶初始状态相结合,构建完整的测试场景。
6) 输出:船舶追越测试场景。
本文使用船舶AIS数据中的真实追越轨迹对提出的高风险追越场景泛化生成方法进行实验验证。使用SAM-SeqGAN生成OS的高风险追越轨迹,然后结合基于安全距离的初始追越状态约束模型生成的船舶初始状态,实现测试场景的泛化生成。在仿真实验中,本文结合航海实际设置实验参数,借助python3.8和PyTorch1.10.0搭建深度学习框架。
SAM-SeqGAN包含3个超参数:生成器训练步数gstep、判别器训练步数dstep以及真实追越轨迹与生成轨迹数量的比例ratio。在模型训练过程中,增加真实追越轨迹数量能够同步降低生成器和判别器的训练误差。然而,由于SAM-SeqGAN模型的训练误差包含奖励机制,当判别器的能力远超生成器时,生成器会优先追求高奖励而非欺骗判别器,从而导致生成的追越轨迹严重偏离实际情况。因此,本文将超参数选取为gstep = 1,dstep = 1,ratio = 3,能够在满足小样本轨迹数据增强需求时兼顾对抗训练与高收益值。在初始状态约束模型中,使用TS在研究水域中的安全航速7 kn和100 m船长来验证方法的有效性。
实验参数的具体取值如表1所示。
测试场景中生成轨迹与实际轨迹的相似度越高,说明轨迹生成算法的准确性越好。因此,本文以真实轨迹位置的上下界建立轨迹缓冲区,通过比较生成轨迹的位置分布以及目标船切入时刻的速度分布来验证生成轨迹的准确性。轨迹缓冲区以及生成的高风险追越轨迹分布如图7所示。在SAM-SeqGAN模型生成的500条追越轨迹样本中,超过89.3%的轨迹位于预设的真实轨迹缓冲区内,经初始状态约束后,保留的轨迹中有97.3%位于真实轨迹缓冲区内,说明生成的高风险追越轨迹符合真实轨迹的位置分布。通过引入高风险特征和系统性调整关键参数,能够提升追越测试场景的风险性和多样性。与真实数据相比,生成场景具有更高的场景覆盖度,能够更高效地暴露自主避碰系统的性能边界。
轨迹中不仅包含位置特征,还含有速度特征。为了验证生成轨迹的速度相似性,本文对比了真实追越轨迹与生成轨迹中OS切入阶段起始时刻的速度分布,如图8所示。其中,生成的OS速度服从N(8.351, 3.4328)的正态分布,实际OS速度服从N(8.143, 3.5893)的正态分布。结果表明,生成的OS切入阶段起始时刻的速度分布与实际速度分布接近。
生成对抗网络的损失值降低通常意味着生成数据质量的提升。为了验证本文所提SAM-SeqGAN算法的生成性能,将其与传统GAN,SeqGAN网络的损失值进行了对比,结果如图9所示。SAM-SeqGAN的损失值比其他2种网络的都要低,并且收敛速度更快。这一结果表明,SAM-SeqGAN的训练效率以及生成数据质量优于未经优化的GAN和SeqGAN网络,模型能够更加有效地捕捉数据分布特征,生成更符合实际情况的船舶轨迹序列。
高风险追越轨迹生成以后,利用2.2节所述追越初始状态约束模型计算得到的每个轨迹所对应的船舶初始状态。每条高风险追越轨迹都对应一个特定的初始状态,共同构成一个完整的高风险测试用例。通过调整切入时刻两船相对航速、相对距离等参数,可以获取具有不同碰撞风险的测试场景,从而满足不同的测试需求。设定TS的初始位置位于原点(0,0),以7 kn的航速沿x轴方向行驶。OS与TS的初始横向距离设置为1.35 n mile,并以10 kn的航速进行追越,从坐标(1.5,1.3)处开始切入,该场景的仿真持续时间为174 s。
图10所示,蓝色曲线表示OS的轨迹,红色曲线表示TS的轨迹,红色曲线的终点表示TS采取向OS反方向转向避碰后,与OS保持相同航向,并成功避免碰撞。
将生成的高风险追越轨迹及船舶初始状态导入Unreal Engine虚拟引擎,搭建基于虚拟测试的高风险追越场景。在虚拟引擎中复制实际海上环境,把生成的OS轨迹和计算得到的TS初始位置坐标导入数据库,通过获取数据库中的船舶运动状态信息,构建船舶追越测试场景,其效果如图11所示。图中,蓝色船表示OS,红色表示TS。
虚拟仿真追越测试场景呈现了船舶追越的动态演化过程,依次呈现了起始、转向、并行及切入4个阶段。在起始阶段,OS从船尾方向逼近TS,以模拟实际航行过程中的典型追越态势;在转向阶段,通过航向与航速的协同调整加剧两船交互的风险;在并行阶段,验证自主避碰系统对动态干扰的感知能力;最终,在切入阶段通过转向操作触发碰撞临界状态,暴露自主避碰系统的决策延迟与容错边界。该场景通过系统性地注入高风险特征,复现了追越过程中因操纵失误或环境耦合导致的极端工况,可为评估自主避碰算法在复杂场景下的鲁棒性提供可量化、可复现的测试基准。
本文提出了一种数据−模型驱动的船舶自主避碰追越测试场景生成方法,能够生成大量符合航海实际的高风险追越场景,减少测试所需的时间和资源,同时为智能船舶自主避碰系统在追越场景下的性能评估提供可行的测试场景。首先,利用注意力机制优化SeqGAN网络,构建SAM-SeqGAN模型,高效生成符合航海实际的高风险船舶追越轨迹;然后,基于安全距离设计初始状态约束模型,避免极端场景,增强测试合理性;最后,结合泛化算法,实现追越测试场景的自动化生成。实验结果表明,采用本文方法能够有效模拟船舶实际追越过程中的复杂动态,生成兼具危险性和真实性的测试场景。这种主动风险注入机制可高效暴露自主避碰系统的性能边界,同时减少对稀缺真实危险数据的依赖,能为智能船舶自主避碰测试提供一种可扩展的解决方案,有助于减少测试时间和资源消耗,加速自主避碰技术的研发及商业化应用。
需要特别指出的是,本文的场景生成目标是通过极限压力测试暴露系统潜在的缺陷,而非鼓励智能船舶在实际航行中主动逼近安全边界。测试结果将用于优化算法逻辑,确保自主避碰系统在追越过程中合理操纵船舶以维持安全冗余。未来的研究将聚焦于优化轨迹生成模型和安全距离约束模型,用以提升场景生成能力,并以此为基础扩展研究智能船舶在其他会遇场景下的自主避碰性能。
参考文献 引证文献
排序方式:
1
ZHANG X Y, WANG C B, JIANG L L, et al. Collision-avoidance navigation systems for maritime autonomous surface ships: a state of the art survey[J]. Ocean Engineering, 2021, 235: 109380.
2
中国船级社. 智能船舶规范[S]. 北京: 人民交通出版社, 2024.
China Classification Society. Rules for intelligent ships[S]. Beijing: China Communications Press, 2024 (in Chinese).
3
郑茂, 丁世淦, 兰加芬, 等. 内河航道避碰测试会遇场景建模方法研究[J]. 中国舰船研究, 2023, 18(5): 121–132.
ZHENG M, DING S G, LAN G F, et al. Study on scenario modeling method for collision avoidance test in inland waterway[J]. Chinese Journal of Ship Research, 2023, 18(5): 121–132 (in Chinese).
4
张笛, 赵银祥, 崔一帆, 等. 智能船舶的研究现状可视化分析与发展趋势[J]. 交通信息与安全, 2021, 39(1): 7–16, 34.
ZHANG D, ZHAO Y X, CUI Y F, et al. A visualization analysis and development trend of intelligent ship studies[J]. Journal of Transport Information and Safety, 2021, 39(1): 7–16, 34 (in Chinese).
5
GUO W Q, ZHANG X Y, LIU W, et al. Mixed traffic conditions of autonomous and human-driven ships: assessing channel traffic capacity bounds and optimizing channel management[J]. Ocean Engineering, 2024, 314: 119734.
6
张浩杰, 苏治宝, 杨甜甜. 基于USARSim和ROS的无人平台编队仿真系统[J]. 自动化学报, 2021, 47(6): 1390–1400.
ZHANG H J, SU Z B, YANG T T. Design of team formation simulation system for unmanned ground vehicles based on USARSim and ROS[J]. Acta Automatica Sinica, 2021, 47(6): 1390–1400 (in Chinese).
7
张英俊, 翟鹏宇. 海运船舶自主避碰技术研究进展与趋势[J]. 大连海事大学学报, 2022, 48(3): 1–11.
ZHANG Y J, ZHAI P Y. Research progress and trend of autonomous collision avoidance technology for marine ships[J]. Journal of Dalian Maritime University, 2022, 48(3): 1–11 (in Chinese).
8
ISO. Road vehicles—safety of the intended functionality: ISO/PAS 21448: 2019[S]. Vernier: ISO, 2019.
9
翟强, 程洪, 黄瑞, 等. 智能汽车中人工智能算法应用及其安全综述[J]. 电子科技大学学报, 2020, 49(4): 490–498, 510.
ZHAI Q, CHENG H, HUANG R, et al. A survey: artificial intelligence and its security in intelligent vehicle[J]. Journal of University of Electronic Science and Technology of China, 2020, 49(4): 490–498, 510 (in Chinese).
10
陈立家, 刘岩植, 许毅, 等. AIS数据驱动的船舶智能避碰测试场景构建方法[J]. 武汉理工大学学报(交通科学与工程版), 2025, 49(1): 195–202, 209.
CHEN L J, LIU Y Z, XU Y, et al. Construction method of intelligent collision avoidance test scene for ships based on AIS data-driven[J]. Journal of Wuhan University of Technology (Transportation Science & Engineering), 2025, 49(1): 195–202, 209 (in Chinese).
11
STANKIEWICZ P G, MULLINS G E. Improving evaluation methodology for autonomous surface vessel COLREGs compliance[C]//OCEANS 2019-Marseille. Marseille, France: IEEE, 2019: 1−7.
12
PORRES I, AZIMI S, LILIUS J. Scenario-based testing of a ship collision avoidance system[C]//2020 46th Euromicro Conference on Software Engineering and Advanced Applications (SEAA). Portoroz, Slovenia: IEEE, 2020: 545−552.
13
WANG X Y, WANG Y H, MA I, et al. Research on the process-based generation method of scenarios database for maritime autonomous surface ship test[C]//ICETIS 2022; 7th International Conference on Electronic Technology and Information Science. Harbin, China: VDE, 2022: 1−8.
14
陈立家, 王凯, 黄立文, 等. 船舶自主航行虚拟测试场景研究进展[J]. 中国舰船研究, 2025, 20(1): 25–37.
CHEN L J, WANG K, HUANG L W, et al. Research progress on test scenario of ship autonomous navigation[J]. Chinese Journal of Ship Research, 2025, 20(1): 25–37 (in Chinese).
15
SNIJDERS R, ELROFAI H. Scenario identification for safety assessment of autonomous shipping using AIS data[C]//15th International Naval Engineering Conference & Exhibition. Delft, 2020.
16
BOLBOT V, GKEREKOS C, THEOTOKATOS G, et al. Automatic traffic scenarios generation for autonomous ships collision avoidance system testing[J]. Ocean engineering, 2022, 254: 111309.
17
丁志国, 张新宇, 王程博, 等. 基于驾驶实践的无人船智能避碰决策方法[J]. 中国舰船研究, 2021, 16(1): 96–104, 113.
DING Z G, ZHANG X Y, WANG C B, et al. Intelligent collision avoidance decision-making method for unmanned ships based on driving practice[J]. Chinese Journal of Ship Research, 2021, 16(1): 96–104, 113 (in Chinese).
18
CHAUVIN C, LARDJANE S. Decision making and strategies in an interaction situation: collision avoidance at sea[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2008, 11(4): 259–269.
19
赵祥模, 赵玉钰, 景首才, 等. 面向自动驾驶测试的危险变道场景泛化生成[J]. 自动化学报, 2023, 49(10): 2211–2223.
ZHAO X M, ZHAO Y Y, JING S C, et al. Generalization generation of hazardous lane-changing scenarios for automated vehicle testing[J]. Acta Automatica Sinica, 2023, 49(10): 2211–2223 (in Chinese).
20
GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems. Montreal: MIT Press, 2014: 2672−2680.
21
JING S C, ZHAO Y Y, ZHAO X M, et al. An efficient high-risk lane-changing scenario edge cases generation method for autonomous vehicle safety testing[J/OL]. IEEE Transactions on Intelligent Vehicles, 2024. https://dx.doi.org/10.1109/TIV.2024.3414828.
22
ZHANG H, GOODFELLOW I, METAXAS D, et al. Self-attention generative adversarial networks[C]//Proceedings of the 36th International Conference on Machine Learning. Long Beach, CA, USA: PMLR, 2019: 7354−7363.
23
刘钊, 谢金鹏, 张明阳, 等. 考虑避让紧迫度的数字化船舶领域模型[J]. 中国航海, 2021, 44(2): 15–20.
LIU Z, XIE J P, ZHANG M Y, et al. A numerical model of ship domain reflecting collision urgency ditribution[J]. Navigation of China, 2021, 44(2): 15–20 (in Chinese).
24
KIJIMA K, FURUKAWA Y. Automatic collision avoidance system using the concept of blocking area[J]. IFAC Proceedings Volumes, 2003, 36(21): 223–228.
25
李元奎, 彭飞, 贾梓淇, 等. 基于行船视距的船舶会遇安全距离研究[J]. 安全与环境学报, 2024, 24(4): 1442–1449.
LI Y K, PENG F, JIA Z Q, et al. Study on safe distances of ships encounter situations based on sight distance[J]. Journal of Safety and Environment, 2024, 24(4): 1442–1449 (in Chinese).
2026年第21卷第2期
PDF下载
18
5
引用本文
BibTeX
文章信息
doi: 10.19693/j.issn.1673-3185.04271
  • 接收时间:2024-11-13
  • 首发时间:2026-05-20
  • 出版时间:2026-04-30
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-11-13
  • 修回日期:2025-03-04
基金
作者信息
    1大连海事大学 航海学院,辽宁 大连 116026
    2大连海事局,辽宁 大连 116001

通讯作者:

* 张新宇
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/zgjcyj/CN/10.19693/j.issn.1673-3185.04271
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