Article(id=1281326802639499379, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1281326672238580175, articleNumber=null, orderNo=null, doi=10.3969/j.issn.1007-7294.2026.01.015, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1758297600000, receivedDateStr=2025-09-20, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1783421715650, onlineDateStr=2026-07-07, pubDate=1768406400000, pubDateStr=2026-01-15, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1783421715650, onlineIssueDateStr=2026-07-07, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1783421715650, creator=13701087609, updateTime=1783421715650, updator=13701087609, issue=Issue{id=1281326672238580175, tenantId=1146029695717560320, journalId=1240685776644648972, year='2026', volume='30', issue='1', pageStart='1', pageEnd='176', issueExtLink='null', onlineDate='null', pubDate='1768406400000', pubDateStr='2026-01-15', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1783421684559, creator='13701087609', updateTime=1783422118948, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1281328494261026863, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1281326672238580175, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1281328494261026864, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1281326672238580175, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=159, endPage=167, ext={EN=ArticleExt(id=1281326802899546228, articleId=1281326802639499379, tenantId=1146029695717560320, journalId=1240685776644648972, language=EN, title=Line spectrum feature extraction based on machine learning, columnId=1241023038381158513, journalTitle=Journal of Ship Mechanics, columnName=Hydro/Structural Acoustics, runingTitle=null, highlight=null, articleAbstract=

This paper is to study the principle of line spectrum feature extraction. In view of the deficiency of manual line spectrum feature extraction method, a line spectrum feature extraction method based on machine learning was proposed. The Encoder-Decoder based on convolution neural network was built, and the attention mechanism was introduced between the convolution and pooling layers, so that the important features of the input data could occupy a higher weight to enhance the accuracy of feature extraction. The model was compared with U-Net model and TPSW algorithm in the case of low signal-to-noise ratio, and tested on the actual data. The experimental results show that the improved model achieves a line positioning accuracy of 0.823 at a signal-to-noise ratio of 5 dB. This performance is better than that of the U-Net model and TPSW algorithm with in the 0~5 dB range. Thus the model effectively extracts line spectrum information and improves the accuracy of underwater target detection.

, authors=Jun-jie SHI1, 2, 3, 4, Ling-shuang XIONG4, Da-jun SUN1, 2, 3, authorsList=Jun-jie SHI, Ling-shuang XIONG, Da-jun SUN, authorCompany=null, correspAuthors=Ling-shuang XIONG, authorNote=null, correspAuthorsNote=null, copyrightStatement=Copyright ©2026 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=1281326811309125815, articleId=1281326802639499379, tenantId=1146029695717560320, journalId=1240685776644648972, language=CN, title=基于机器学习的线谱特征提取方法, columnId=1241023038515376243, journalTitle=船舶力学, columnName=流体与结构声学, runingTitle=null, highlight=null, articleAbstract=

针对人工提取线谱特征方法的不足,本文对线谱特征提取原理进行研究,提出一种基于机器学习的线谱特征提取方法。搭建基于卷积神经网络的编码器–解码器,并在卷积和池化层间引入注意力机制,使输入数据的重要特征占据更高的权重,从而提高特征提取的准确度。将模型与U–Net模型和TPSW算法在低信噪比情况下进行比较,并在实际数据上进行测试。实验结果表明,在谱级信噪比为5 dB时,改进模型线定位精度可达到0.823,在0~5 dB时均优于U–Net模型和TPSW算法,达到了提取线谱有效信息,提高水下目标检测准确率的目的。

, authors=师俊杰1, 2, 3, 4, 熊凌霜4, 孙大军1, 2, 3, authorsList=师俊杰, 熊凌霜, 孙大军, authorCompany=null, correspAuthors=熊凌霜, authorNote=

师俊杰(1980–),男,博士,副教授

, correspAuthorsNote=
熊凌霜(2000–),女,硕士,通讯作者,E-mail:
, copyrightStatement=版权所有©《船舶力学》编辑部2026, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=PEZnKe1aWur1xLwa9/QXLg==, magXml=x3Rzl+Zh+yACsCcAlsz7zA==, pdfUrl=null, pdf=ByRx1s79nTnI+e8PL9Xfmg==, pdfFileSize=2873193, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=qjTrA+lOO7P+ER4RqmfslQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=ePEL3kiBRD/gEHVa3J9sdQ==, mapNumber=null, fund=null)}, authors=[Author(id=1281326811841802437, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=1281326811950854346, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326811841802437, language=EN, stringName=Jun-jie SHI, firstName=Jun-jie, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, 4, address=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China
2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China
3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China
4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326812022157515, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326811841802437, language=CN, stringName=师俊杰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, 4, address=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001
2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001
3.哈尔滨工程大学 水声工程学院,哈尔滨 150001
4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000, bio={"content":"

师俊杰(1980–),男,博士,副教授

"}, bioImg=null, bioContent=

师俊杰(1980–),男,博士,副教授

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326811518841016, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=1., ext=[AuthorCompanyExt(id=1281326811523035321, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China), AuthorCompanyExt(id=1281326811531423930, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001)]), AuthorCompany(id=1281326811602727099, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=2., ext=[AuthorCompanyExt(id=1281326811611115708, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China), AuthorCompanyExt(id=1281326811615310013, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001)]), AuthorCompany(id=1281326811686613182, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=3., ext=[AuthorCompanyExt(id=1281326811695001791, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China), AuthorCompanyExt(id=1281326811699196096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.哈尔滨工程大学 水声工程学院,哈尔滨 150001)]), AuthorCompany(id=1281326811753722049, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=4., ext=[AuthorCompanyExt(id=1281326811762110658, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China), AuthorCompanyExt(id=1281326811770499267, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000)])]), Author(id=1281326812080877773, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=xls0016@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1281326812160569551, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812080877773, language=EN, stringName=Ling-shuang XIONG, firstName=Ling-shuang, middleName=null, lastName=XIONG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326812219289808, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812080877773, language=CN, stringName=熊凌霜, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326811753722049, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=4., ext=[AuthorCompanyExt(id=1281326811762110658, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China), AuthorCompanyExt(id=1281326811770499267, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000)])]), Author(id=1281326812282204370, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=1281326812382867670, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812282204370, language=EN, stringName=Da-jun SUN, firstName=Da-jun, middleName=null, lastName=SUN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China
2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China
3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326812458365143, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812282204370, language=CN, stringName=孙大军, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001
2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001
3.哈尔滨工程大学 水声工程学院,哈尔滨 150001, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326811518841016, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=1., ext=[AuthorCompanyExt(id=1281326811523035321, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China), AuthorCompanyExt(id=1281326811531423930, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001)]), AuthorCompany(id=1281326811602727099, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=2., ext=[AuthorCompanyExt(id=1281326811611115708, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China), AuthorCompanyExt(id=1281326811615310013, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001)]), AuthorCompany(id=1281326811686613182, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=3., ext=[AuthorCompanyExt(id=1281326811695001791, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China), AuthorCompanyExt(id=1281326811699196096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.哈尔滨工程大学 水声工程学院,哈尔滨 150001)])])], keywords=[Keyword(id=1281326812571611352, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, orderNo=1, keyword=line spectrum feature extraction), Keyword(id=1281326812651303129, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, orderNo=2, keyword=machine learning), Keyword(id=1281326812726800602, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, orderNo=3, keyword=Encoder-Decoder), Keyword(id=1281326812781326555, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, orderNo=4, keyword=underwater target detection), Keyword(id=1281326812848435420, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, orderNo=1, keyword=线谱特征提取), Keyword(id=1281326812915544285, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, orderNo=2, keyword=机器学习), Keyword(id=1281326812982653150, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, orderNo=3, keyword=编码–解码器), Keyword(id=1281326813053956319, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, orderNo=4, keyword=水下目标检测)], refs=[Reference(id=1281326818162618610, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=吴国清, 李 靖, 陈耀明, 等. 舰船噪声识别(I)──总体框架、线谱分析和提取[J]. 声学学报, 1998, 5: 394‒400., articleTitle=null, refAbstract=null), Reference(id=1281326818233921779, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=Wu G Q, Li J, Chen Y M, et al. Ship noise recognition (I)−Overall framework, line spectrum analysis and extraction[J]. Acta Acustica, 1998, 5: 394‒400. (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1281326818326196468, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=Zhao L, Li B. A novel method for extracting frequency line on lofargram based on feature function[C]//The 4th International Conference on Image, Video and Signal Processing, 2022., articleTitle=null, refAbstract=null), Reference(id=1281326818384916725, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=李 山, 王德俊, 王海斌. 一种基于线谱特征函数提取LOFAR图线谱的方法[J]. 声学技术, 2016, 35(4): 373‒377., articleTitle=null, refAbstract=null), Reference(id=1281326820003918070, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=Li S, Wang D J, Wang H B. A method for extracting LOFAR graph line spectrum based on line spectrum feature function[J]. Acoustic Technology, 2016, 35(4): 373‒377. (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1281326820066832631, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=Chen C H, Lee J D, Lin M. Classification of underwater signals using neural networks[J]. Journal of Applied Science and Engineering, 2000, 3(1): 31‒48., articleTitle=null, refAbstract=null), Reference(id=1281326820154913016, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=陈敬军, 陆佶人, 刘 淼. 基于人工智能的线谱检测技术[J]. 船舶工程, 2004, 3: 68‒71., articleTitle=null, refAbstract=null), Reference(id=1281326820226216185, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=Chen J J, Lu J R, Liu M. Line spectrum detection technology based on artificial intelligence[J]. Ship Engineering, 2004, 3: 68‒71. (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1281326820297519354, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=Jung D J, Park J, Lee S H, et al. Deep learning based tonal detection for passive sonar signals[C]//UTD, 2019., articleTitle=null, refAbstract=null), Reference(id=1281326820377211131, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen J, Han B, Ma X, et al. Underwater target recognition based on multi-decision LOFAR spectrum enhancement: A deep-learning approach[J]. Future Internet, 2021, 13: 265., articleTitle=null, refAbstract=null), Reference(id=1281326820456902908, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[J]. arXiv, 2017., articleTitle=null, refAbstract=null), Reference(id=1281326820515623165, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Paris Sebastien, Jauffret C, et al. Frequency line tracking using HMM-based schemes[J]. IEEE Transactions on Aerospace & Electronic Systems, 2003, 39(2): 439‒449., articleTitle=null, refAbstract=null), Reference(id=1281326820591120638, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=Lampert T A, O'Keefe S E M. On the detection of tracks in spectrogram images[J]. Pattern Recognit, 2013, 46: 1396‒1408., articleTitle=null, refAbstract=null), Reference(id=1281326820649840895, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Martino J C D, Tabbone S. An approach to detect lofar lines[J]. Pattern Recognition Letters, 1996, 17(1): 37‒46., articleTitle=null, refAbstract=null), Reference(id=1281326820708561152, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Pratt W K. Digital image processing[M]. Second Edition. New York: Wiley Interscience, 1991., articleTitle=null, refAbstract=null), Reference(id=1281326820775670017, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical imagesegmentation[J]. Springer International Publishing, 2015., articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1281326811518841016, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=1., ext=[AuthorCompanyExt(id=1281326811523035321, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China), AuthorCompanyExt(id=1281326811531423930, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001)]), AuthorCompany(id=1281326811602727099, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=2., ext=[AuthorCompanyExt(id=1281326811611115708, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China), AuthorCompanyExt(id=1281326811615310013, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001)]), AuthorCompany(id=1281326811686613182, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=3., ext=[AuthorCompanyExt(id=1281326811695001791, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China), AuthorCompanyExt(id=1281326811699196096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.哈尔滨工程大学 水声工程学院,哈尔滨 150001)]), AuthorCompany(id=1281326811753722049, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=4., ext=[AuthorCompanyExt(id=1281326811762110658, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China), AuthorCompanyExt(id=1281326811770499267, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000)])], figs=[ArticleFig(id=1281326813150425312, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.1, caption=Architecture of Encoder-Decoder, figureFileSmall=nNq5CQAFBj5JpiFgV03c9Q==, figureFileBig=qjTrA+lOO7P+ER4RqmfslQ==, tableContent=null), ArticleFig(id=1281326813225922785, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图1, caption=编码器–解码器网络架构, figureFileSmall=nNq5CQAFBj5JpiFgV03c9Q==, figureFileBig=qjTrA+lOO7P+ER4RqmfslQ==, tableContent=null), ArticleFig(id=1281326813439832290, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.2, caption=Schematic diagram of network, figureFileSmall=gRleZY2pL67mcp7Wl5Evgg==, figureFileBig=o5Vp6vQSIHBBWDZOa4KFZg==, tableContent=null), ArticleFig(id=1281326813502746851, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图2, caption=网络示意图, figureFileSmall=gRleZY2pL67mcp7Wl5Evgg==, figureFileBig=o5Vp6vQSIHBBWDZOa4KFZg==, tableContent=null), ArticleFig(id=1281326813611798756, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.3, caption=Diagram of the attention layer, figureFileSmall=TL9YK9+in1AKm12fon+TFw==, figureFileBig=kxGvUSw+/AD54rGpdJMGrA==, tableContent=null), ArticleFig(id=1281326813708267749, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图3, caption=注意力层示意图, figureFileSmall=TL9YK9+in1AKm12fon+TFw==, figureFileBig=kxGvUSw+/AD54rGpdJMGrA==, tableContent=null), ArticleFig(id=1281326813800542438, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.4, caption=Architecture of attention mechanism, figureFileSmall=wLy61ut5detOmedhaMYcNQ==, figureFileBig=K6sTRqjx0BWhWDP/h5QZSA==, tableContent=null), ArticleFig(id=1281326813863456999, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图4, caption=注意力架构, figureFileSmall=wLy61ut5detOmedhaMYcNQ==, figureFileBig=K6sTRqjx0BWhWDP/h5QZSA==, tableContent=null), ArticleFig(id=1281326815499235560, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.5, caption=Dataset, figureFileSmall=qq70NkUILUrZcCVcfSxXMg==, figureFileBig=favx5sKVQvT13/2Pp6lESA==, tableContent=null), ArticleFig(id=1281326815570538729, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图5, caption=数据集构成, figureFileSmall=qq70NkUILUrZcCVcfSxXMg==, figureFileBig=favx5sKVQvT13/2Pp6lESA==, tableContent=null), ArticleFig(id=1281326815646036202, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.6, caption=Line spectra of different shapes, figureFileSmall=aX9aJodKcwx5dHWmu2NL9A==, figureFileBig=Bjgza7bCaGIvfIsRzoQ2Og==, tableContent=null), ArticleFig(id=1281326815721533675, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图6, caption=不同形态线谱, figureFileSmall=aX9aJodKcwx5dHWmu2NL9A==, figureFileBig=Bjgza7bCaGIvfIsRzoQ2Og==, tableContent=null), ArticleFig(id=1281326817680273644, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.7, caption=ROC curves of different models, figureFileSmall=FwtB7RvO+S9bOo6X+TQjpA==, figureFileBig=+Y+k2eaTsXGRdB7CHeHETA==, tableContent=null), ArticleFig(id=1281326817768354029, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图7, caption=不同模型ROC曲线, figureFileSmall=FwtB7RvO+S9bOo6X+TQjpA==, figureFileBig=+Y+k2eaTsXGRdB7CHeHETA==, tableContent=null), ArticleFig(id=1281326817835462894, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.8, caption=Comparison of LLA of different models, figureFileSmall=N2yyGK3Jc6hwMn4b65hOyA==, figureFileBig=TH6ZIM6en6sxNv0gan8IVQ==, tableContent=null), ArticleFig(id=1281326817910960367, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图8, caption=不同模型LLA对比, figureFileSmall=N2yyGK3Jc6hwMn4b65hOyA==, figureFileBig=TH6ZIM6en6sxNv0gan8IVQ==, tableContent=null), ArticleFig(id=1281326817973874928, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=EN, label=Fig.9, caption=Feature extraction results of sea trial LOFAR spectral, figureFileSmall=YrmyTkF8B202KWoSqvRmsQ==, figureFileBig=b2N0bwt3oIzVhjVEhbT+dQ==, tableContent=null), ArticleFig(id=1281326818057761009, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, language=CN, label=图9, caption=海试LOFAR谱特征提取结果, figureFileSmall=YrmyTkF8B202KWoSqvRmsQ==, figureFileBig=b2N0bwt3oIzVhjVEhbT+dQ==, tableContent=null)], 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.1007-7294.2026.01.015, detailUrlEn=https://castjournals.cast.org.cn/joweb/cblx/EN/10.3969/j.issn.1007-7294.2026.01.015, pdfUrlCn=https://castjournals.cast.org.cn/joweb/cblx/CN/PDF/10.3969/j.issn.1007-7294.2026.01.015, pdfUrlEn=https://castjournals.cast.org.cn/joweb/cblx/EN/PDF/10.3969/j.issn.1007-7294.2026.01.015, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1768406400000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于机器学习的线谱特征提取方法
收藏切换
PDF下载
师俊杰 1, 2, 3, 4 , 熊凌霜 4 , 孙大军 1, 2, 3
船舶力学 | 流体与结构声学 2026,30(1): 159-167
收起
收藏切换
船舶力学 |流体与结构声学 2026 , 30 (1) : 159 -167
基于机器学习的线谱特征提取方法
全屏
[Author(id=1281326811841802437, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=1281326811950854346, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326811841802437, language=EN, stringName=Jun-jie SHI, firstName=Jun-jie, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, 4, address=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China
2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China
3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China
4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326812022157515, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326811841802437, language=CN, stringName=师俊杰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, 4, address=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001
2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001
3.哈尔滨工程大学 水声工程学院,哈尔滨 150001
4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000, bio={"content":"

师俊杰(1980–),男,博士,副教授

"}, bioImg=null, bioContent=

师俊杰(1980–),男,博士,副教授

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326811518841016, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=1., ext=[AuthorCompanyExt(id=1281326811523035321, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China), AuthorCompanyExt(id=1281326811531423930, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001)]), AuthorCompany(id=1281326811602727099, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=2., ext=[AuthorCompanyExt(id=1281326811611115708, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China), AuthorCompanyExt(id=1281326811615310013, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001)]), AuthorCompany(id=1281326811686613182, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=3., ext=[AuthorCompanyExt(id=1281326811695001791, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China), AuthorCompanyExt(id=1281326811699196096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.哈尔滨工程大学 水声工程学院,哈尔滨 150001)]), AuthorCompany(id=1281326811753722049, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=4., ext=[AuthorCompanyExt(id=1281326811762110658, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China), AuthorCompanyExt(id=1281326811770499267, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000)])]), Author(id=1281326812080877773, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=xls0016@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1281326812160569551, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812080877773, language=EN, stringName=Ling-shuang XIONG, firstName=Ling-shuang, middleName=null, lastName=XIONG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326812219289808, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812080877773, language=CN, stringName=熊凌霜, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326811753722049, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=4., ext=[AuthorCompanyExt(id=1281326811762110658, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China), AuthorCompanyExt(id=1281326811770499267, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811753722049, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000)])]), Author(id=1281326812282204370, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, 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=1281326812382867670, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812282204370, language=EN, stringName=Da-jun SUN, firstName=Da-jun, middleName=null, lastName=SUN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China
2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China
3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1281326812458365143, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, authorId=1281326812282204370, language=CN, stringName=孙大军, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001
2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001
3.哈尔滨工程大学 水声工程学院,哈尔滨 150001, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1281326811518841016, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=1., ext=[AuthorCompanyExt(id=1281326811523035321, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China), AuthorCompanyExt(id=1281326811531423930, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811518841016, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001)]), AuthorCompany(id=1281326811602727099, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=2., ext=[AuthorCompanyExt(id=1281326811611115708, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China), AuthorCompanyExt(id=1281326811615310013, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811602727099, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001)]), AuthorCompany(id=1281326811686613182, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, xref=3., ext=[AuthorCompanyExt(id=1281326811695001791, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China), AuthorCompanyExt(id=1281326811699196096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1281326802639499379, companyId=1281326811686613182, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.哈尔滨工程大学 水声工程学院,哈尔滨 150001)])])]
师俊杰1, 2, 3, 4, 熊凌霜4 , 孙大军1, 2, 3
作者信息
  • 1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001
  • 2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001
  • 3.哈尔滨工程大学 水声工程学院,哈尔滨 150001
  • 4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000
通讯作者:
熊凌霜(2000–),女,硕士,通讯作者,E-mail:
作者简介:

师俊杰(1980–),男,博士,副教授

Line spectrum feature extraction based on machine learning
Jun-jie SHI1, 2, 3, 4, Ling-shuang XIONG4 , Da-jun SUN1, 2, 3
Affiliations
  • 1.National Key Laboratory of underwater Acoustic Technology, Harbin Engineering University Harbin 150001, China
  • 2.Key Laboratory of Marine Information Acquisition and Security (Harbin Engineering University), Ministry of Industry and Information Technology, Harbin 150001, China
  • 3.College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China
  • 4.Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao 266000, China
出版时间: 2026-01-15 doi: 10.3969/j.issn.1007-7294.2026.01.015
文章导航
收藏切换

针对人工提取线谱特征方法的不足,本文对线谱特征提取原理进行研究,提出一种基于机器学习的线谱特征提取方法。搭建基于卷积神经网络的编码器–解码器,并在卷积和池化层间引入注意力机制,使输入数据的重要特征占据更高的权重,从而提高特征提取的准确度。将模型与U–Net模型和TPSW算法在低信噪比情况下进行比较,并在实际数据上进行测试。实验结果表明,在谱级信噪比为5 dB时,改进模型线定位精度可达到0.823,在0~5 dB时均优于U–Net模型和TPSW算法,达到了提取线谱有效信息,提高水下目标检测准确率的目的。

线谱特征提取  /  机器学习  /  编码–解码器  /  水下目标检测

This paper is to study the principle of line spectrum feature extraction. In view of the deficiency of manual line spectrum feature extraction method, a line spectrum feature extraction method based on machine learning was proposed. The Encoder-Decoder based on convolution neural network was built, and the attention mechanism was introduced between the convolution and pooling layers, so that the important features of the input data could occupy a higher weight to enhance the accuracy of feature extraction. The model was compared with U-Net model and TPSW algorithm in the case of low signal-to-noise ratio, and tested on the actual data. The experimental results show that the improved model achieves a line positioning accuracy of 0.823 at a signal-to-noise ratio of 5 dB. This performance is better than that of the U-Net model and TPSW algorithm with in the 0~5 dB range. Thus the model effectively extracts line spectrum information and improves the accuracy of underwater target detection.

line spectrum feature extraction  /  machine learning  /  Encoder-Decoder  /  underwater target detection
师俊杰, 熊凌霜, 孙大军. 基于机器学习的线谱特征提取方法. 船舶力学, 2026 , 30 (1) : 159 -167 . DOI: 10.3969/j.issn.1007-7294.2026.01.015
Jun-jie SHI, Ling-shuang XIONG, Da-jun SUN. Line spectrum feature extraction based on machine learning[J]. Journal of Ship Mechanics, 2026 , 30 (1) : 159 -167 . DOI: 10.3969/j.issn.1007-7294.2026.01.015
近年来,热门的机器学习技术在许多领域都取得了优于常规方法的成就,人类如何运用机器学习融合多领域的研究,正受到越来越多的关注。高效的目标检测技术,是把握机器学习发展方向的关键之一,水声目标检测与机器学习的结合逐渐成为近年来水声领域最热门的话题之一。水声目标检测的研究涉及许多方面,线谱检测是水声目标检测领域中一种重要的研究技术[1]。考虑到信号的局部平稳性,一般使用低频分析记录(Low Frequency Analysis Record, LOFAR)谱进行线谱特征提取[2]
传统水声目标检测算法在信噪比降低时没有很好的适应能力,因此传统算法的准确率已经难以满足对水声目标检测准确率日益增长的需求。对LOFAR谱采用图像处理、神经网络等机器学习领域的方法进行技术研究,可以提高线谱特征提取的精确度,并获得LOFAR谱中所需要的有效信息,从而提升水声目标检测的准确率。
干扰水下信号的噪声是多种噪声源的综合叠加,因此水下信号的LOFAR谱十分复杂,对于LOFAR谱的线谱检测往往会从图像处理和神经网络等角度[3]出发。从图像滤波角度把LOFAR谱看成一幅包含需要提取特征的图像,通过统计和图像处理手段来处理线谱。Chen等[4]采用图像滤波技术中的双通过分离窗方法,检测高度不规则时变线谱信号时效果较好,但在处理低信噪比信号时效果比较差。从神经网络角度处理LOFAR谱,是将高维的数据输入网络,让神经网络从中提取空间特征。陈敬军等[5]根据声呐兵的检测流程,融合线谱识别、检测和跟踪后提出了人工智能方法,实现了低信噪比情况下的自动检测,能在低信噪比的情况下自动地以高检测概率和低虚警概率检测线谱。Jung等[6]利用U–Net对LOFAR谱特征进行提取,在模拟的数据集上得到了92%的召回率,提高了低信噪比情况下的线谱特征提取准确率,但其没有在真实数据上进行实验,无从得知面对真实情况下的复杂线谱时能否很好地处理线谱的特征。Chen等[7]开发了一种基于LOFAR谱的卷积神经网络(Convolutional Neural Networks, CNN)进行在线识别。利用CNN强大的特征提取能力,他们所提出的LOFAR–CNN可以优化特征质量,进一步提高后续识别精度。
基于机器学习的特征提取任务的核心是通过数据预处理将目标的声音信号转化为频谱,从而将声音信号的识别任务转化为频谱的处理任务,通过对频谱的处理和识别提取出特征信息。
本文提出一种基于机器学习的编码器–解码器模型来提高水下目标信号LOFAR谱的特征提取能力。模型基于卷积神经网络搭建编码–解码结构,并在卷积和池化层间引入注意力机制,使输入数据的重要特征占据更高的权重。在此基础上,向神经网络输入由LOFAR谱图构建的数据集,经过对编码–解码神经网络的多次训练和测试得到模型,实现对水下声信号的LOFAR谱特征提取,从而提升特征提取的准确度。
编码–解码器是依靠最小化输入与重构信号之间的误差来得到输入的隐含层特征表示。本文所搭建的编码–解码网络如图1所示。将LOFAR谱输入网络后,首先经过编码器部分,编码器由卷积层、池化层和注意力层构成。所有卷积核大小都统一为3×3。在样本上和通道上的步幅设置为1,填充的值由算法内部根据卷积核大小计算,目的是让输出尺寸和输入相等。激活函数选择ReLU激活函数。经过卷积层的滤波后,数据输入至池化层进行降维并压缩参数数量,减小网络的过拟合程度,在设计网络时使用2×2的最大池化尺寸。在第二部分卷积层和池化层中间引入了自注意力机制,用于更好地关注输入数据的特征。经过编码模块后得到特征矩阵,输入至解码模块进行上采样升维,将提取到的特征解码为与输入相同的结构并输出网络。
如果说卷积神经网络相当于编码器加上一个分类层用作解码,那么基于卷积神经网络的编码器–解码器模型则是传统网络的扩建和续写,通过它的编码–解码过程可以实现端到端的任务。编码–解码模型主体结构由编码器和解码器组成,编码过程就是将输入序列编码转化为一个固定长度的特征向量,解码过程则是将特征向量再解码升维转化为输出,网络的输入和输出的尺寸维度都是相同的。
首先,将输入x通过编码过程的函数f转变为变量y,因此,
$ y = f(x) $
然后,通过解码过程的函数g得到重构的信号r,得到
$ \tilde x = g(y) = g(f(x)) $
定义误差e为原始输入x与重构信号$ \tilde x $之间的误差,有
$ e = x - \tilde x $
接着,通过训练模型,将输入x和输出$ \tilde x $进行对比,直到其无限接近,也就是e达到最小值时获得满意的输出结果。编码–解码网络的主要目标是获得与输入相同的输出,可以通过编码学习到数据的隐含特征,再通过解码用这些特征重构原始的输入数据,在这过程中,先学习数据的重要特征,再让输出数据与输入数据的差异最小化来达到目的。
由于本文所获得的真实数据目标信号的频带范围约为250 Hz,因此将输入谱图的尺寸设置为256×256,即可满足对目标线谱特征进行提取的目的。在经过多次降维后,得到尺寸为32×32×256的特征矩阵,然后再通过解码器将编码器输出的低维特征矩阵转换回原始的数据空间,用以恢复原始输入。
编码模块由卷积神经网络的卷积层和池化层组成。卷积层的主要操作是使用卷积核对输入执行卷积操作,并检测输入局部的连接特性,以获得新的输出。简单来说,卷积层的作用就是提取局部特征,卷积核就相当于特征提取器。在多层卷积层中,前一卷积层的输出就是后一卷积层的输入,第一层是原始数据,第二层就是第一层输出的激活映射。为了使激活映射拥有更复杂的特征,可以进行多次卷积操作。如果输入矩阵为I,卷积核矩阵为Km×n),mn为核矩阵的长和宽,输出矩阵为O,有
$ O(i,j) = (K * I)(i,j) = \sum\limits_m {\sum\limits_n {I(i - m,j - n)K(m,n)} } $
式中:$ * $表示卷积符号;Oi, j)表示输出矩阵O中位置(i, j)的输出值。卷积层的输出尺寸由输入和卷积核的尺寸决定。
编码器将原始数据转换为低维度的表示,以捕捉数据的关键特征,将其编码为低维的特征矩阵。与编码器对应的是解码器,解码器将编码器输出的低维特征矩阵转换回原始的数据空间,用以恢复原始输入。解码器由卷积层和上采样层构成,卷积层用于恢复通道数,上采样层则用于恢复在编码器中被压缩的维度。
将多次卷积、池化操作后得到的编码作为特征隐层的表示,即输入数据的特征矩阵,而后再输入解码器进行上采样恢复维度。基于卷积神经网络的编码器–解码器网络示意图如图2所示。
使用上采样恢复维度的最后一步是将结果进行回归分类,分类问题的本质是使用机器学习模型将训练数据集的特征和标签转化为一个决策面。
在机器学习中,逻辑回归算法是常见二分类问题的核心算法。逻辑回归的输入是训练数据的特征及标签,经过逻辑回归模型后输出分类结果,即线谱图中的像素点是否为线谱特征点。在最后一层作为输出层的卷积层引入逻辑回归函数sigmoid进行逻辑回归,逻辑函数的公式为
$ y = \frac{1}{{1 + {e^{ - z}}}} $
其中,回归方程$z = ax + b$y表示分类结果标签为1时x的概率值,ab为常数。然后根据需要,在后续处理中将x的概率值通过决策面分类映射到标签0或1上。逻辑函数获得预测数值后,根据决策面输出分类结果。
自注意力机制恰如其名,就是模仿人类会对感兴趣的东西更关注的特点,对输入数据的某些部分给予更高的关注度,根据不同的注意度分配不同的权重,从而可以从信息庞杂的数据中提取出更显著、更需要的特征。它的核心思想为,对序列中的每个元素,计算其与其他元素之间的相似度,并将这些相似度归一化为注意力权重。然后,通过将每个元素与对应的注意力权重进行加权求和,可以得到自注意力机制的输出[8]
近年来,注意力机制发展迅速,传统注意力机制的输入和输出的内容表现形式是不同的,可以被认为是输入和输出中每个信息组的对齐模型,而自注意力机制是针对输入或输出的内部所有元素,两者所作用的对象不同。由于并不是全部的LOFAR谱都对线谱特征提取任务同等重要,只有线谱所在的区域才需要投以更多注意力,因此借鉴了自注意力机制的思想,在网络里寻找出输入的空间信息聚集最多的区域并加大权重,从而提升特征提取的准确度。
图3所示为在卷积层后加入注意力(Attention)层。计算注意力时,输入是卷积层的输出xi,卷积层的权重为W,偏置为B,输入为${h_i}$,经过ReLU激活函数的注意力层输入可以表示为
$ {x_i} = {\text{ReLU}}(W{h_i} + B) $
注意力模块在输入的卷积层后加入一个使用了softmax激活函数的全连接层,它的神经元个数与输入向量的维度相同,权重为${W^w}$,偏置为${B^w}$,主要作用是通过softmax从输入中选择对特征贡献最大的维度作为权重wi,其表达式为
$ {w_i} = {\text{softmax}}({W^w}{x_i} + {B^w}) $
然后全连接层所输出的权重与卷积层的输出进行位对位(Element Wise)的乘法,相当于使用注意力机制将动态捕捉到的信息对卷积层输出结果进行动态的加权,然后将其连接起来生成一个有效的特征描述符,有助于汇合各层网络捕捉提取到的特征信息,输入下一层网络。自注意力机制的架构如图4所示。
机器学习中的自注意力机制在图像视觉和序列建模等方向都有着出色表现。LOFAR谱图中线谱特征提取也可以看做是一种图像任务,利用卷积神经网络强大的特征提取能力和注意力机制更关注输入数据中更重要目标特征区域的特点,通过编码器-解码器网络端到端线谱特征提取架构,可以实现LOFAR谱图中的频率线和环境噪声的快速区分,从而达到将LOFAR谱图中的特征线谱提取出来的目的。
为了检验在低信噪比下检测不规则波动频率线的能力,采用Paris[9]提出的模型生成信号
$ {s_t}{\text{ = }}{a_{{t}}}\sin (2{\text{π}} {f_t} + {\phi _t}) + {\varepsilon _t}\quad t = 0,\cdots ,T - 1 $
其中,观测值st被假定为被加性零均值高斯白噪声εt ~ N(0, σ2)扰乱的正弦波,瞬时频率ft是一个随机游走模型,每个时刻的频率值都是由前一个时刻的频率值加上一个随机扰动得到的。设计谱级信噪比为
$ {P_{{\mathrm{SNR}}}} = 10{\log _{10}}\frac{{{P_{\text{s}}}}}{{{P_{\mathrm{n}}}}}{\text{ = }}10{\log _{10}}\frac{{{a_t}^2 \cdot B}}{{{\sigma ^2}}} $
其中,Ps为信号的功率;Pn表示噪声功率;σ2为加性零均值高斯白噪声的方差,即噪声的平均功率;B为噪声频带范围。生成信号后进行短时傅里叶(Short–time Fourier Transform, STFT)变换计算得到LOFAR谱。与传统的傅里叶变换要求信号稳定性不同,短时傅里叶变换适用于非平稳信号,它利用了信号的短期平稳特性。在对信号进行加窗和分帧后,进行傅里叶变换,得到时频信号,从而更准确地刻画出信号频率分量和时间节点的分布。计算公式如下:
$ {S_{{\mathrm{STFT}}}}\{ s(t)\} {\text{ = }}\int_{ - \infty }^\infty {s(t)} \omega (t - \tau ){{\mathrm{e}}^{ - {\mathrm{j}}wt}}{\mathrm{d}}t $
其中,st)为信号,$\omega (t)$是窗函数。
为了达到良好的训练效果,生成数目不同、形态各异的数据集,由数据集谱级信噪比为0~10 dB范围内的线谱构成,包含4950张训练集、550张测试集。图5为数据集中训练集和标签的构成示例。
图6是为了模拟海试时可能遇到的真实线谱形状而生成的多种形态的线谱示例,这些示例在各信噪比上均匀分布。
使用神经网络来实现线谱的特征提取,相当于对LOFAR谱图内每个像素点进行预测是否为线谱点,可以理解为对线谱中的每个像素进行二分类,最后输出矩阵的每个值都是该像素点为线谱点的概率值,大小在0到1之间。当使用神经网络进行分类和预测时,通常使用交叉熵损失函数,这是由于交叉熵损失函数能在一定程度上避免梯度消失。此外,当模型对某个样本的预测结果与实际标签差距较大时,交叉熵损失函数会给出一个较大的梯度值,因此,当权重更新快时代表此时误差较大,权重更新慢时代表此时误差较小,这样能更快地更新模型参数,加快训练速度,使用交叉熵损失函数能更好地提升神经网络的性能。在交叉熵损失函数中有专门针对二分类问题的二分类交叉熵损失(Binary Cross Entropy loss, BCE loss),其激活函数为sigmoid。激活函数sigmoid和BCE loss相结合时采用了类间竞争的机制,这种方法注重正样本预测的准确率,可以很好地学习类与类之间的关系,优化类间距离,即谱图中的线谱点和非线谱点之间的细节区分会更明显。BCE loss的计算公式为
$ {\mathrm{BCE}}\;loss = \frac{1}{N}\sum\limits_{n = 1}^N {{l_n}} $
其中,N表示样本数量,${l_n}$为第n个样本对应的损失,有
$ {l_n} = - ({y_n}\log \delta ({a_n})) + (1 - {y_n})\log (1 - \delta ({a_n})) $
在反向传播训练期间使用交叉熵损失函数,权重会一直改变,也就是每次训练模型都能学到新的信息。
使用受试者工作特性(Receiver Operating Characteristic Curve, ROC)曲线和线定位精度[10-11](Line Location Accuracy, LLA)来分别评估检测性能和恢复频率线的精度。
模型召回率(Recall)也被叫做真阳率(True Positive Rate, TPR)或命中率,其反映的是模型对测试集中正样本的预测能力,也就是本实验中神经网络对线谱特征正确提取的能力,一般来说,将正样本预测为正样本的称为真阳性(True Positive, TP),将正样本预测为负样本的称为假阴性(False negative, FN)。召回率反映的是在LOFAR谱图中模型预测出的正样本像素点占总的真实正样本像素点的比例,即
$ {\mathrm{recall}} = {\mathrm{TPR}} = \frac{{{\mathrm{TP}}}}{{{\mathrm{TP}} + {\mathrm{FN}}}} $
误报率(False Alarm Rate, FPR)也被称为虚警率,能检测出将负样本预测为正样本占总的负样本的比例,FPR的数值越接近0,即模型的虚警率越低,表示模型的提取性能越好,有
$ {\mathrm{FPR}} = \frac{{{\mathrm{FP}}}}{{{\mathrm{TN}} + {\mathrm{FP}}}} $
其中,假阳性率(False Positive Rate, FP)表示标签中是负样本却被预测为正样本的像素点数,真阴性率(True Negative Rate, TN)表示标签中是负样本且被预测为负样本的像素点数。
ROC(Receiver Operating Characteristic, ROC)曲线反映了模型在不同的判定标准下对同一信号刺激的反应,改变不同的阈值门限可以得到多组TPR和FPR,将这些(TPR, FPR)对连接起来就得到了模型的ROC曲线。曲线中越接近左上角的点召回率越高,虚警率越低,代表在此阈值下模型的检测性能越好。绘制出ROC曲线后,可以得到曲线下的面积,该面积又被称为AUC(Area Under the Curve, AUC)。AUC是一种用于评估分类模型性能的常用指标,代表分类器在不同阈值下的预测能力,通常取值范围在0到1之间。AUC越大的分类器,正确率越高。
线位置精度(LLA)是LOFAR谱图特征提取的准确率的另一种常用的评判标准,它是指检测出的轨迹和真实轨迹之间的平均距离,这种标准可以用来比较不同检测算法在同一数据集上的性能。
线位置精度的评判步骤为:首先,将真实轨迹和检测出的轨迹转化为坐标列表,然后计算每个坐标点与对应的真实轨迹之间的欧氏距离,求出每个坐标点对应的最小欧氏距离d,最后求出所有距离的平均值F[12]F越小,表示检测出的轨迹越接近真实轨迹,即线位置精度越高。线位置精度的表达式为
$ F{\text{ = }}\frac{1}{{\max ({I_l},{I_A})}}\sum\limits_{i = 1}^{{I_A}} {\frac{1}{{1 + \alpha {d^2}(i)}}} $
其中,Il为理想正样本像素点个数;IA为实际检测出的正样本像素点个数;di)表示第i个实际点到理想点的欧氏距离;α为一个常数,一般取1;当Il = IAdi)=0时,F取最大值1。
将5 dB的单根频率漂移的线谱输入模型进行多次测试,设置阈值范围0.0001~ 0.9,连接各个点,画出ROC曲线并计算出AUC值,图7是本文的模型与无注意力层的编码器–解码器模型和Ronneberger[13]提出的U–Net模型的ROC曲线。在5 dB时各个神经网络模型的检测能力排序为:改进编码器–解码器 > 编码器–解码器 > U–Net。
图8是本文的模型与无注意力层的编码器–解码器、U–Net以及传统TPSW算法的LLA对比。在谱级信噪比0~5 dB时,U–Net模型的线定位精度为0.068~0.458,无注意力层的线定位精度为0.238~0.686,而传统TPSW算法的线定位精度为0.025~0.123,本文模型的线定位精度为0.360~0.823,特征提取效果最好。
图9是本文模型与U–Net和TPSW算法在真实海试数据的LOFAR谱特征提取上的提取结果对比。图9(a)为真实数据的其中一段LOFAR谱,该数据包含受背景噪声影响强弱不同的两个信号;图9(b)为U–Net模型的提取结果,可以看见强的线谱大致轮廓基本能提取出来,低信噪比线谱有很大程度的特征缺失;图9(c)为TPSW算法提取结果,能够看见大量背景噪声也被误判为线谱频率点,线谱不连贯,低信噪比线谱被高信噪比线谱和背景噪声所抑制,很难判断出形状与细节;图9(d)为本文模型所提取出的线谱,可以看见其很好地保留了线谱细节和特征,无论对较高信噪比还是低信噪比的线谱均表现出良好的提取效果。
在舰船水下目标检测的工程实践中,提取线谱特征的传统方法常受制于信噪比条件,在低信噪比情况下往往效果不佳。为了提升线谱特征提取在低信噪比情况下的准确率,引入机器学习的方法对线谱特征进行提取。搭建了编码器–解码器网络模型并引入注意力机制,赋予重要特征更高的权重,提高了模型在低信噪比下的特征提取准确率。未来计划加深网络层数,进一步优化模型,提高模型的泛化能力。

参考文献 引证文献
排序方式:
1
吴国清, 李 靖, 陈耀明, 等. 舰船噪声识别(I)──总体框架、线谱分析和提取[J]. 声学学报, 1998, 5: 394‒400.
Wu G Q, Li J, Chen Y M, et al. Ship noise recognition (I)−Overall framework, line spectrum analysis and extraction[J]. Acta Acustica, 1998, 5: 394‒400. (in Chinese)
2
Zhao L, Li B. A novel method for extracting frequency line on lofargram based on feature function[C]//The 4th International Conference on Image, Video and Signal Processing, 2022.
3
李 山, 王德俊, 王海斌. 一种基于线谱特征函数提取LOFAR图线谱的方法[J]. 声学技术, 2016, 35(4): 373‒377.
Li S, Wang D J, Wang H B. A method for extracting LOFAR graph line spectrum based on line spectrum feature function[J]. Acoustic Technology, 2016, 35(4): 373‒377. (in Chinese)
4
Chen C H, Lee J D, Lin M. Classification of underwater signals using neural networks[J]. Journal of Applied Science and Engineering, 2000, 3(1): 31‒48.
5
陈敬军, 陆佶人, 刘 淼. 基于人工智能的线谱检测技术[J]. 船舶工程, 2004, 3: 68‒71.
Chen J J, Lu J R, Liu M. Line spectrum detection technology based on artificial intelligence[J]. Ship Engineering, 2004, 3: 68‒71. (in Chinese)
6
Jung D J, Park J, Lee S H, et al. Deep learning based tonal detection for passive sonar signals[C]//UTD, 2019.
7
Chen J, Han B, Ma X, et al. Underwater target recognition based on multi-decision LOFAR spectrum enhancement: A deep-learning approach[J]. Future Internet, 2021, 13: 265.
8
Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[J]. arXiv, 2017.
9
Paris Sebastien, Jauffret C, et al. Frequency line tracking using HMM-based schemes[J]. IEEE Transactions on Aerospace & Electronic Systems, 2003, 39(2): 439‒449.
10
Lampert T A, O'Keefe S E M. On the detection of tracks in spectrogram images[J]. Pattern Recognit, 2013, 46: 1396‒1408.
11
Martino J C D, Tabbone S. An approach to detect lofar lines[J]. Pattern Recognition Letters, 1996, 17(1): 37‒46.
12
Pratt W K. Digital image processing[M]. Second Edition. New York: Wiley Interscience, 1991.
13
Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical imagesegmentation[J]. Springer International Publishing, 2015.
2026年第30卷第1期
PDF下载
91
46
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.1007-7294.2026.01.015
  • 接收时间:2025-09-20
  • 首发时间:2026-07-07
  • 出版时间:2026-01-15
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-09-20
基金
作者信息
    1.哈尔滨工程大学 水声技术全国重点实验室,哈尔滨 150001
    2.海洋信息获取与安全工信部重点实验室(哈尔滨工程大学)工业和信息化部,哈尔滨 150001
    3.哈尔滨工程大学 水声工程学院,哈尔滨 150001
    4.哈尔滨工程大学 青岛创新发展基地,山东 青岛 266000

通讯作者:

熊凌霜(2000–),女,硕士,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.1007-7294.2026.01.015
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