Article(id=1281326861871452194, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1281326807345500788, articleNumber=null, orderNo=null, doi=10.3969/j.issn.1007-7294.2025.12.003, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1751558400000, receivedDateStr=2025-07-04, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1783421729772, onlineDateStr=2026-07-07, pubDate=1765728000000, pubDateStr=2025-12-15, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1783421729772, onlineIssueDateStr=2026-07-07, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1783421729772, creator=13701087609, updateTime=1783421729772, updator=13701087609, issue=Issue{id=1281326807345500788, tenantId=1146029695717560320, journalId=1240685776644648972, year='2025', volume='29', issue='12', pageStart='1827', pageEnd='1990', issueExtLink='null', onlineDate='null', pubDate='1765728000000', pubDateStr='2025-12-15', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1783421716772, creator='13701087609', updateTime=1783422145004, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1281328603572977733, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1281326807345500788, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1281328603572977734, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1281326807345500788, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=1848, endPage=1861, ext={EN=ArticleExt(id=1281326862148276259, articleId=1281326861871452194, tenantId=1146029695717560320, journalId=1240685776644648972, language=EN, title=Data-driven and machine learning-based optimization design for AUV shape, columnId=1241023037940748650, journalTitle=Journal of Ship Mechanics, columnName=Hydrodynamics, runingTitle=null, highlight=null, articleAbstract=

The performance of Autonomous Underwater Vehicles (AUVs) is significantly influenced by their shape design. This study presents a novel AUV shape optimization method that integrates data-driven approaches and machine learning technique to focus on the impact of the head and tail profiles, and the configuration of the fins and rudders on AUV performance. A parameterized hydrodynamic analysis workflow was developed to automatically generate the hydrodynamic parameters required during the optimization process. Based on this workflow, a data-driven framework was constructed for multi-objective optimization of AUV shapes, with the aims of minimizing drag and maximizing maneuverability. Surrogate models for the two optimization objectives above were built using Multilayer Perceptron (MLP) neural networks and ensemble learning methods respectively, and their performances were compared with traditional surrogate models. The optimization problem was solved using the Non-dominated Sorting Genetic Algorithm II. Comparative analysis of the initial and optimized AUV shapes demonstrates significant improvements in hydrodynamic performance, confirming the feasibility and effectiveness of the proposed method.

, authors=Yun-tian LI1, 2, Wei-zheng CHEN1, Qing HAI1, Wei-ye CHEN1, authorsList=Yun-tian LI, Wei-zheng CHEN, Qing HAI, Wei-ye CHEN, authorCompany=null, correspAuthors=Wei-zheng CHEN, authorNote=null, correspAuthorsNote=null, copyrightStatement=Copyright ©2025 Journal of Ship Mechanics. 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, fund=null), CN=ArticleExt(id=1281326900945588661, articleId=1281326861871452194, tenantId=1146029695717560320, journalId=1240685776644648972, language=CN, title=基于数据驱动和机器学习的AUV外形优化设计, columnId=1241023038087549292, journalTitle=船舶力学, columnName=流体力学, runingTitle=null, highlight=null, articleAbstract=

自主水下航行器(AUV)的性能在很大程度上受其外形设计的影响。本文提出了一种AUV外形优化方法,通过集成数据驱动与机器学习技术,重点关注头部与尾部线型以及鳍舵布局对AUV性能的影响。本研究开发了一个参数化的水动力分析流程,用于自动生成优化过程中所需的水动力参数。我们基于此流程,构建了一个数据驱动框架,用于实现AUV外形的多目标优化,目标是最小化阻力并最大化机动性。针对这两个优化目标,本研究分别采用多层感知器(MLP)神经网络和集成学习方法构建了代理模型,并将它们的性能与传统代理模型进行了对比。优化问题通过非支配排序遗传算法求解。对比分析初始与优化后的AUV外形表明,其水动力性能显著提升,从而证实了所提方法的可行性与有效性。

, authors=李云天1, 2, 陈伟政1, 海青1, 陈炜烨1, authorsList=李云天, 陈伟政, 海青, 陈炜烨, authorCompany=null, correspAuthors=陈伟政, authorNote=

李云天(2000–),男,硕士研究生

, correspAuthorsNote=
陈伟政(1974−),男,博士,研究员,通讯作者,E-mail:
, copyrightStatement=版权所有©《船舶力学》编辑部2025, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=67dUPRiBJ0uuFfzxEX/sTg==, magXml=ZDoq4ViIRDlg8O+XL3QNDQ==, pdfUrl=null, pdf=euA184bq59Ro4+LQCknMVw==, pdfFileSize=874139, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=Xias9E6wN5YaVC9K8CuGlQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=dvZ8cOBE2A3X9CR9oTtvdQ==, mapNumber=null, fund=null)}, authors=[Author(id=1281326901427933629, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1281326901532791232, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901427933629, language=EN, stringName=Yun-tian LI, firstName=Yun-tian, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.China Ship Scientific Research Center, Wuxi 214082, China
2.CSSC Marine Power Zhenjiang Co., Ltd., Zhenjiang 212005, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326901612483009, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901427933629, language=CN, stringName=李云天, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.中国船舶科学研究中心,江苏 无锡 214082
2.中船动力镇江有限公司,江苏 镇江 212005, bio={"content":"

李云天(2000–),男,硕士研究生

"}, bioImg=null, bioContent=

李云天(2000–),男,硕士研究生

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)]), AuthorCompany(id=1281326901323076025, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=2., ext=[AuthorCompanyExt(id=1281326901331464634, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901323076025, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.CSSC Marine Power Zhenjiang Co., Ltd., Zhenjiang 212005, China), AuthorCompanyExt(id=1281326901339853243, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901323076025, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中船动力镇江有限公司,江苏 镇江 212005)])]), Author(id=1281326901717340611, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=chenwz.mail@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1281326901809615301, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901717340611, language=EN, stringName=Wei-zheng CHEN, firstName=Wei-zheng, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326901885112774, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901717340611, language=CN, stringName=陈伟政, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])]), Author(id=1281326901964804552, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=1281326902052884938, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901964804552, language=EN, stringName=Qing HAI, firstName=Qing, middleName=null, lastName=HAI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326902124188107, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901964804552, language=CN, stringName=海青, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])]), Author(id=1281326902203879885, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=1281326902275183055, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326902203879885, language=EN, stringName=Wei-ye CHEN, firstName=Wei-ye, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326902354874832, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326902203879885, language=CN, stringName=陈炜烨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])])], keywords=[Keyword(id=1281326902455538129, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, orderNo=1, keyword=Autonomous Underwater Vehicle), Keyword(id=1281326902535229906, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, orderNo=2, keyword=shape design), Keyword(id=1281326902610727379, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, orderNo=3, keyword=CFD simulation), Keyword(id=1281326902686224852, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, orderNo=4, keyword=multi-objective optimization), Keyword(id=1281326902774305237, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, orderNo=5, keyword=machine learning), Keyword(id=1281326902849802710, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, orderNo=6, keyword=parametric design), Keyword(id=1281326902929494487, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, orderNo=1, keyword=自主水下航行器), Keyword(id=1281326903193735640, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, orderNo=2, keyword=外形设计), Keyword(id=1281326903273427417, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, orderNo=3, keyword=CFD仿真), Keyword(id=1281326903340536282, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, orderNo=4, keyword=多目标优化), Keyword(id=1281326903420228059, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, orderNo=5, keyword=机器学习), Keyword(id=1281326903499919836, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, orderNo=6, keyword=参数化设计)], refs=[Reference(id=1281326907115409923, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=宋保维, 潘 光, 张立川, 等. 自主水下航行器发展趋势及关键技术[J]. 中国舰船研究, 2022, 17(5): 27‒44., articleTitle=null, refAbstract=null), Reference(id=1281326907186713092, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=Song B W, Pan G, Zhang L C, et al. Development trend and key technologies of autonomous underwater vehicles[J]. Chinese Journal of Ship Research, 2022, 17(5): 27‒44. (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1281326907287376389, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=Hu F, Huang Y, Xie Z, et al. Conceptual design of a long-range autonomous underwater vehicle based on multidisciplinary optimization framework[J]. Ocean Engineering, 2022, 248: 110684., articleTitle=null, refAbstract=null), Reference(id=1281326907409011206, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=Sener M Z, Aksu E. The effects of head form on resistance performance and flow characteristics for a streamlined AUV hull design[J]. Ocean Engineering, 2022, 257: 111630., articleTitle=null, refAbstract=null), Reference(id=1281326907463537159, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen X, Yu L, Liu L Y, et al. Multi-objective shape optimization of autonomous underwater vehicle by coupling CFD simulation with genetic algorithm[J]. Ocean Engineering, 2023, 286: 115722., articleTitle=null, refAbstract=null), Reference(id=1281326907526451720, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen W, Dong H, Wang P, et al. Causal-relationship-assisted shape design optimization for blended-wing-body underwater gliders[J]. Engineering Optimization, 2023, 6., articleTitle=null, refAbstract=null), Reference(id=1281326907593560585, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=Abbasi S, Zeinali M, Nejadabbasi P. Autonomous underwater vehicle hull geometry optimization using a multi-objective algorithm approach[J]. International Journal of Engineering, 2018, 31(9): 1593‒1601., articleTitle=null, refAbstract=null), Reference(id=1281326907669058058, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Hou Y H, Liang X. Uncertainty optimization design of an axially symmetric body using speed perturbation[J]. Journal of Marine Science and Technology, 2019, 24(2): 612‒619., articleTitle=null, refAbstract=null), Reference(id=1281326907731972619, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=Lu Y, Yuan J, Si Q, et al. Study on the optimal design of a shark-like shape AUV based on the CFD method[J]. Journal of Marine Science and Engineering, 2023, 11(10): 1869., articleTitle=null, refAbstract=null), Reference(id=1281326907807470092, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Wen Q, Feng R, An X, et al. Optimal design of an autonomous underwater helicopter’s shape based on combinatorial optimization strategy[J]. Ocean Engineering, 2022, 266: 113015., articleTitle=null, refAbstract=null), Reference(id=1281326907887161869, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=Luo W, Guo X, Dai J, et al. Hull optimization of an underwater vehicle based on dynamic surrogate model[J]. Ocean Engineering, 2021, 230: 109050., articleTitle=null, refAbstract=null), Reference(id=1281326907962659342, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu M, Song Y, Yang S, et al. Rapid data-driven individualised shape design of AUVs based on CFD and machine learning[J]. Ships and Offshore Structures, 2024, 19(10112): 2050−2064., articleTitle=null, refAbstract=null), Reference(id=1281326908042351119, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=Xu Y, Liu P, Wang L, et al. Main body shape optimization of non-body-of-revolution underwater vehicles by using CNN and genetic algorithm[J]. Ocean Engineering, 2024, 295: 116938., articleTitle=null, refAbstract=null), Reference(id=1281326908101071376, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen X, Wang P, Zhang D, et al. Gradient-based multidisciplinary design optimization of an autonomous underwater vehicle[J]. Applied Ocean Research, 2018, 80: 101‒111., articleTitle=null, refAbstract=null), Reference(id=1281326908172374545, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Granville P S. Geometrical characteristics of streamlined shapes[R]. 1969., articleTitle=null, refAbstract=null), Reference(id=1281326908243677714, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Saha S, Gan Z, Cheng L, et al. Hierarchical deep learning neural network (HiDeNN): An artificial intelligence (AI) framework for computational science and engineering[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 373: 113452., articleTitle=null, refAbstract=null), Reference(id=1281326908306592275, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=潘 光. 鱼雷力学 [M]. 西安: 陕西师范大学出版社, 2013: 120−150., articleTitle=null, refAbstract=null), Reference(id=1281326908407255572, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=Pan G. Torpedo mechanics[M]. Xi’an: Shaanxi Normal University Press, 2013: 120−150. (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1281326908491141653, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=宋保维, 王 鹏. 鱼雷多学科设计优化理论与应用研究[M]. 西安: 西北工业大学出版社, 2016: 18−53., articleTitle=null, refAbstract=null), Reference(id=1281326908562444822, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Song B W, Wang P. Research on the theory and application of torpedo multidisciplinary design optimization[M]. Xi’an: Northwestern Polytechnical University Press, 2016: 18−53. (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1281326908625359383, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang J, Li Y, Gao R X, et al. Hybrid physics-based and data-driven models for smart manufacturing: Modelling, simulation, and explainability[J]. Journal of Manufacturing Systems, 2022, 63: 381‒391., articleTitle=null, refAbstract=null), Reference(id=1281326908709245464, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=Yang M, Wang Y, Chen Y, et al. Data-driven optimization design of a novel pressure hull for AUV[J]. Ocean Engineering, 2022, 257: 111562., articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)]), AuthorCompany(id=1281326901323076025, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=2., ext=[AuthorCompanyExt(id=1281326901331464634, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901323076025, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.CSSC Marine Power Zhenjiang Co., Ltd., Zhenjiang 212005, China), AuthorCompanyExt(id=1281326901339853243, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901323076025, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中船动力镇江有限公司,江苏 镇江 212005)])], figs=[ArticleFig(id=1281326903659303389, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.1, caption=Workflow of hydrodynamic analysis, figureFileSmall=axNaOlf84aGwWG7YmB1/JQ==, figureFileBig=Xias9E6wN5YaVC9K8CuGlQ==, tableContent=null), ArticleFig(id=1281326903726412254, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图1, caption=水动力分析流程, figureFileSmall=axNaOlf84aGwWG7YmB1/JQ==, figureFileBig=Xias9E6wN5YaVC9K8CuGlQ==, tableContent=null), ArticleFig(id=1281326904007430623, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.2, caption=Parameter definitions of AUV, figureFileSmall=P3LaDTMdJxXAoQm4rDRN1A==, figureFileBig=hr02zB1K+30Hb7RWmK5xzA==, tableContent=null), ArticleFig(id=1281326904082928096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图2, caption=AUV外形参数定义, figureFileSmall=P3LaDTMdJxXAoQm4rDRN1A==, figureFileBig=hr02zB1K+30Hb7RWmK5xzA==, tableContent=null), ArticleFig(id=1281326904187785697, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.3, caption=Computational domain of AUV, figureFileSmall=pRuCDK+5uAZ9rhEnzaInSg==, figureFileBig=EplQIOdkTHNElyadWfib7A==, tableContent=null), ArticleFig(id=1281326904263283170, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图3, caption=AUV计算域, figureFileSmall=pRuCDK+5uAZ9rhEnzaInSg==, figureFileBig=EplQIOdkTHNElyadWfib7A==, tableContent=null), ArticleFig(id=1281326904351363555, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.4, caption=Automatically generated unstructured mesh for the AUV, figureFileSmall=JQX+s2R+Wr6vTdlqzC6p1A==, figureFileBig=UVSII+PhD5uBKibUYfxIVg==, tableContent=null), ArticleFig(id=1281326904439443940, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图4, caption=自动生成的非结构化网格, figureFileSmall=JQX+s2R+Wr6vTdlqzC6p1A==, figureFileBig=UVSII+PhD5uBKibUYfxIVg==, tableContent=null), ArticleFig(id=1281326904502358501, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.5, caption=Values of $ {{{y}}^ + } $ under a total number of 3.01 million cells, figureFileSmall=uwHegckQh49ZrsiVMSJtwg==, figureFileBig=ikL03ULiqf9qg6eRQjzTew==, tableContent=null), ArticleFig(id=1281326904565273062, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图5, caption=301万网格下的$ {{{y}}^ + } $, figureFileSmall=uwHegckQh49ZrsiVMSJtwg==, figureFileBig=ikL03ULiqf9qg6eRQjzTew==, tableContent=null), ArticleFig(id=1281326904653353447, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.6, caption=Data-driven framework for AUV hydrodynamic shape design, figureFileSmall=2qpI0666OB2ea/1BGOUAIw==, figureFileBig=f8MgU+bre0XOzwxKuIGbSg==, tableContent=null), ArticleFig(id=1281326904724656616, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图6, caption=AUV外形设计的数据驱动框架, figureFileSmall=2qpI0666OB2ea/1BGOUAIw==, figureFileBig=f8MgU+bre0XOzwxKuIGbSg==, tableContent=null), ArticleFig(id=1281326904800154089, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.7, caption=Platform of data collection, figureFileSmall=Bj3W8V/FLRNwTzyiqDLWPw==, figureFileBig=Qu9fQ/47UcniN1RDGvloaw==, tableContent=null), ArticleFig(id=1281326904879845866, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图7, caption=数据收集平台, figureFileSmall=Bj3W8V/FLRNwTzyiqDLWPw==, figureFileBig=Qu9fQ/47UcniN1RDGvloaw==, tableContent=null), ArticleFig(id=1281326904951149035, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.8, caption=Distribution of new data points, figureFileSmall=yyYv7gOIFPp6LBSjbG2myA==, figureFileBig=J1UM6LmoSukABz5qZaTANw==, tableContent=null), ArticleFig(id=1281326905026646508, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图8, caption=新样本点的空间分布, figureFileSmall=yyYv7gOIFPp6LBSjbG2myA==, figureFileBig=J1UM6LmoSukABz5qZaTANw==, tableContent=null), ArticleFig(id=1281326905118921197, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.9, caption=Comparison between actual and predicted value, figureFileSmall=RH9r+cm7NJYcLzUMJwuU/g==, figureFileBig=Au3l77I5VtFjTtpWDy2aBQ==, tableContent=null), ArticleFig(id=1281326905190224366, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图9, caption=实际值与预测值对比图, figureFileSmall=RH9r+cm7NJYcLzUMJwuU/g==, figureFileBig=Au3l77I5VtFjTtpWDy2aBQ==, tableContent=null), ArticleFig(id=1281326905261527535, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.10, caption=Histogram of error distribution, figureFileSmall=3hT7vwJAT2Gsyk4W9T923Q==, figureFileBig=otwn115F+OuQIRKUvedRpg==, tableContent=null), ArticleFig(id=1281326905337025008, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图10, caption=误差分布直方图, figureFileSmall=3hT7vwJAT2Gsyk4W9T923Q==, figureFileBig=otwn115F+OuQIRKUvedRpg==, tableContent=null), ArticleFig(id=1281326905433494001, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.11, caption=Diagram of sensitivity analysis, figureFileSmall=4eiRHdUxHNz3ppsf0OvcAA==, figureFileBig=IrvJVOhJXtbIiemp8EK+FQ==, tableContent=null), ArticleFig(id=1281326905513185778, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图11, caption=灵敏度分析图, figureFileSmall=4eiRHdUxHNz3ppsf0OvcAA==, figureFileBig=IrvJVOhJXtbIiemp8EK+FQ==, tableContent=null), ArticleFig(id=1281326905592877555, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.12, caption=Multi-objective optimization results, figureFileSmall=6/XljW50ZLb0xNfxeh0+Cg==, figureFileBig=6jeVH/c1+gWF0pc7fwc6Pg==, tableContent=null), ArticleFig(id=1281326905655792116, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图12, caption=多目标优化结果, figureFileSmall=6/XljW50ZLb0xNfxeh0+Cg==, figureFileBig=6jeVH/c1+gWF0pc7fwc6Pg==, tableContent=null), ArticleFig(id=1281326905731289589, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Fig.13, caption=Comparison of AUV models before and after optimization, figureFileSmall=cJzPIGYJvhX6RC7jNuzz3g==, figureFileBig=Yln97uEGgHBqJ0fhh6oV5A==, tableContent=null), ArticleFig(id=1281326905802592758, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=图13, caption=优化前后AUV模型对比, figureFileSmall=cJzPIGYJvhX6RC7jNuzz3g==, figureFileBig=Yln97uEGgHBqJ0fhh6oV5A==, tableContent=null), ArticleFig(id=1281326905911644663, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Tab.1, caption=

Design variables of AUV

, figureFileSmall=null, figureFileBig=null, tableContent=
设计变量名称符号/单位设计变量名称符号/单位
头部线型可调参数 $ q{h_1},q{h_2} $根弦长 ${b_0}/{\text{m}}$
尾部线型可调参数 $ q{t_1},q{t_2} $梢弦长 ${b_1}/{\text{m}}$
鳍舵总展长 $ a/{\text{m}} $舵后缘距尾端面距离 ${b_{\text{s}}}/{\text{m}}$
), ArticleFig(id=1281326906003919352, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=表1, caption=

AUV设计变量

, figureFileSmall=null, figureFileBig=null, tableContent=
设计变量名称符号/单位设计变量名称符号/单位
头部线型可调参数 $ q{h_1},q{h_2} $根弦长 ${b_0}/{\text{m}}$
尾部线型可调参数 $ q{t_1},q{t_2} $梢弦长 ${b_1}/{\text{m}}$
鳍舵总展长 $ a/{\text{m}} $舵后缘距尾端面距离 ${b_{\text{s}}}/{\text{m}}$
), ArticleFig(id=1281326906071028217, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Tab.2, caption=

Comparison under different total grid number

, figureFileSmall=null, figureFileBig=null, tableContent=
网格数 $ {{{y}}^ + } $ $ {C_{xS}} $
255万600.10959
301万550.10809
351万500.10794
401万500.10786
511万500.10765
), ArticleFig(id=1281326906163302906, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=表2, caption=

不同网格数$ {C_{xS}} $$ {{{y}}^ + } $值比较

, figureFileSmall=null, figureFileBig=null, tableContent=
网格数 $ {{{y}}^ + } $ $ {C_{xS}} $
255万600.10959
301万550.10809
351万500.10794
401万500.10786
511万500.10765
), ArticleFig(id=1281326906234606075, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Tab.3, caption=

Sample space of design variables

, figureFileSmall=null, figureFileBig=null, tableContent=
设计变量取值范围数据类型设计变量取值范围数据类型
$q{h_1}$1.0~10.0实数 $a$300.0~500.0实数
$q{h_2}$0.0~20.0实数 ${b_0}$400.0~800.0实数
$q{t_1}$3.0~11.0实数 ${b_1}$200.0~600.0实数
$q{t_2}$3.0~11.0实数 ${b_{\mathrm{s}}}$90.0~300.0实数
), ArticleFig(id=1281326906343657980, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=表3, caption=

设计变量的样本空间

, figureFileSmall=null, figureFileBig=null, tableContent=
设计变量取值范围数据类型设计变量取值范围数据类型
$q{h_1}$1.0~10.0实数 $a$300.0~500.0实数
$q{h_2}$0.0~20.0实数 ${b_0}$400.0~800.0实数
$q{t_1}$3.0~11.0实数 ${b_1}$200.0~600.0实数
$q{t_2}$3.0~11.0实数 ${b_{\mathrm{s}}}$90.0~300.0实数
), ArticleFig(id=1281326906402378237, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Tab.4, caption=

Comparison of predictive performance of various surrogate models

, figureFileSmall=null, figureFileBig=null, tableContent=
代理模型最大预测误差 $ {R^2} $
Kriging model0.012050.4855
RSM0.030900.8221
RBF0.040750.7701
MLP0.00950.9490
), ArticleFig(id=1281326906565956094, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=表4, caption=

不同代理模型的预测性能比较

, figureFileSmall=null, figureFileBig=null, tableContent=
代理模型最大预测误差 $ {R^2} $
Kriging model0.012050.4855
RSM0.030900.8221
RBF0.040750.7701
MLP0.00950.9490
), ArticleFig(id=1281326906662425087, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Tab.5, caption=

Performance comparison between ensemble and individual models

, figureFileSmall=null, figureFileBig=null, tableContent=
预测模型MSE $ {R^2} $
随机森林13.25380.5323
SVR16.78160.4079
弹性网络9.61410.6608
集成学习方法10.89700.7155
), ArticleFig(id=1281326906746311168, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=表5, caption=

集成模型与单个模型的性能比较

, figureFileSmall=null, figureFileBig=null, tableContent=
预测模型MSE $ {R^2} $
随机森林13.25380.5323
SVR16.78160.4079
弹性网络9.61410.6608
集成学习方法10.89700.7155
), ArticleFig(id=1281326906935054849, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=EN, label=Tab.6, caption=

Performance comparison between ensemble method and individual models

, figureFileSmall=null, figureFileBig=null, tableContent=
变量名 $q{h_1}$ $q{h_2}$ $q{t_1}$ $q{t_2}$ $a$ ${b_0}$ ${b_1}$ ${b_{\mathrm{s}}}$ ${C_{xS}}$ $R$
初始值2.08.03.03.0338.5400.0320900.111011.7056
优化后值1.74.04.05.0464.1533.8397.1141.70.105910.6140
), ArticleFig(id=1281326907018940930, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, language=CN, label=表6, caption=

集成模型与单个模型的性能比较

, figureFileSmall=null, figureFileBig=null, tableContent=
变量名 $q{h_1}$ $q{h_2}$ $q{t_1}$ $q{t_2}$ $a$ ${b_0}$ ${b_1}$ ${b_{\mathrm{s}}}$ ${C_{xS}}$ $R$
初始值2.08.03.03.0338.5400.0320900.111011.7056
优化后值1.74.04.05.0464.1533.8397.1141.70.105910.6140
)], attaches=null, journal=Journal(id=1240685663704625163, delFlag=0, nameCn=船舶力学, nameEn=Journal of Ship Mechanics, nameHistory1=null, nameHistory2=null, issn=1007-7294, eissn=null, cn=32-1468/U, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=8VABIsgSw5I4HF/4goaIlA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Ship Mechanics, journalRemark=null, publicationField=null, createdTime=1773732113054, updatedTime=1784019383023, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=C, firstLetterEn=C, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=8VABIsgSw5I4HF/4goaIlA==, picEn=Ve/8A/FRYceS/z8o7GCJ5g==, jcr=null, cjcr=null, exts=[JournalExt(id=1283833602848834028, language=CN, name=船舶力学, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1784019383394, updatedTime=1784019383394, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=1, submissionEditorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=3, submissionReviewUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1283833602983051757, language=EN, name=Journal of Ship Mechanics, 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=1784019383426, updatedTime=1784019383426, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=1, submissionEditorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=3, submissionReviewUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1240685776644648972, websiteList=[Website(id=1240686760611607145, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1240685776644648972, 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/cblx/CN, language=CN, createTime=1773732374576, createBy=18614031015, updateTime=1773732394922, updateBy=18614031015, name=船舶力学-中文, tplId=1146099689490845704, title=船舶力学, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1240687166741869216, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=articleTextType, value=kx, createTime=1773732471404, updateTime=1773732471404, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166725091997, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=banner, value=null, createTime=1773732471400, updateTime=1773732471400, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166758646435, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=grayFlag, value=0, createTime=1773732471408, updateTime=1773732471408, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166716703388, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=logo, value=https://castjournals.cast.org.cn/joweb/cblx/CN/file/pic?fileId=/ZVjg3ao2vf5wYxw5Kn1Uw==, createTime=1773732471398, updateTime=1773732471398, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166771229349, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=minRunFlag, value=0, createTime=1773732471411, updateTime=1773732471411, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166737674911, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/cblx/CN/file/pic, createTime=1773732471403, updateTime=1773732471403, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166767035044, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=silenceFlag, value=0, createTime=1773732471410, updateTime=1773732471410, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166729286302, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1773732471401, updateTime=1773732471401, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166750257825, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=themeColor, value=null, createTime=1773732471406, updateTime=1773732471406, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166754452130, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=themeStyle, value=null, createTime=1773732471407, updateTime=1773732471407, creator=18614031015, updator=18614031015)]), Website(id=1240686760674521707, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1240685776644648972, 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/cblx/EN, language=EN, createTime=1773732374590, createBy=18614031015, updateTime=1773732413338, updateBy=18614031015, name=船舶力学-英文, tplId=1146101810881728533, title=Journal of Ship Mechanics, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1240687194269078438, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=articleTextType, value=kx, createTime=1773732477967, updateTime=1773732477967, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194248106915, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=banner, value=null, createTime=1773732477962, updateTime=1773732477962, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194285855657, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=grayFlag, value=0, createTime=1773732477971, updateTime=1773732477971, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194239718306, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=logo, value=https://castjournals.cast.org.cn/joweb/cblx/EN/file/pic?fileId=/ZVjg3ao2vf5wYxw5Kn1Uw==, createTime=1773732477960, updateTime=1773732477960, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194302632875, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=minRunFlag, value=0, createTime=1773732477975, updateTime=1773732477975, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194260689829, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/cblx/EN/file/pic, createTime=1773732477965, updateTime=1773732477965, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194294244266, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=silenceFlag, value=0, createTime=1773732477973, updateTime=1773732477973, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194256495524, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1773732477964, updateTime=1773732477964, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194277467047, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=themeColor, value=null, createTime=1773732477969, updateTime=1773732477969, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194281661352, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=themeStyle, value=null, createTime=1773732477970, updateTime=1773732477970, creator=18614031015, updator=18614031015)])], journalTitle=船舶力学, weixinUrl=null, journalUrl=https://cblx.cbpt.cnki.net/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Ship Mechanics, journalPhotoCn=8VABIsgSw5I4HF/4goaIlA==, journalPhotoEn=Ve/8A/FRYceS/z8o7GCJ5g==, 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, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/cblx/CN/10.3969/j.issn.1007-7294.2025.12.003, detailUrlEn=https://castjournals.cast.org.cn/joweb/cblx/EN/10.3969/j.issn.1007-7294.2025.12.003, pdfUrlCn=https://castjournals.cast.org.cn/joweb/cblx/CN/PDF/10.3969/j.issn.1007-7294.2025.12.003, pdfUrlEn=https://castjournals.cast.org.cn/joweb/cblx/EN/PDF/10.3969/j.issn.1007-7294.2025.12.003, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1765728000000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于数据驱动和机器学习的AUV外形优化设计
收藏切换
PDF下载
李云天 1, 2 , 陈伟政 1 , 海青 1 , 陈炜烨 1
船舶力学 | 流体力学 2025,29(12): 1848-1861
收起
收藏切换
船舶力学 |流体力学 2025 , 29 (12) : 1848 -1861
基于数据驱动和机器学习的AUV外形优化设计
全屏
[Author(id=1281326901427933629, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1281326901532791232, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901427933629, language=EN, stringName=Yun-tian LI, firstName=Yun-tian, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.China Ship Scientific Research Center, Wuxi 214082, China
2.CSSC Marine Power Zhenjiang Co., Ltd., Zhenjiang 212005, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326901612483009, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901427933629, language=CN, stringName=李云天, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.中国船舶科学研究中心,江苏 无锡 214082
2.中船动力镇江有限公司,江苏 镇江 212005, bio={"content":"

李云天(2000–),男,硕士研究生

"}, bioImg=null, bioContent=

李云天(2000–),男,硕士研究生

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)]), AuthorCompany(id=1281326901323076025, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=2., ext=[AuthorCompanyExt(id=1281326901331464634, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901323076025, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.CSSC Marine Power Zhenjiang Co., Ltd., Zhenjiang 212005, China), AuthorCompanyExt(id=1281326901339853243, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901323076025, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中船动力镇江有限公司,江苏 镇江 212005)])]), Author(id=1281326901717340611, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=chenwz.mail@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1281326901809615301, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901717340611, language=EN, stringName=Wei-zheng CHEN, firstName=Wei-zheng, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326901885112774, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901717340611, language=CN, stringName=陈伟政, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])]), Author(id=1281326901964804552, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=1281326902052884938, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901964804552, language=EN, stringName=Qing HAI, firstName=Qing, middleName=null, lastName=HAI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326902124188107, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326901964804552, language=CN, stringName=海青, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])]), Author(id=1281326902203879885, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, 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=1281326902275183055, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326902203879885, language=EN, stringName=Wei-ye CHEN, firstName=Wei-ye, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326902354874832, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, authorId=1281326902203879885, language=CN, stringName=陈炜烨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326901222412726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, xref=1., ext=[AuthorCompanyExt(id=1281326901230801335, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1281326901239189944, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326861871452194, companyId=1281326901222412726, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])])]
李云天1, 2, 陈伟政1 , 海青1, 陈炜烨1
作者信息
  • 1.中国船舶科学研究中心,江苏 无锡 214082
  • 2.中船动力镇江有限公司,江苏 镇江 212005
通讯作者:
陈伟政(1974−),男,博士,研究员,通讯作者,E-mail:
作者简介:

李云天(2000–),男,硕士研究生

Data-driven and machine learning-based optimization design for AUV shape
Yun-tian LI1, 2, Wei-zheng CHEN1 , Qing HAI1, Wei-ye CHEN1
Affiliations
  • 1.China Ship Scientific Research Center, Wuxi 214082, China
  • 2.CSSC Marine Power Zhenjiang Co., Ltd., Zhenjiang 212005, China
出版时间: 2025-12-15 doi: 10.3969/j.issn.1007-7294.2025.12.003
文章导航
收藏切换

自主水下航行器(AUV)的性能在很大程度上受其外形设计的影响。本文提出了一种AUV外形优化方法,通过集成数据驱动与机器学习技术,重点关注头部与尾部线型以及鳍舵布局对AUV性能的影响。本研究开发了一个参数化的水动力分析流程,用于自动生成优化过程中所需的水动力参数。我们基于此流程,构建了一个数据驱动框架,用于实现AUV外形的多目标优化,目标是最小化阻力并最大化机动性。针对这两个优化目标,本研究分别采用多层感知器(MLP)神经网络和集成学习方法构建了代理模型,并将它们的性能与传统代理模型进行了对比。优化问题通过非支配排序遗传算法求解。对比分析初始与优化后的AUV外形表明,其水动力性能显著提升,从而证实了所提方法的可行性与有效性。

自主水下航行器  /  外形设计  /  CFD仿真  /  多目标优化  /  机器学习  /  参数化设计

The performance of Autonomous Underwater Vehicles (AUVs) is significantly influenced by their shape design. This study presents a novel AUV shape optimization method that integrates data-driven approaches and machine learning technique to focus on the impact of the head and tail profiles, and the configuration of the fins and rudders on AUV performance. A parameterized hydrodynamic analysis workflow was developed to automatically generate the hydrodynamic parameters required during the optimization process. Based on this workflow, a data-driven framework was constructed for multi-objective optimization of AUV shapes, with the aims of minimizing drag and maximizing maneuverability. Surrogate models for the two optimization objectives above were built using Multilayer Perceptron (MLP) neural networks and ensemble learning methods respectively, and their performances were compared with traditional surrogate models. The optimization problem was solved using the Non-dominated Sorting Genetic Algorithm II. Comparative analysis of the initial and optimized AUV shapes demonstrates significant improvements in hydrodynamic performance, confirming the feasibility and effectiveness of the proposed method.

Autonomous Underwater Vehicle  /  shape design  /  CFD simulation  /  multi-objective optimization  /  machine learning  /  parametric design
李云天, 陈伟政, 海青, 陈炜烨. 基于数据驱动和机器学习的AUV外形优化设计. 船舶力学, 2025 , 29 (12) : 1848 -1861 . DOI: 10.3969/j.issn.1007-7294.2025.12.003
Yun-tian LI, Wei-zheng CHEN, Qing HAI, Wei-ye CHEN. Data-driven and machine learning-based optimization design for AUV shape[J]. Journal of Ship Mechanics, 2025 , 29 (12) : 1848 -1861 . DOI: 10.3969/j.issn.1007-7294.2025.12.003
自主水下航行器(AUV)是一种智能设备,主要由潜艇或水面舰船支持,能够长期自主进行水下航行和回收。它可搭载各种传感器、专用设备或武器模块以执行特定任务。AUV的自主性、灵活性和多功能性使其能够替代人类在恶劣环境中执行繁琐和危险的任务。它具有优越的机动性、卓越的适应性和生存能力,能够最大限度地降低人员出现伤亡的风险,其制造和维护同时具有经济、高效的特点[1]
续航能力和机动性对于AUV成功执行复杂任务至关重要,这些能力受到诸如阻力、设备负载、电池数量以及鳍舵的形状和布置等因素的显著影响。AUV的外形以及鳍舵的形状和布置不仅影响AUV的阻力,还影响其内部空间布局,在决定续航能力和机动性方面起着关键作用。然而,具有较低阻力的AUV外形并不一定具有最佳的机动性。因此,在AUV外形的优化过程中,必须采用多目标设计优化算法,以全面考虑这两个可能存在冲突的优化目标[2]
AUV外形以及鳍舵布置的优化涉及多个因素,研究人员从不同角度开展了相关研究。Sener和Aksu[3]全面分析了头部形状对AUV船体阻力性能和流动特性的影响,随后确定了在最小化阻力的同时最大化内部体积的最佳外形。Chen等[4]基于计算流体动力学(CFD)方法和遗传算法,提出了一种适用于拖曳式和自航式AUV的新型外形参数多目标优化方法。Chen等[5]研究了基于代理模型和因果图的翼身融合水下滑翔机外形的设计方法。Abbasi等[6]提出了一种基于混合多项式的优化方法,解决了水下滑翔机优化中的参数不确定性问题。Hou和Liang[7]提出了一种新的不确定性外形优化设计方法,用于受速度扰动影响的轴对称体。他们将该方法应用于AUV船体外形优化,并提供了使用不同不确定性优化方法和参数的示例。传统方法往往针对固定需求进行优化,需求变化时需重启设计流程,导致周期延长、资源浪费。相比之下,本文提出的数据驱动优化方法依托参数化分析与自动化代理建模,具备更高的灵活性与适应性。在目标或约束变化时,仅需调整输入参数即可快速生成新方案,无需重构流程,显著提升了设计效率。
在优化AUV外形以及鳍舵的形状和布置过程中,需要进行反复的CFD模拟以获得相关的优化目标值。因此,许多研究者使用代理模型来预测设计样本点的目标值,从而节省计算资源并提高优化效率。代理模型在航空、汽车和海洋工程等领域的优化设计中得到广泛应用。Lu等[8]使用最优拉丁超立方抽样和径向基函数构建了一个近似CFD的代理模型。Wen等[9]使用BP神经网络训练和拟合样本点以创建代理模型。Luo等[10]引入了一个包含置信区域和下置信界的动态代理模型,用于优化AUV艇体外形、阻力和能源效率。
近年来,人工智能技术的快速发展推动了计算机辅助设计的创新。在此背景下,许多研究者开始探索机器学习领域的神经网络方法,将其应用于AUV外部构型的优化。这种方法旨在提高性能的同时,缩短开发周期。Liu等[11]提出了一种结合计算流体动力学(CFD)和机器学习的数据驱动方法,用于有效优化AUV在不同速度、攻角和体积下的水动力外形。Xu等[12]将轮廓线的灰度图像作为输入特征,并将CFD模拟得到的阻力系数作为输出标签构建了数据集,通过构建卷积神经网络(CNNs)预测阻力系数,并结合体积约束确定了在给定体积下最小化阻力系数的旋转体外形。
尽管AUV构型设计已有广泛研究,但多数工作仍集中于局部变形,且依赖传统的数据采集与模型训练流程,存在样本生成耗时、缺乏自动化等问题,难以适应设计需求的变化。为此,本文提出一种集成化的数据驱动AUV外形优化方法,融合参数化建模、自动网格生成、CFD仿真、代理建模与多目标优化等模块。在AUV头尾线型及舵鳍布置方面,针对阻力系数与最小回转半径这两个优化目标,分别构建了基于多层感知器(MLP)和集成学习的代理模型。最终,采用调优的非支配排序遗传算法(NSGA–Ⅱ)求解多目标优化问题,该算法在确保全局搜索能力的同时能保持解的多样性,从而显著提升了外形优化的效率与最终性能。
数据驱动的AUV外形的多目标优化依赖于高质量的CFD原始数据。由于优化需要反复迭代,因此构建参数化的CFD流程至关重要,这可显著提升效率,而该流程涵盖了参数化建模、自动预处理、网格生成与CFD计算等关键环节。
为此,本文开发了一套集成了UG 12.0、SCDM 2022R2与FLUENT 2022R2三者的参数化水动力分析流程(见图1),通过CAD/CAE文件接口实现三者的协同工作。该流程可根据设计变量进行多轮CFD计算,输出各外形对应的水动力参数,用于优化分析。
鱼雷形AUV的几何构型主要包括两个核心部分:主体几何和鳍舵几何。主体部分进一步分为四个关键部分:头部、平行中段、尾部和尾锥段。为了显著提高AUV的机动性,并简化建模及后续分析过程,本研究采用了全动舵设计,取代了传统的鳍舵结构。这种简化方法不仅增加了有效舵面积,还提高了操作灵活性[13]图2展示了AUV的两个主要视图,并提供了全动舵的局部特写。
图2中,$L$表示AUV的总长度,$D$表示头部前端面直径,${L_{\text{H}}}$表示头部曲线段轴向长度,${L_{\text{C}}}$代表平行中段长度,${L_{{\text{TC}}}}$是尾部曲线段截取后剩余长度,${L_{{\text{TD}}}}$表示尾部整体曲线段被截取部分长度,${L_{{\text{TE}}}}$是尾锥段轴向长度,$ {D_{\text{E}}} $表示尾端面直径,$a$表示鳍舵总展长,${t_{\text{r}}}$表示舵的厚度,${b_0}$是根弦长,${b_1}$是梢弦长,${b_{\text{s}}}$表示舵后缘距尾端面距离,$\alpha $是尾锥半角,$\alpha = {0^ \circ }$
AUV主体线型的头部和尾部采用了格兰韦尔曲线[14]。具体来说,头部采用双参数立方多项式平头线型,尾部采用格兰韦尔尖尾线型。头部和尾部线型数学表达式如下
$\begin{aligned}&\qquad {y}_{\text{h}}^{3}=\dfrac{1}{q{h}_{1}}6x{(}{x-1})^4+q{h}_{2}\dfrac{1}{2}{x}^{2}{(x-1)}^{3}+1-{(x-1)}^{4}(4x+1)\\&{y}_{\text{t}}^{2}=-q{t}_{2}{x}^{2}{(x-1)}^{3}-q{t}_{1}{x}^{3}{(x-1)}^{2}+{x}^{3}\left(6{x}^{2}-15x+10\right),\quad (0\leqslant x \leqslant 1)\end{aligned} $
式中:$ q{h_1} $$ q{h_2} $是头部线型可调参数,$ q{t_1} $$ q{t_2} $是尾部线型可调参数。此外,$0 \leqslant {y_{\mathrm{h}}} \leqslant 1$$ 0 \leqslant {y_{\mathrm{t}}} \leqslant 1 $
在三维CAD软件(如UG)中实现参数化建模时,首要任务是将相关参数转换为具有实际意义的形式,这一过程涉及将无量纲的格兰韦尔数学线型表达式转化为有量纲的物理线型表达式。头部和尾部截面的物理线型表达式如下
$ \begin{gathered} \qquad \qquad \qquad \qquad {Y_{\text{h}}} = \frac{D}{2}{y_{\text{h}}}\left( {\frac{X}{{{L_{\text{H}}}}}} \right),\;\;\; \left( {0 \leqslant X \leqslant {L_{\text{H}}}} \right) \\ {Y_{\text{t}}} = \frac{D}{2}{y_{\text{t}}}\left( {\frac{{{L_{\text{H}}} + {L_{\text{C}}} + {L_{{\text{TC}}}} + {L_{{\text{TD}}}} - X}}{{{L_{{\text{TC}}}} + {L_{{\text{TD}}}}}}} \right),\;\;\; \left( {{L_{\text{H}}} + {L_{\text{C}}} \leqslant X \leqslant {L_{\text{H}}} + {L_{\text{C}}} + {L_{{\text{TC}}}}} \right) \\ \end{gathered} $
AUV的外形受到多种复杂因素的影响。本文选取了若干关键参数(如表1所定义),作为AUV参数化建模的核心设计变量。使用UG 12.0软件处理所有形状参数文件,并通过自行开发的程序生成了AUV模型。随后,生成的模型被导出为prt格式文件,以适应后续的预处理步骤。
根据AUV参数化模型的长度L和最大直径D调整相应计算域的尺寸,为此本文编写了Python脚本以自动化模型前处理过程。SpaceClaim软件运行这些包含模型预处理操作的脚本,并将预处理后的模型输出为scdm文件。预处理步骤主要涉及AUV模型的几何处理和相关计算域的设置。
计算域设计为圆柱形,AUV模型位于其中心。域的总长度设为$ 5.0 \; L $,宽度和高度均为$5.0D$(其中$ L $$ D $分别代表AUV的长度和最大直径)。为全面观察AUV的尾流特性,AUV后方的计算域长度设置为$ 3.0 \; L $,前方长度设为$ 1.0 \; L $。计算域前端的入口速度设定为AUV的设计速度(23.13 m/s),计算域末端的出口压力设为0 Pa。计算域的其他表面均定义为壁面边界条件。图3展示了整个计算域的设置。
Fluent软件通过运行编写的脚本自动导入scdm文件,基于脚本中指定的参数生成网格,并随后导出网格文件以用于后续的CFD仿真。脚本文件中的参数精确控制整个网格生成过程。
鉴于AUV复杂的几何结构,在计算域内采用了多面体非结构化网格。这种网格类型具有卓越的灵活性和适应性,能有效捕捉复杂的几何细节,同时显著减少网格单元数量并提高计算精度。考虑到AUV周围区域对CFD仿真结果的关键影响,应用了“体加密”(Body of Influence,BOI)方法对AUV附近区域进行局部网格加密。此外,采用棱柱层网格生成精细的边界层单元。图4展示了AUV周围自动生成的非结构化网格分布。
计算参数通过Fluent的日志脚本(Jou文件)进行修改,随后由FLUENT 2022R2执行计算。通过Jou文件导入自动生成的网格文件以进行数值模拟。模拟采用雷诺平均纳维–斯托克斯(RANS)方程和$ {{{{k}}- \omega }} $剪切应力输运(SST)湍流模型。流体介质为水,控制方程用于求解不可压缩粘性流动。应用体积流体(VOF)方法进行离散化,使用SIMPLEC进行压力修正,压力离散化采用二阶格式,动量、湍流动能和湍流耗散率采用二阶迎风格式。所有松弛因子保持默认值。
在AUV外形的多目标优化过程中,最小化阻力是关键优化目标之一。然而,由于设计约束和变量的范围广泛,整个样本空间内的阻力值变化范围很大。为确保数据拟合的准确性和可靠性,本研究引入无量纲阻力系数作为评估指标[15]。无量纲阻力系数$ {C_{xS}} $定义如下
$ {C_{xS}} = \frac{{{D_{{\text{rag}}}}}}{{0.5\rho {v^2}S}} $
式中:$ {D_{{\text{rag}}}} $表示AUV所受阻力,$ \rho $为流体密度,$ v $为巡航速度,$ S $为AUV的最大横截面积。
靠近AUV表面的区域,尤其是边界层,对CFD仿真结果有显著影响,因此,这一区域需要特别关注。第一网格层的无量纲壁面距离($ {{{y}}^ + } $值)是衡量网格质量的关键指标,可通过公式(4)计算和确定。
$ {{{y}}^ + } = {{y}}{{{u}}_\tau }/\upsilon $
式中:$ {{{u}}_\tau } $为壁面摩擦速度,$ \upsilon $为运动粘性系数。
为了在保证计算精度的前提下降低网格数量,本研究采用壁面函数方法进行湍流建模,目标将$ {{{y}}^ + } $值控制在推荐范围内(30~300)。图5展示了总单元数为301万时模型表面的$ {{{y}}^ + } $值分布情况。可以观察到,模型大部分区域的$ {{{y}}^ + } $值分布集中于40至60之间,基本符合壁面函数的要求。然而,在AUV头部及尾部舵面附近,由于流动加速与几何突变,$ {{{y}}^ + } $值分别出现了偏高(>90)与偏低(<10)现象,后续研究将考虑在这些区域进一步加密网格或采用不同壁面处理策略。
在水动力分析中,应用CFD技术需要评估其精度和有效性。为确保网格尺寸无关性,本研究针对AUV在$ {Re} = 6.4 \times {10^7} $$ v = 23.13{\text{ m/s}} $$ \rho = 998.2{\text{ kg/}}{{\text{m}}^{\text{3}}} $和攻角$ \alpha = {0^ \circ } $的条件下,采用五种不同的网格尺寸进行了对比,从而验证计算结果的可靠性。这五种网格的单元数量分别为255万个单元、301万个单元、351万个单元、401万个单元和511万个单元。
表2比较了不同网格数量下CFD模拟获得的$ {{{y}}^ + } $值和阻力系数值$ {C_{xS}} $。通过对比这些网格下得到的阻力系数结果,发现最粗糙和最精细网格之间的$ {C_{xS}} $值差异仅为3.32%,表明随着网格数量的增加,仿真结果趋于收敛。为节约计算资源,提高优化效率,后续优化分析统一选用了301万个单元的中等数量网格。
尽管其他水动力作用力和力矩对AUV的运动有显著影响,但其CFD仿真耗时较长,且本文主要关注AUV的操纵性。因此,本研究中采用了经验公式来计算相关的水动力系数[16]
AUV的机动性是指其在导航过程中改变运动状态的能力。衡量AUV机动性的一个简单参数是其在水平面上作定常回旋运动时的回转半径。水平面上定常回旋运动的最小回转半径可以使用下式计算[17]
$ \frac{L}{R} = \frac{{\left( {C_z^\beta + {C_{xS}}} \right)m_y^{{\delta _v}} + C_z^{{\delta _v}}m_y^\beta }}{{\left( {C_z^\beta + {C_{xS}}} \right)m_y^{{\omega _y}} - \left( {{\mu _S} - C_z^{{\omega _y}}} \right)m_y^\beta }}{\delta _{v\max }} $
式中:$R$代表最小回转半径,${\mu _s}$表示AUV的相对密度,${\mu _s} = 2 \; m/(\rho SL)$$m$是AUV的质量,$m = 220{\text{ kg}}$$ C_z^{{\omega _y}} $表示侧向力系数关于角速度的旋转导数,具体计算公式如下
$ C_z^{{\omega _y}} = C_z^\beta \frac{{{L_{vf}}}}{{L(1 - w)}} $
式中:$ {L_{vf}} $表示垂直鳍的升力作用点到浮心的距离,$ w $表示伴流系数,$ m_y^{{\omega _y}} $表示偏航力矩系数关于角速度的旋转导数,具体计算公式如下
$ m_y^{{\omega _y}} = m_y^\beta \frac{{{L_{vf}}}}{{L(1 - w)}} $
式中:$ m_y^\beta $代表AUV偏航力矩对侧滑角$\beta $的位置导数,且$ m_y^\beta = m_z^\alpha $$ m_z^\alpha $表示AUV俯仰力矩对攻角的位置导数,具体计算公式如下
$ m_z^\alpha = m_{zh}^\alpha + m_{zhf}^\alpha $
式中:$ m_{zh}^\alpha $代表AUV艇体俯仰力矩对攻角的位置导数,具体计算公式如下
$ m_{zh}^\alpha = 2(0.62 + 0.13(L/D))\phi $
式中:$ \phi $表示艇体丰满度系数,$\phi = 4 \; V/\left( {\text{π} {D^2}L} \right)$$V$为AUV体积。式(8)中的$ m_{zhf}^\alpha $表示水平鳍俯仰力矩对攻角的位置导数,具体计算公式如下
$ m_{zhf}^\alpha = - C_{yhf}^\alpha \frac{{{L_{hf}}}}{L} $
式中:$ C_{yhf}^\alpha $表示水平鳍对攻角的位置导数,$ {L_{hf}} $表示水平鳍的升力作用点到浮心的距离。$ {\delta _{v\max }} $表示最大垂直舵角,在本文中取${10^ \circ }$$ C_z^\beta $表示侧力系数对侧滑角的位置导数,由对称性可知,$ C_z^\beta = - C_y^\alpha $$ C_y^\alpha $为AUV升力系数的位置导数,其定义如下
$ C_y^\alpha = {\left. {\frac{{\partial {C_y}}}{{\partial \alpha }}} \right|_{\alpha = {\delta _h} = 0}} = C_{yh}^\alpha + C_{yhf}^\alpha $
式中:$ C_{yh}^\alpha $表示艇体升力对攻角的位置导数,$ C_{yhf}^\alpha $表示水平鳍舵对攻角的位置导数,两者的具体计算公式分别如公式(12)和公式(13)所示。
$ C_{yh}^\alpha = \frac{\text{π} }{2}\left( {\frac{1}{f}} \right)(0.25 + 0.013f)\left( {\frac{{{S_M}}}{S}} \right) $
式中:$f$表示AUV的细长比,$f = L/D$$S$是AUV的最大横截面积,${S_M}$是AUV的最大纵截面面积。
$ C_{yhf}^\alpha = 5.6 \times \frac{{5.5\lambda }}{{5.5\lambda + 5.6}}\frac{{2{A_{hf}}}}{S} $
式中:$S$是AUV的最大横截面积,$\lambda $表示单个水平鳍的展弦比,$\lambda = a_{hf}^2/{A_{hf}}$${a_{hf}}$是单个水平鳍的长度,单位是m,${A_{hf}}$是水平鳍的平面几何面积,单位是m2
由对称性可知,$m_y^{{\delta _v}} = m_z^{{\delta _h}}$$m_z^{{\delta _h}}$表示AUV对舵角的俯仰力矩,$ m_z^{{\delta _h}} = m_{zhf}^{{\delta _h}} $$ m_{zhf}^{{\delta _h}} $表示水平鳍舵俯仰力矩对舵角的位置导数,具体计算公式如下
$ m_{zhf}^{{\delta _h}} = - C_{yhf}^{{\delta _h}}\frac{{{L_{hf}}}}{L} $
式中:$ C_{yhf}^{{\delta _h}} $表示水平鳍舵对舵角的位置导数,$ C_{yhf}^{{\delta _h}} = nC_{yhf}^\alpha $$n$表示修正后的舵的效率系数,$n = 0.9$
由对称性可知,$C_z^{{\delta _v}} = C_y^{{\delta _h}}$$C_y^{{\delta _h}}$表示AUV对舵角的位置导数,具体计算公式如下
$ C_y^{{\delta _h}} = {\left. {\frac{{\partial {C_y}}}{{\partial {\delta _h}}}} \right|_{\alpha = {\delta _h} = 0}} = C_{yhf}^{{\delta _h}} $
数据驱动方法基于实验或仿真的数据构建模型,以量化输入与输出之间的关系,适用于传统方法难以处理的复杂系统[18]。该方法借助历史数据与算法支持,可用于建模、预测与决策。
在AUV设计中,数据驱动方法有助于优化设计参数、提升性能,并为多工况仿真和测试提供高效评估手段。为解决AUV外形多目标优化问题,本文构建了一个数据驱动优化框架(见图6),涵盖数据采集、代理模型构建、优化建模及性能分析等环节。
第一步,根据自主水下航行器的外形参数和鳍舵设计参数确定设计变量和输出响应。随后采用最优拉丁超立方抽样(OLHS)作为试验设计方法创建初始设计点${N_{{\text{ini }}}}$,在整个样本空间生成初始样本点[10]。再利用CFD仿真计算这些样本点的响应值,从而将这些样本点及其对应的响应值作为初始数据集${D_{{\text{all }}}}$输出。
第二步,将初始数据${D_{{\text{all }}}}$输入数据驱动模型以构建代理模型。将初始数据集划分为训练集和验证集。如果精度不足,则添加额外的样本点,并重复第一步和第二步,直至达到所需精度。在此步骤中,需评估不同代理模型(如Kriging法和响应面法(RSM)[19])的精度,以确定最适合的模型。
第三步,基于代理模型构建数据驱动优化模型。具体而言,在此步骤中,将目标函数、优化变量和设计约束(包括优化约束)输入优化算法,以求在最小阻力和最佳机动性之间获得平衡的最优外形。每个最优外形对应的设计变量值构成最优数据集。
第四步,将数据驱动优化模型获得的优化结果与最优样本点的实际CFD模拟结果进行比较。分析这些结果之间的误差。如果误差在可接受范围内,则认为该方法可行。否则,添加额外的样本点,并重复上述这四个步骤。
应用机器学习中的神经网络方法构建代理模型时,首先需要获取初始数据集。模型训练所需的数据通过CFD仿真计算获得。为简化数据收集过程,基于Isight软件和第1章所述的参数化水动力分析工作流程,本文开发了一个多学科集成平台。该平台能够自动计算并记录不同AUV外形的水动力相关参数。图7展示了此数据收集平台的架构。
首先,在UG软件中创建了AUV的参数化模型,定义了8个形状参数来控制AUV模型。这些形状参数作为后续优化的设计变量。表3列出了它们的设计空间范围。使用最优拉丁超立方抽样(OLHS)方法在此空间内生成初始采样点,记为${N_{{\text{ini}}}}$。OLHS方法通过在每个设计变量的范围内随机采样,同时确保样本点在整个空间内均匀分布,从而充分代表整个设计空间,因此无需预先设定固定的取值间隔。随后,将参数化模型导入SpaceClaim软件进行预处理,同时设置计算域。接着,将预处理后的模型导入Fluent Meshing软件进行网格划分。生成的网格随后输入Fluent进行数值计算,其中包含预设的AUV速度$v$。最后,将上述所有模块集成到Isight软件中,用于试验设计和初始数据集${D_{{\text{all }}}}$的收集。
由于CFD仿真计算的成本高昂,优化过程中对大量样本的需求构成了主要挑战。为此,本研究引入代理模型以替代高成本仿真。相较于传统方法(如克里金插值、响应面法(RSM)和径向基函数(RBF)),基于神经网络的代理模型展现出更强大的性能。在本研究中,我们构建了神经网络代理模型,用于高效预测AUV的水动力性能参数。
生物启发算法,特别是神经网络,在分类和回归问题中得到了广泛应用。这些网络模仿了生物神经元的信号感知和激活机制。在神经网络中,每个人工神经元处理加权输入,将它们求和后通过激活函数来确定其输出。这个感知过程可以用数学公式描述如下
$ y = f\left( {\sum\limits_{i = 1}^n {{w_i}} {x_i} + b} \right) $
式中:$ y $是神经元的输出,$ {x_i} $是输入信号,$ {w_i} $是相应的权重,$ b $是偏置项,$ f $是激活函数。神经网络通常由输入层、一个或多个隐藏层和输出层组成。神经网络系统通过在学习阶段修改权重和偏置来适应需求,这一过程使神经网络能够识别复杂的数据关系,从而高效地执行各种预测和分析任务。
基于上述理论基础,本研究采用双管齐下的方法构建AUV设计优化的代理模型。为预测阻力系数$ {C_{{xS}}} $,本文构建并训练了多层感知器(MLP)神经网络作为代理模型,针对最小回转半径$R$的预测,则采用了集成学习方法,融合了随机森林、支持向量回归和弹性网络模型。
神经网络通过模拟生物神经元的加权感知与非线性激活机制,可精准建模复杂输入输出关系。基于通用近似理论,本文采用多层感知器(MLP)神经网络来拟合CFD计算得到的阻力系数${C_{xS}}$,并使用Adam优化器最小化均方误差(MSE)。CFD数据的非线性特性使得MLP在捕捉${C_{xS}}$与几何参数间映射方面具备优势。先前在类似流体动力学预测任务中的研究,也已证明了MLP的优越性能[12]
MSE损失函数定义如下
$ {\text{MSE}} = \frac{1}{n}\sum\limits_{i = 1}^n {{{\left( {{y_i} - {{\hat y}_i}} \right)}^2}} $
式中:$ {y_i} $是真实值,$ {\hat y_i} $是预测值,$n$是样本数量。
与此不同,最小回转半径$R$系由MATLAB经验公式计算,更适合采用集成学习策略。本文结合随机森林(RF)、支持向量回归(SVR)与弹性网络模型,构建更稳健的预测框架。该集成策略融合了各模型的优势,降低过拟合风险,提升广泛参数区间内的泛化能力,并增强对参数间非线性关联的刻画能力。其中RF提供特征重要性排序,SVR具备良好的高维泛化能力,弹性网络则兼顾特征选择与多重共线性处理。
随机森林模型由100棵决策树组成,擅长捕捉特征间的非线性关系和交互。它提供了特征重要性排序的额外优势,揭示了不同设计参数对$R$的相对影响。支持向量回归模型使用径向基函数(RBF)核,特别擅长处理高维空间,在各种工程应用中表现出良好的泛化能力。弹性网络模型结合了L1和L2正则化,在特征选择和处理多重共线性之间取得平衡,有利于解决AUV设计参数间潜在的相互影响。
这些模型在数据集上独立训练,通过k折交叉验证调整超参数以确保最佳性能。$R$的最终预测通过个体模型预测的加权平均获得,即
$ {R_{{\text{predicted }}}} = {w_{{\text{RF}}}} \cdot {R_{{\text{RF}}}} + {w_{{\text{SVR}}}} \cdot {R_{{\text{SVR}}}} + {w_{{\text{EN}}}} \cdot {R_{{\text{EN}}}} $
式中:$ {w_{{\text{RF}}}} $$ {w_{{\text{SVR}}}} $$ {w_{{\text{EN}}}} $分别是分配给随机森林、SVR和弹性网络模型的权重,在本研究中权重均取$1/3$,这是因为验证结果显示上述各模型表现相近,且简单均值能够有效平滑各模型的偏差。
基于建立的代理模型,本文以AUV艇型参数以及鳍舵的外形参数作为设计变量,以阻力系数和最小回转半径为目标,开展多目标优化研究,得到阻力最小和机动性最优之间的Pareto前沿解。具体的设计变量取值范围如表3中所示。
为了确保优化过程的合理性及最终优化的AUV外形具有合理的体积,使用头部和尾部的丰满系数作为约束条件。头部丰满系数定义如下
$ {\varPsi }_{\text{H}}=\frac{4}{\text{π} {D}^{2}{L}_{\text{H}}}{\displaystyle {\int }_{0}^{{L}_{\text{H}}}\text{π} }{Y}_{h}^{2} \text{d}X $
式中:${L_{\text{H}}}$为AUV头部长度,$ {Y_{\mathrm{h}}} $为格兰韦尔头部线型曲线。尾部丰满系数定义如下
$ {\varPsi }_{\text{T}}=\frac{4}{\text{π} {D}^{2}{L}_{\text{T}}}{\displaystyle {\int }_{0}^{{L}_{\text{T}}}\text{π} }{Y}_{t}^{2} \text{d}X $
式中:${L_{\text{T}}}$为AUV头部长度,$ {Y_{\mathrm{t}}} $为格兰韦尔尾部线型曲线。
综上,优化问题可以描述为以下形式
$ \begin{array}{*{20}{l}} {{\text{优化目标参数}}}&{C_{xS}}&R \\ {{\text{设计变量}}}&{1.0 \leqslant q{h_1} \leqslant 10.0}&{0.0 \leqslant q{h_2} \leqslant 20} \\ {{\text{取值范围}}}& {3.0 \leqslant q{t_1} \leqslant 11.0}&{3.0 \leqslant q{t_2} \leqslant 11.0} \\ &{300.0 \leqslant a \leqslant 500.0}&{400.0 \leqslant {b_0} \leqslant 800.0} \\ & {200.0 \leqslant {b_1} \leqslant 600.0}&{90.0 \leqslant {b_{\mathrm{s}}} \leqslant 300.0} \\ {{\text{约束条件}}}&{{\Psi _{\text{H}}} \geqslant 0.85}&{{\Psi _{\text{T}}} \geqslant 0.55} \end{array} $
本研究所提出的优化问题为多目标优化问题,目标函数分别为最小化阻力$ C_{xS} $和最小回转半径$R$。在该问题中,阻力和回转半径被作为两个独立的目标进行优化,而非通过加权组合成单一目标。为求解此多目标问题,本文采用了非支配排序遗传算法(NSGA−Ⅱ),该算法能够同时优化两个目标,并生成一组非支配解。设置算法中的交叉概率为0.7、变异概率为0.4、种群规模为300、最大迭代次数为40次。
代理模型的性能通过测试数据集进行了严格评估,该数据集占总可用数据的20%。此外,本文比较了使用各种方法建立的代理模型的预测性能,包括MLP神经网络、Kriging法、RBF(径向基函数)模型和响应面法。决定系数$ {R^2} $被用作评估预测性能的标准。$ {R^2} $是一个统计量,表示回归模型与观测数据的拟合程度。它的范围从0到1,值越接近1表示拟合越好,$ {R^2} $定义如下
$ {R^2} = 1 - \dfrac{{\displaystyle\sum\limits_{i = 1}^n {{{\left( {{y_i} - {{\hat y}_i}} \right)}^2}} }}{{\displaystyle\sum\limits_{i = 1}^n {{{\left( {{y_i} - \bar y} \right)}^2}} }} $
式中:$ {y_i} $表示观测值,$ {\hat y_i} $表示模型预测值,$ \bar y $是观测值的平均值,$n$是观测数量。表4展示了不同代理模型预测阻力系数${C_{xS}}$值的性能。
表4的结果可以看出,MLP神经网络在捕捉设计变量与目标变量之间复杂非线性关系方面表现出卓越能力。与传统的代理建模技术相比,该模型显著提高了预测准确性。通过MLP模型获得的最优解在满足约束的同时,证明了该模型是一种用于CFD模拟代理建模的可行且高效的计算策略。
表5展示了不同机器学习模型在预测最小回转半径$ R $时的性能比较。从表中可以看出,弹性网络在均方误差(MSE)指标上取得了最低值(9.6141),而集成方法在决定系数$ {R^2} $指标上达到了最高值(0.7155)。这种现象表明,单一的弹性网络模型虽然在降低预测误差方面表现较好,但从结合了随机森林、支持向量回归(SVR)和弹性网络的集成方法预测结果分析,集成方法能够更好地解释数据变异并提高整体预测的稳定性。因此,对于最小回转半径的预测,本文采用了集成学习方法,以期在降低误差与提高模型泛化能力之间达到更优的平衡。
根据表5所示,不同机器学习模型对最小回转半径$ R $的预测效果较为有限,这主要可能源于在本研究的优化设计流程中,对鳍和舵的约束条件相对较少,导致其对机动性影响的数据分布较为宽泛且噪声较大。此外,最小回转半径$ R $的计算涉及复杂的流体力学模拟和经验公式,这也可能引入一定的计算误差。未来的工作将重点加强对鳍舵参数的约束,并考虑采用更多的数据样本和更精细的数据处理方法来提升$ R $的预测性能。
本研究得到了预测值的空间分布图,以便将代理模型在整个设计空间中的性能可视化。图8展示了通过最优拉丁超立方采样生成的新样本点在设计空间中的分布情况,以及相应的代理模型预测值的空间分布。该图主要用于验证采样方法在整个设计空间内的均匀性,确保样本点能够充分覆盖所有可能的设计变量组合。此外,图8中预测值的连续变化趋势表明,代理模型在全局范围内具有较好的预测一致性,从而间接验证了模型的合理性。虽然图8未直接揭示各单个设计参数对阻力系数和最小回转半径的具体影响,但其整体分布情况为后续进行敏感性分析和多目标优化研究提供了有力的数据支持。
本文绘制了实际值与预测值的对比图以及误差直方图(图9图10),以进一步评估代理模型的误差和性能。图9展示了${C_{xS}}$$ R $的实际值与预测值的散点图。对于${C_{xS}}$,点紧密地靠近对角线,表明MLP模型具有较高的预测准确性。$ R $的预测结果来自集成学习方法,同样显示出良好的对齐趋势,由此证明结合了多个模型后的集成方法在捕捉最小回转半径复杂关系方面具备的有效性。
图10中的误差直方图进一步证实了这些观察结果。MLP 模型的误差分布较窄且呈峰值,大多数误差都落在实际值的较小范围内。集成模型的误差分布虽然稍宽一些,但仍显示出明显的集中趋势,反映了本文所用模型在处理、预测固有复杂性方面的能力。
敏感性分析被广泛用于估计输入参数对输出参数的影响。在本研究中,使用Sobol方法分析了每个设计参数对AUV阻力系数和最小回转半径的影响,结果如图11所示。在设计参数中,$a$对阻力系数${C_{xS}}$和最小回转半径$ R $有显著影响,而${b_0}$$q{t_2}$的影响较小。显然,相比头部和尾部可调参数,阻力和机动性对舵的展弦比$a$、弦长${b_0}$、弦长${b_1}$和纵向位置${b_{\mathrm{s}}}$的变化更为敏感。这表明舵的设计和布置对AUV的阻力和机动性有显著影响。
非支配排序遗传算法成功生成了具有明显Pareto前沿的非支配解集。图12显示了基于NSGA–Ⅱ得到的Pareto前沿(蓝色圆点)以及其余可行解(灰色圆点)。为更直观地了解优化过程,本文在图中同时保留了被支配解和非支配解。在非支配解集中,本研究通过比较所有满足约束条件的解中阻力系数值${C_{xS}}$,确定了最佳解。具体而言,由于实际工程中降低阻力对 AUV 性能至关重要,本文在两个目标中赋予${C_{xS}}$较高的权重,最终选择了${C_{xS}}$值最小的解。该最优解在图中以红色星形标记。这个前沿展示了${C_{xS}}$$ R $之间的权衡关系。如图12所示,${C_{xS}}$$ R $之间存在负相关关系:随着${C_{xS}}$的减少,$ R $往往会增加,反映了两个目标间的冲突特性。Pareto前沿的形状呈现出显著的非线性特征,表明两个目标之间的权衡在解空间的不同区域有所变化。特别是在${C_{xS}}$相对较低的区域,$ R $的变化更为明显,这暗示在某些设计点上可能存在显著的性能改进潜力。此外,需要指出的是,在当前约束条件下,可行解的数量有限,因此Pareto前沿上点数较少。今后工作中将进一步放宽或调整设计变量及约束条件,以探索更广泛的可行解空间并获得更丰富的Pareto解集。
表6显示了AUV的优化结果及其对应的初始设计。对AUV优化设计结果的分析表明,优化后的AUV阻力系数比初始设计降低了4.59%,回转半径减小了9.32%。经计算,优化后的丰满度系数满足约束要求。初始和优化后的3D模型如图13所示。根据模型图可以看出,优化后的艇体更具流线型,舵的形状、尺寸和位置也得到了优化。
尽管本研究通过网格无关性分析验证了数值计算的一致性,但由于缺乏针对当前AUV外形的实验数据,尚未进行直接验证。因此,本文侧重于各设计方案之间的相对性能趋势分析。后续研究将结合实验或标准算例开展进一步验证,以提高结果的可信度。
本文提出了一种数据驱动的鱼雷形AUV外形优化方法,利用CFD仿真生成的数据训练代理模型,提升了设计空间探索的效率与深度。主要贡献如下:
(1) 构建了自动化的AUV水动力分析流程,集成了参数化建模、自动网格生成及CFD仿真,显著提高了建模效率与可重复性。
(2) 提出了一个完整的数据驱动优化框架,明确了设计变量和优化目标(阻力系数与回转半径),对头尾线型与舵鳍布置参数进行多目标优化,最终获得Pareto解集。该框架涵盖了从数据采集到仿真验证的全过程。
(3) 针对不同目标分别构建了高精度代理模型:阻力系数采用多层感知器(MLP)预测,而最小回转半径则采用融合了随机森林、支持向量回归和弹性网络的集成学习方法进行预测。最终,为每个目标选取了精度最高的代理模型用于优化。
(4) 在先对代理模型进行了残差分析与误差评估,验证了其预测准确性与稳定性后,继而通过独立的CFD仿真验证优化结果,并与模型预测值进行对比,进一步证实了该方法的有效性与可行性。

参考文献 引证文献
排序方式:
1
宋保维, 潘 光, 张立川, 等. 自主水下航行器发展趋势及关键技术[J]. 中国舰船研究, 2022, 17(5): 27‒44.
Song B W, Pan G, Zhang L C, et al. Development trend and key technologies of autonomous underwater vehicles[J]. Chinese Journal of Ship Research, 2022, 17(5): 27‒44. (in Chinese)
2
Hu F, Huang Y, Xie Z, et al. Conceptual design of a long-range autonomous underwater vehicle based on multidisciplinary optimization framework[J]. Ocean Engineering, 2022, 248: 110684.
3
Sener M Z, Aksu E. The effects of head form on resistance performance and flow characteristics for a streamlined AUV hull design[J]. Ocean Engineering, 2022, 257: 111630.
4
Chen X, Yu L, Liu L Y, et al. Multi-objective shape optimization of autonomous underwater vehicle by coupling CFD simulation with genetic algorithm[J]. Ocean Engineering, 2023, 286: 115722.
5
Chen W, Dong H, Wang P, et al. Causal-relationship-assisted shape design optimization for blended-wing-body underwater gliders[J]. Engineering Optimization, 2023, 6.
6
Abbasi S, Zeinali M, Nejadabbasi P. Autonomous underwater vehicle hull geometry optimization using a multi-objective algorithm approach[J]. International Journal of Engineering, 2018, 31(9): 1593‒1601.
7
Hou Y H, Liang X. Uncertainty optimization design of an axially symmetric body using speed perturbation[J]. Journal of Marine Science and Technology, 2019, 24(2): 612‒619.
8
Lu Y, Yuan J, Si Q, et al. Study on the optimal design of a shark-like shape AUV based on the CFD method[J]. Journal of Marine Science and Engineering, 2023, 11(10): 1869.
9
Wen Q, Feng R, An X, et al. Optimal design of an autonomous underwater helicopter’s shape based on combinatorial optimization strategy[J]. Ocean Engineering, 2022, 266: 113015.
10
Luo W, Guo X, Dai J, et al. Hull optimization of an underwater vehicle based on dynamic surrogate model[J]. Ocean Engineering, 2021, 230: 109050.
11
Liu M, Song Y, Yang S, et al. Rapid data-driven individualised shape design of AUVs based on CFD and machine learning[J]. Ships and Offshore Structures, 2024, 19(10112): 2050−2064.
12
Xu Y, Liu P, Wang L, et al. Main body shape optimization of non-body-of-revolution underwater vehicles by using CNN and genetic algorithm[J]. Ocean Engineering, 2024, 295: 116938.
13
Chen X, Wang P, Zhang D, et al. Gradient-based multidisciplinary design optimization of an autonomous underwater vehicle[J]. Applied Ocean Research, 2018, 80: 101‒111.
14
Granville P S. Geometrical characteristics of streamlined shapes[R]. 1969.
15
Saha S, Gan Z, Cheng L, et al. Hierarchical deep learning neural network (HiDeNN): An artificial intelligence (AI) framework for computational science and engineering[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 373: 113452.
16
潘 光. 鱼雷力学 [M]. 西安: 陕西师范大学出版社, 2013: 120−150.
Pan G. Torpedo mechanics[M]. Xi’an: Shaanxi Normal University Press, 2013: 120−150. (in Chinese)
17
宋保维, 王 鹏. 鱼雷多学科设计优化理论与应用研究[M]. 西安: 西北工业大学出版社, 2016: 18−53.
Song B W, Wang P. Research on the theory and application of torpedo multidisciplinary design optimization[M]. Xi’an: Northwestern Polytechnical University Press, 2016: 18−53. (in Chinese)
18
Wang J, Li Y, Gao R X, et al. Hybrid physics-based and data-driven models for smart manufacturing: Modelling, simulation, and explainability[J]. Journal of Manufacturing Systems, 2022, 63: 381‒391.
19
Yang M, Wang Y, Chen Y, et al. Data-driven optimization design of a novel pressure hull for AUV[J]. Ocean Engineering, 2022, 257: 111562.
2025年第29卷第12期
PDF下载
119
63
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.1007-7294.2025.12.003
  • 接收时间:2025-07-04
  • 首发时间:2026-07-07
  • 出版时间:2025-12-15
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-07-04
基金
作者信息
    1.中国船舶科学研究中心,江苏 无锡 214082
    2.中船动力镇江有限公司,江苏 镇江 212005

通讯作者:

陈伟政(1974−),男,博士,研究员,通讯作者,E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/cblx/CN/10.3969/j.issn.1007-7294.2025.12.003
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