Article(id=1149768571858895360, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1149768563956826506, articleNumber=null, orderNo=null, doi=10.12404/j.issn.1671-1815.2406880, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1726156800000, receivedDateStr=2024-09-13, revisedDate=1742400000000, revisedDateStr=2025-03-20, acceptedDate=null, acceptedDateStr=null, onlineDate=1752055789198, onlineDateStr=2025-07-09, pubDate=1749312000000, pubDateStr=2025-06-08, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1752055789198, onlineIssueDateStr=2025-07-09, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1752055789198, creator=13701087609, updateTime=1752055789198, updator=13701087609, issue=Issue{id=1149768563956826506, tenantId=1146029695717560320, journalId=1146123166801305609, year='2025', volume='25', issue='16', pageStart='6587', pageEnd='7021', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1752055787314, creator=13701087609, updateTime=1768456850262, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1218559607937618069, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1149768563956826506, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1218559607937618070, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1149768563956826506, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=6733, endPage=6741, ext={EN=ArticleExt(id=1149768572047639044, articleId=1149768571858895360, tenantId=1146029695717560320, journalId=1146123166801305609, language=EN, title=Design of T-beam Formwork Polishing Robot Based on Deep Learning, columnId=1156262732765717457, journalTitle=Science Technology and Engineering, columnName=Papers·Mechanical and Instrumental Industry, runingTitle=null, highlight=null, articleAbstract=

A deep learning based T-beam formwork polishing robot was designed for the problems of difficult and time consuming polishing of T-beam formwork for variable cross-section. Firstly, an adaptive polishing structure was proposed to solve the technical problem that the existing polishing device cannot fit the inner variable cross-section of the T-beam formwork, and the polishing roller was easy to get stuck in the T-beam formwork partition. Secondly, in order to realize the quantitative monitoring of the polishing quality, a YOLOv8n-DSE algorithm was proposed to identify concrete dirt and stains on the formwork, the DySample dynamic up-sampling module was introduced to enhance the anti-interference ability of the model and accelerate the calculation speed, to improve the accuracy of small target detection, the SOEP (small object enhance pyramid) module was designed to improve the detection performance of small target detection through the SPDConv(space to depth convolution) to obtain the information features of the small target and give them to the CSP(cross stage partial)-Omni-Kernel for the integration of the features. Finally, the EMA(exponential moving average)-SlideLoss was replaced to make the model more concerned with the quantitative monitoring of the concrete, allowed the model to focus more on difficult targets, which can improve the effect on difficult case detection. The accuracy, recall, and mAP(mean average precision) values are improved by 3.1%, 9.7%, and 3.2%, respectively, compared with those before the improvement. The improved model was deployed to the robot and tested in the field. The results show that the equipment meets the plant's needs for polishing variable-section T-beam formwork.

, correspAuthors=Jun-feng SHI, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Ying-quan CHEN, Jun-feng SHI, Lei ZHAI, Chen LIU, Hui-dong LI), CN=ArticleExt(id=1149768583883964513, articleId=1149768571858895360, tenantId=1146029695717560320, journalId=1146123166801305609, language=CN, title=基于深度学习的T梁模板打磨机器人设计, columnId=1156262732954461139, journalTitle=科学技术与工程, columnName=论文·机械、仪表工业, runingTitle=null, highlight=null, articleAbstract=

针对变截面T梁模板打磨工作难度大,费时长等问题,设计一种基于深度学习的T梁模板打磨机器人。首先提出一种自适应打磨结构,解决了现有打磨装置无法贴合T梁模板内侧变截面,打磨辊容易卡在T梁模板隔断内的技术问题;其次为实现打磨质量的量化监测,提出一种YOLOv8n-DSE算法,对模板上的混凝土痂块及污渍进行识别,引入DySample动态上采样模块,增强模型的抗干扰能力并加快计算速度;为提高小目标检测的精度,设计SOEP(small object enhance pyramid)模块,通过SPDConv(space to depth convolution)获得小目标信息特征给到CSP(cross stage partial)-Omni-Kernel进行特征整合提高小目标的检测检测性能;最后替换EMA(exponential moving average)-Slide Loss让模型更加关注困难目标,可以改善在难例检测上的效果。准确率、召回率和均值平均精度(mean average precision,mAP)值相较改进前分别提升了3.1%,9.7%和3.2%,将改进的模型部署至机器人并进行现场测试。结果表明:设备满足工厂对变截面T梁模板的打磨需求。

, correspAuthors=石峻峰, authorNote=null, correspAuthorsNote=
* 石峻峰(1974—),男,汉族,吉林辽源人,博士,讲师。研究方向:智能建造、工程结构BIM技术。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=Zi+KOvN3CKpXV1z8yPftMw==, magXml=LIZqHw1NiMttg9i4l9eItg==, pdfUrl=null, pdf=s6ZTByKm/KWzJXcxWJukFQ==, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=kFZKhDeaDlpfmnDgvkLqWg==, mapNumber=null, authorCompany=null, fund=null, authors=

陈映全(2000—),男,汉族,湖北黄石人,硕士研究生。研究方向:智能建造、桥梁智能建造装备研发。E-mail:

, authorsList=陈映全, 石峻峰, 翟磊, 刘郴, 李惠东)}, authors=[Author(id=1177979127350051764, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=2454648925@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1177979127429743543, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979127350051764, language=EN, stringName=Ying-quan CHEN, firstName=Ying-quan, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 School of Civil, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1177979127526212536, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979127350051764, language=CN, stringName=陈映全, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 湖北工业大学土木建筑与环境学院, 武汉 430068, bio={"content":"

陈映全(2000—),男,汉族,湖北黄石人,硕士研究生。研究方向:智能建造、桥梁智能建造装备研发。E-mail:

"}, bioImg=null, bioContent=

陈映全(2000—),男,汉族,湖北黄石人,硕士研究生。研究方向:智能建造、桥梁智能建造装备研发。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1177979127144530857, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=1, ext=[AuthorCompanyExt(id=1177979127152919466, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127144530857, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Civil, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China), AuthorCompanyExt(id=1177979127161308075, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127144530857, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 湖北工业大学土木建筑与环境学院, 武汉 430068)])]), Author(id=1177979127584932794, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=592297026@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1177979127702373308, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979127584932794, language=EN, stringName=Jun-feng SHI, firstName=Jun-feng, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 School of Civil, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1177979127769482173, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979127584932794, language=CN, stringName=石峻峰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 湖北工业大学土木建筑与环境学院, 武汉 430068, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1177979127144530857, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=1, ext=[AuthorCompanyExt(id=1177979127152919466, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127144530857, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Civil, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China), AuthorCompanyExt(id=1177979127161308075, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127144530857, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 湖北工业大学土木建筑与环境学院, 武汉 430068)])]), Author(id=1177979127840785343, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, 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=1177979127995974593, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979127840785343, language=EN, stringName=Lei ZHAI, firstName=Lei, middleName=null, lastName=ZHAI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 National Key Laboratory of Intelligent and Green Bridge Construction, Wuhan 430050, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1177979128050500546, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979127840785343, language=CN, stringName=翟磊, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 桥梁智能与绿色建造全国重点实验室, 武汉 430050, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1177979127224222636, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=2, ext=[AuthorCompanyExt(id=1177979127232611245, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127224222636, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 National Key Laboratory of Intelligent and Green Bridge Construction, Wuhan 430050, China), AuthorCompanyExt(id=1177979127240999854, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127224222636, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 桥梁智能与绿色建造全国重点实验室, 武汉 430050)])]), Author(id=1177979128109220804, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, 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=1177979128172135366, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979128109220804, language=EN, stringName=Chen LIU, firstName=Chen, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3 China Railway Bridge Science Research Institute Ltd. , Wuhan 430034, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1177979128251827143, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979128109220804, language=CN, stringName=刘郴, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3 中铁大桥科学研究院有限公司, 武汉 430034, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1177979127291331503, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=3, ext=[AuthorCompanyExt(id=1177979127295525808, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127291331503, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 China Railway Bridge Science Research Institute Ltd. , Wuhan 430034, China), AuthorCompanyExt(id=1177979127303914417, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127291331503, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 中铁大桥科学研究院有限公司, 武汉 430034)])]), Author(id=1177979128390239177, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1177979128495096779, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979128390239177, language=EN, stringName=Hui-dong LI, firstName=Hui-dong, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 National Key Laboratory of Intelligent and Green Bridge Construction, Wuhan 430050, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1177979128578982860, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, authorId=1177979128390239177, language=CN, stringName=李惠东, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 桥梁智能与绿色建造全国重点实验室, 武汉 430050, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1177979127224222636, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=2, ext=[AuthorCompanyExt(id=1177979127232611245, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127224222636, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 National Key Laboratory of Intelligent and Green Bridge Construction, Wuhan 430050, China), AuthorCompanyExt(id=1177979127240999854, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127224222636, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 桥梁智能与绿色建造全国重点实验室, 武汉 430050)])])], keywords=[Keyword(id=1177979128700617677, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, orderNo=1, keyword=T-beam formwork for variable cross-section), Keyword(id=1177979128776115150, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, orderNo=2, keyword=deep learning), Keyword(id=1177979128839029711, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, orderNo=3, keyword=polishing robot), Keyword(id=1177979128906138576, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, orderNo=4, keyword=adaptive sanding), Keyword(id=1177979128960664529, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, orderNo=5, keyword=YOLOv8n), Keyword(id=1177979129027773394, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, orderNo=1, keyword=变截面T梁模板), Keyword(id=1177979129094882259, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, orderNo=2, keyword=深度学习), Keyword(id=1177979129157796820, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, orderNo=3, keyword=打磨机器人), Keyword(id=1177979129220711381, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, orderNo=4, keyword=自适应打磨), Keyword(id=1177979129271043030, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, orderNo=5, keyword=YOLOv8n)], refs=[Reference(id=1177979131783431157, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2020, volume=36, issue=11, pageStart=7, pageEnd=22, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=李兰冰, 高雪莲, 黄玖立, journalName=管理世界, refType=null, unstructuredReference=李兰冰, 高雪莲, 黄玖立. “十四五”时期中国新型城镇化发展重大问题展望[J]. 管理世界, 2020, 36(11): 7-22., articleTitle=“十四五”时期中国新型城镇化发展重大问题展望, refAbstract=null), Reference(id=1177979131850540022, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2020, volume=36, issue=11, pageStart=7, pageEnd=22, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=Li Lanbing, Gao Xuelian, Huang Jiuli, journalName=Management World, refType=null, unstructuredReference=Li Lanbing, Gao Xuelian, Huang Jiuli. Prospects for major issues of China's new urbanization development during the “14th Five-Year Plan” period[J]. Management World, 2020, 36(11): 7-22., articleTitle=Prospects for major issues of China's new urbanization development during the “14th Five-Year Plan” period, refAbstract=null), Reference(id=1177979131917648887, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2024, volume=201, issue=null, pageStart=112014, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=Zhang R, Guo T, Li A, journalName=Thin-Walled Structures, refType=null, unstructuredReference=Zhang R, Guo T, Li A. Experimental investigation into demountable dry connections for fully precast frame structures through shaking table tests[J]. Thin-Walled Structures, 2024, 201: 112014., articleTitle=Experimental investigation into demountable dry connections for fully precast frame structures through shaking table tests, refAbstract=null), Reference(id=1177979131997340664, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2022, volume=50, issue=4, pageStart=32, pageEnd=40, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=查道宏, 李军堂, 胡勇, journalName=世界桥梁, refType=null, unstructuredReference=查道宏, 李军堂, 胡勇. 大跨度斜拉桥钢筋混凝土桥塔快速施工技术发展与展望[J]. 世界桥梁, 2022, 50(4): 32-40., articleTitle=大跨度斜拉桥钢筋混凝土桥塔快速施工技术发展与展望, refAbstract=null), Reference(id=1177979132081226745, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2022, volume=50, issue=4, pageStart=32, pageEnd=40, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=Cha Daohong, Li Juntang, Hu Yong, journalName=World Bridges, refType=null, unstructuredReference=Cha Daohong, Li Juntang, Hu Yong. Development and prospects of rapid construction technology for reinforced concrete pylons of long-span cable-stayed bridges[J]. World Bridges, 2022, 50(4): 32-40., articleTitle=Development and prospects of rapid construction technology for reinforced concrete pylons of long-span cable-stayed bridges, refAbstract=null), Reference(id=1177979132135752698, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[4], rfOrder=5, authorNames=李勇海, 余秀平, 刘宁, journalName=null, refType=null, unstructuredReference=李勇海, 余秀平, 刘宁, 等. 铁路预制箱梁的钢模板用打磨机: CN202121692391.X[P]. 2021-12-21., articleTitle=null, refAbstract=null), Reference(id=1177979132202861563, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=Li Yonghai, Yu Xiuping, Liu Ning, journalName=null, refType=null, unstructuredReference=Li Yonghai, Yu Xiuping, Liu Ning, et al. Grinding machine for steel mould plates for prefabricated box girders for railways: CN202121692391.X[P]. 2021-12-21., articleTitle=null, refAbstract=null), Reference(id=1177979132253193212, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2023, volume=7, issue=5, pageStart=304, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=Li Y, Fan Q, Huang H, journalName=Drones, refType=null, unstructuredReference=Li Y, Fan Q, Huang H, et al. A modified YOLOv8 detection network for UAV aerial image recognition[J]. Drones, 2023, 7(5):304., articleTitle=A modified YOLOv8 detection network for UAV aerial image recognition, refAbstract=null), Reference(id=1177979132307719165, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=28, pageStart=12159, pageEnd=12167, url=null, language=null, rfNumber=[6], rfOrder=8, authorNames=蒲玲玲, 杨柳, journalName=科学技术与工程, refType=null, unstructuredReference=蒲玲玲, 杨柳. 改进YOLOv5的多车辆目标实时检测及跟踪算法[J]. 科学技术与工程, 2023, 23(28): 12159-12167., articleTitle=改进YOLOv5的多车辆目标实时检测及跟踪算法, refAbstract=null), Reference(id=1177979132383216638, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=28, pageStart=12159, pageEnd=12167, url=null, language=null, rfNumber=[6], rfOrder=9, authorNames=Pu Lingling, Yang Liu, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Pu Lingling, Yang Liu. Improved real-time detection and tracking algorithm for multi vehicle targets in YOLOv5[J]. Science Technology and Engineering, 2023, 23(28): 12159-12167., articleTitle=Improved real-time detection and tracking algorithm for multi vehicle targets in YOLOv5, refAbstract=null), Reference(id=1177979132513240063, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=35, pageStart=15148, pageEnd=15157, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=张世强, 史卫亚, 张绍文, journalName=科学技术与工程, refType=null, unstructuredReference=张世强, 史卫亚, 张绍文, 等. 基于改进YOLOv5算法的钢铁表面缺陷检测[J]. 科学技术与工程, 2023, 23(35): 15148-15157., articleTitle=基于改进YOLOv5算法的钢铁表面缺陷检测, refAbstract=null), Reference(id=1177979132571960320, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=35, pageStart=15148, pageEnd=15157, url=null, language=null, rfNumber=[7], rfOrder=11, authorNames=Zhang Shiqiang, Shi Weiya, Zhang Shaowen, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Zhang Shiqiang, Shi Weiya, Zhang Shaowen, et al. Steel surfacedefect detection based on improved YOLOv5 algorithm[J]. Science Technology and Engineering, 2023, 23(35): 15148-15157., articleTitle=Steel surfacedefect detection based on improved YOLOv5 algorithm, refAbstract=null), Reference(id=1177979132655845376, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=18, pageStart=7833, pageEnd=7842, url=null, language=null, rfNumber=[8], rfOrder=12, authorNames=王俊岭, 王晨晨, 熊玉华, journalName=科学技术与工程, refType=null, unstructuredReference=王俊岭, 王晨晨, 熊玉华. 基于YOLOv5l-Im的排水管道缺陷检测方法及效果分析[J]. 科学技术与工程, 2024, 24(18): 7833-7842., articleTitle=基于YOLOv5l-Im的排水管道缺陷检测方法及效果分析, refAbstract=null), Reference(id=1177979132760702977, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=18, pageStart=7833, pageEnd=7842, url=null, language=null, rfNumber=[8], rfOrder=13, authorNames=Wang Junling, Wang Chenchen, Xiong Yuhua, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Wang Junling, Wang Chenchen, Xiong Yuhua. Drainage pipe defect detection and effect based on the improved YOLOv5l-Im[J]. Science Technology and Engineering, 2024, 24(18): 7833-7842., articleTitle=Drainage pipe defect detection and effect based on the improved YOLOv5l-Im, refAbstract=null), Reference(id=1177979132928475138, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=8759, pageEnd=8768, url=null, language=null, rfNumber=[9], rfOrder=14, authorNames=Liu S, Qi L, Qin H, journalName=Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, refType=null, unstructuredReference=Liu S, Qi L, Qin H, et al. Path aggregation network for instance segmentation[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 8759-8768., articleTitle=Path aggregation network for instance segmentation, refAbstract=null), Reference(id=1177979132999778307, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=443, pageEnd=459, url=null, language=null, rfNumber=[10], rfOrder=15, authorNames=Sunkara R, Luo T, journalName=Joint European Conference on Machine Learning and Knowledge Discovery in databases, refType=null, unstructuredReference=Sunkara R, Luo T. No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects[C]// Joint European Conference on Machine Learning and Knowledge Discovery in databases. Cham: Springer Nature Switzerland, 2022: 443-459., articleTitle=No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects, refAbstract=null), Reference(id=1177979133146578948, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=1571, pageEnd=1580, url=null, language=null, rfNumber=[11], rfOrder=16, authorNames=Wang C Y, Yuan H, Wu Y H, journalName=Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, refType=null, unstructuredReference=Wang C Y, Yuan H, Wu Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Seattle: IEEE, 2020: 1571-1580., articleTitle=CSPNet: a new backbone that can enhance learning capability of CNN, refAbstract=null), Reference(id=1177979133213687813, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=1426, pageEnd=1434, url=null, language=null, rfNumber=[12], rfOrder=17, authorNames=Cui Y N, Ren W Q, Alois K, journalName=Proceedings of the AAAI Conference on Artificial Intelligence, refType=null, unstructuredReference=Cui Y N, Ren W Q, Alois K. Omni-kernel network for image restoration[C]// Proceedings of the AAAI Conference on Artificial Intelligence. Vancouver: AAAI, 2024: 1426-1434., articleTitle=Omni-kernel network for image restoration, refAbstract=null), Reference(id=1177979133284990982, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=6027, pageEnd=6037, url=null, language=null, rfNumber=[13], rfOrder=18, authorNames=Liu W, Lu H, Fu H, journalName=Proceedings of the IEEE/CVF International Conference on Computer Vision, refType=null, unstructuredReference=Liu W, Lu H, Fu H, et al. Learning to upsample by learning to sample[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Paris: IEEE, 2023: 6027-6037., articleTitle=Learning to upsample by learning to sample, refAbstract=null), Reference(id=1177979133410820103, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2024, volume=155, issue=null, pageStart=110714, pageEnd=null, url=null, language=null, rfNumber=[14], rfOrder=19, authorNames=Yu Z, Huang H, Chen W, journalName=Pattern Recognition, refType=null, unstructuredReference=Yu Z, Huang H, Chen W, et al. YOLO-facev2: a scale and occlusion aware face detector[J]. Pattern Recognition, 2024, 155: 110714., articleTitle=YOLO-facev2: a scale and occlusion aware face detector, refAbstract=null), Reference(id=1177979133586980872, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2021, volume=34, issue=12, pageStart=1143, pageEnd=1151, url=null, language=null, rfNumber=[15], rfOrder=20, authorNames=吕君环, 许柯, 王东, journalName=模式识别与人工智能, refType=null, unstructuredReference=吕君环, 许柯, 王东. 基于指数移动平均知识蒸馏的神经网络低比特量化方法[J]. 模式识别与人工智能, 2021, 34(12): 1143-1151., articleTitle=基于指数移动平均知识蒸馏的神经网络低比特量化方法, refAbstract=null), Reference(id=1177979133675061257, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, doi=null, pmid=null, pmcid=null, year=2021, volume=34, issue=12, pageStart=1143, pageEnd=1151, url=null, language=null, rfNumber=[15], rfOrder=21, authorNames=Lü Junhuan, Xu Ke, Wang Dong, journalName=Pattern Recognition and Artificial Intelligence, refType=null, unstructuredReference= Junhuan, Xu Ke, Wang Dong. Low-bit quantization of neural network based on exponential moving average knowledge distillation[J]. Pattern Recognition and Artificial Intelligence, 2021, 34(12): 1143-1151., articleTitle=Low-bit quantization of neural network based on exponential moving average knowledge distillation, refAbstract=null)], funds=[Fund(id=1177979131519190003, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, awardId=2023-重大-02, language=CN, fundingSource=中国中铁股份有限公司科技研究开发计划课题(2023-重大-02), fundOrder=null, country=null), Fund(id=1177979131590493172, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, awardId=GS.2023-01, language=CN, fundingSource=中铁大桥局集团有限公司科学技术研究与开发课题(GS.2023-01), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1177979127144530857, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=1, ext=[AuthorCompanyExt(id=1177979127152919466, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127144530857, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 School of Civil, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China), AuthorCompanyExt(id=1177979127161308075, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127144530857, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 湖北工业大学土木建筑与环境学院, 武汉 430068)]), AuthorCompany(id=1177979127224222636, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=2, ext=[AuthorCompanyExt(id=1177979127232611245, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127224222636, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 National Key Laboratory of Intelligent and Green Bridge Construction, Wuhan 430050, China), AuthorCompanyExt(id=1177979127240999854, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127224222636, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 桥梁智能与绿色建造全国重点实验室, 武汉 430050)]), AuthorCompany(id=1177979127291331503, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, xref=3, ext=[AuthorCompanyExt(id=1177979127295525808, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127291331503, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 China Railway Bridge Science Research Institute Ltd. , Wuhan 430034, China), AuthorCompanyExt(id=1177979127303914417, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, companyId=1177979127291331503, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3 中铁大桥科学研究院有限公司, 武汉 430034)])], figs=[ArticleFig(id=1177979129430426583, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.1, caption=Polishing robot main skeleton design, figureFileSmall=+5ix8hWKGL1t2aAm7zilPQ==, figureFileBig=xTHkUDuWBIoa+ALhctjndQ==, tableContent=null), ArticleFig(id=1177979129489146840, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图1, caption=打磨机器人主骨架设计图, figureFileSmall=+5ix8hWKGL1t2aAm7zilPQ==, figureFileBig=xTHkUDuWBIoa+ALhctjndQ==, tableContent=null), ArticleFig(id=1177979129556255705, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.2, caption=The main skeleton of the polishing robot, figureFileSmall=XZbIjj+wo6jvpjVuxZLp3g==, figureFileBig=AeAv7nUbE3Fk95UOGZJWWA==, tableContent=null), ArticleFig(id=1177979129619170266, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图2, caption=打磨机器人主骨架实物图, figureFileSmall=XZbIjj+wo6jvpjVuxZLp3g==, figureFileBig=AeAv7nUbE3Fk95UOGZJWWA==, tableContent=null), ArticleFig(id=1177979129690473435, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.3, caption=Roller adjustment jacking structure, figureFileSmall=qbMQZPD9mjlx4cug8uJjmg==, figureFileBig=zfKUJAB79y0AyI/zkTOSzA==, tableContent=null), ArticleFig(id=1177979129765970908, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图3, caption=托辊调节顶升结构, figureFileSmall=qbMQZPD9mjlx4cug8uJjmg==, figureFileBig=zfKUJAB79y0AyI/zkTOSzA==, tableContent=null), ArticleFig(id=1177979129858245597, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.4, caption=Adaptive polishing structure, figureFileSmall=AehuW+QClMK9i6Qpo5DJKQ==, figureFileBig=/zEVWL16QxBMZraUS+wm2Q==, tableContent=null), ArticleFig(id=1177979129937937374, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图4, caption=自适应打磨结构, figureFileSmall=AehuW+QClMK9i6Qpo5DJKQ==, figureFileBig=/zEVWL16QxBMZraUS+wm2Q==, tableContent=null), ArticleFig(id=1177979130026017759, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.5, caption=Control flow chart, figureFileSmall=C3kKfFGvmytCC9j6kxZWwQ==, figureFileBig=GyPTRBuulzTFlK81NqqWCw==, tableContent=null), ArticleFig(id=1177979130093126624, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图5, caption=控制流程图, figureFileSmall=C3kKfFGvmytCC9j6kxZWwQ==, figureFileBig=GyPTRBuulzTFlK81NqqWCw==, tableContent=null), ArticleFig(id=1177979130156041185, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.6, caption=Improved YOLOv8 structure diagram, figureFileSmall=bkj2Wy51j+uiVuimre8Tag==, figureFileBig=wXByuNEzOAVsMePPqDjmzQ==, tableContent=null), ArticleFig(id=1177979130214761442, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图6, caption=改进后YOLOv8结构图

CBS为由Conv(卷积层)+BN(batch normalization)为批量归一化+Silu激活函数组成的模块;C2F为由多个CBS+Split(切片)+CONCAT组成的模块;CONCAT为拼接;SPDConv为细粒化卷积模块;SPPF(spatial pyramid pooling fast)为快速空间金字塔池化;DySample为动态上采样模块;CSP-OmniKernel为融合CSP思想的OmniKernel模块

, figureFileSmall=bkj2Wy51j+uiVuimre8Tag==, figureFileBig=wXByuNEzOAVsMePPqDjmzQ==, tableContent=null), ArticleFig(id=1177979130286064611, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.7, caption=DySample structure, figureFileSmall=ux6Mmt/0H+phbmDdU86jSw==, figureFileBig=1ZPjIehRRyUHqEidtlypcA==, tableContent=null), ArticleFig(id=1177979130348979172, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图7, caption=DySample结构图, figureFileSmall=ux6Mmt/0H+phbmDdU86jSw==, figureFileBig=1ZPjIehRRyUHqEidtlypcA==, tableContent=null), ArticleFig(id=1177979130403505125, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.8, caption=Comparison of detection effects before and after improvement, figureFileSmall=D7rG7eoTUXGDxcbf1or4gg==, figureFileBig=vNeGb7Y/zVd8QjD25MLvIw==, tableContent=null), ArticleFig(id=1177979130487391206, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图8, caption=改进前后检测效果对比, figureFileSmall=D7rG7eoTUXGDxcbf1or4gg==, figureFileBig=vNeGb7Y/zVd8QjD25MLvIw==, tableContent=null), ArticleFig(id=1177979130537722855, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.9, caption=Visual inspection system deployment, figureFileSmall=4nqqVhKgFAQ/HwT5WrHpgQ==, figureFileBig=b1VSytQab4fUbKWs30mptw==, tableContent=null), ArticleFig(id=1177979130600637416, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图9, caption=视觉检测系统部署, figureFileSmall=4nqqVhKgFAQ/HwT5WrHpgQ==, figureFileBig=b1VSytQab4fUbKWs30mptw==, tableContent=null), ArticleFig(id=1177979130680329193, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Fig.10, caption=Polishing robot field test results, figureFileSmall=Gsblb7slazFW49yIu1M6JQ==, figureFileBig=CiGGQMKucKFxR2yYGmwsew==, tableContent=null), ArticleFig(id=1177979130747438058, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=图10, caption=打磨机器人实地测试效果, figureFileSmall=Gsblb7slazFW49yIu1M6JQ==, figureFileBig=CiGGQMKucKFxR2yYGmwsew==, tableContent=null), ArticleFig(id=1177979130797769707, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Table 1, caption=

Main technical parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
参数 设置
工作状态 A3室内T梁模板
整机行走速度(变频)/(m·min-1) 0~6.5
行走功率(四轮驱动)/kW 2×0.55
磨筒转速/(r·min-1) 200
磨筒功率/kW 4×1.5
除吸尘吸嘴数量/个 2×7
除吸尘风量/(m3·s-1) 1
喷涂油喷嘴数量/个 2×7
喷涂油喷幅/mm 400
电压/V 380
动力条件 电源电流类型 3 AC(三相交流电)
交流电频率/Hz 50
整机装机总容量/kW 15
整机自重/t 2
), ArticleFig(id=1177979130869072876, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=表1, caption=

主要技术参数

, figureFileSmall=null, figureFileBig=null, tableContent=
参数 设置
工作状态 A3室内T梁模板
整机行走速度(变频)/(m·min-1) 0~6.5
行走功率(四轮驱动)/kW 2×0.55
磨筒转速/(r·min-1) 200
磨筒功率/kW 4×1.5
除吸尘吸嘴数量/个 2×7
除吸尘风量/(m3·s-1) 1
喷涂油喷嘴数量/个 2×7
喷涂油喷幅/mm 400
电压/V 380
动力条件 电源电流类型 3 AC(三相交流电)
交流电频率/Hz 50
整机装机总容量/kW 15
整机自重/t 2
), ArticleFig(id=1177979130957153261, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Table 2, caption=

Results of ablation experiments

, figureFileSmall=null, figureFileBig=null, tableContent=
序号 P/% R/% mAP50/% mAP50-90/%
1 79.2 71.9 83.6 58.4
2 81.2 75.2 84.9 61.7
3 78.3 77.6 85.2 62.6
4 81.4 79.4 85.6 60.1
5 82.5 75.7 85.5 62.7
6 82.3 81.2 86.8 64.7
), ArticleFig(id=1177979131028456430, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=表2, caption=

消融实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
序号 P/% R/% mAP50/% mAP50-90/%
1 79.2 71.9 83.6 58.4
2 81.2 75.2 84.9 61.7
3 78.3 77.6 85.2 62.6
4 81.4 79.4 85.6 60.1
5 82.5 75.7 85.5 62.7
6 82.3 81.2 86.8 64.7
), ArticleFig(id=1177979131099759599, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Table 3, caption=

Model comparison results

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 P/% R/% mAP50/% mAP50-90/%
YOLOv3 75.2 79.7 83.3 59.2
YOLOv5n 79.8 74.4 87.1 63.2
YOLOv6n 81.2 70.5 83.2 60.5
YOLOv8n 79.2 71.9 83.6 58.4
本文算法 82.3 81.2 86.8 64.7
), ArticleFig(id=1177979131196228592, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=表3, caption=

算法对比结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型 P/% R/% mAP50/% mAP50-90/%
YOLOv3 75.2 79.7 83.3 59.2
YOLOv5n 79.8 74.4 87.1 63.2
YOLOv6n 81.2 70.5 83.2 60.5
YOLOv8n 79.2 71.9 83.6 58.4
本文算法 82.3 81.2 86.8 64.7
), ArticleFig(id=1177979131301086193, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=EN, label=Table 4, caption=

Comparative results of modes of operation

, figureFileSmall=null, figureFileBig=null, tableContent=
作业方式 作业人数 时间/
(h·片-1)
人力成本/
(元·d-1)
日产能/片
传统人工 4 2 1 600 7
打磨机器人 1 0.25~0.5 400 10
), ArticleFig(id=1177979131397555186, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1149768571858895360, language=CN, label=表4, caption=

作业方式对比结果

, figureFileSmall=null, figureFileBig=null, tableContent=
作业方式 作业人数 时间/
(h·片-1)
人力成本/
(元·d-1)
日产能/片
传统人工 4 2 1 600 7
打磨机器人 1 0.25~0.5 400 10
)], attaches=null, journal=Journal(id=1146119176004939786, delFlag=0, nameCn=科学技术与工程, nameEn=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, issn=1671-1815, eissn=, cn=11-4688/T, coden=null, periodic=4, language=CN, oaType=是, 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=UKU/O7GSka5polgCTkbIIw==, journalPrice=null, startedYear=null, abbrevIsoEn=Sci Technol Eng, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1754445529766, createdBy=null, updatedBy=13701087609, firstLetterCn=S, firstLetterEn=S, subjectCode=Natural Sciences, subjectName=自然科学, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=UKU/O7GSka5polgCTkbIIw==, picEn=5hwlULoNwcbj3xUmVi9MAQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1159791870395564357, language=CN, name=科学技术与工程, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.stae.com.cn/jsygc/home, createdTime=1754445529793, updatedTime=1754445529793, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.stae.com.cn/jsygc/site/menus/20090429150146001, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1159791870441701702, language=EN, name=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.stae.com.cn/jsygc/home, createdTime=1754445529804, updatedTime=1754445529804, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1146123166801305609, websiteList=[Website(id=1148243202391400884, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, 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/kxjsygc/CN, language=CN, createTime=1751692112777, createBy=18614031015, updateTime=1753520965431, updateBy=18614031015, name=科学技术与工程-中文站点, tplId=1146099689490845704, title=科学技术与工程, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148622798802673703, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=articleTextType, value=kx, createTime=1751782615614, updateTime=1751782615614, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798781702180, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=banner, value=null, createTime=1751782615609, updateTime=1751782615609, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798769119267, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1751782615606, updateTime=1751782615606, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798794285094, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751782615612, updateTime=1751782615612, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798790090789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751782615611, updateTime=1751782615611, creator=18614031015, updator=18614031015)]), Website(id=1155914124811976731, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, 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/kxjsygc/EN, language=EN, createTime=1753521003206, createBy=18614031015, updateTime=1753521003206, updateBy=18614031015, name=科学技术与工程-英文站点, tplId=1146101810881728533, title=Science Technology and Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155914371227308235, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=articleTextType, value=kx, createTime=1753521061952, updateTime=1753521061952, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371210531016, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=banner, value=null, createTime=1753521061947, updateTime=1753521061947, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371202142407, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1753521061945, updateTime=1753521061945, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371223113930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753521061950, updateTime=1753521061950, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371218919625, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753521061949, updateTime=1753521061949, creator=18614031015, updator=18614031015)])], journalTitle=科学技术与工程, weixinUrl=null, journalUrl=null, iacademicId=null, status=0, seqNo=null, journalTitleEn=Science Technology and Engineering, journalPhotoCn=UKU/O7GSka5polgCTkbIIw==, journalPhotoEn=5hwlULoNwcbj3xUmVi9MAQ==, journalFirstLetter=S, 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=null, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2406880, detailUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/10.12404/j.issn.1671-1815.2406880, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/PDF/10.12404/j.issn.1671-1815.2406880, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/PDF/10.12404/j.issn.1671-1815.2406880, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于深度学习的T梁模板打磨机器人设计
收藏切换
PDF下载
陈映全 1 , 石峻峰 1, * , 翟磊 2 , 刘郴 3 , 李惠东 2
科学技术与工程 | 论文·机械、仪表工业 2025,25(16): 6733-6741
收起
收藏切换
科学技术与工程 | 论文·机械、仪表工业 2025, 25(16): 6733-6741
基于深度学习的T梁模板打磨机器人设计
全屏
陈映全1 , 石峻峰1, * , 翟磊2, 刘郴3, 李惠东2
作者信息
  • 1 湖北工业大学土木建筑与环境学院, 武汉 430068
  • 2 桥梁智能与绿色建造全国重点实验室, 武汉 430050
  • 3 中铁大桥科学研究院有限公司, 武汉 430034
  • 陈映全(2000—),男,汉族,湖北黄石人,硕士研究生。研究方向:智能建造、桥梁智能建造装备研发。E-mail:

通讯作者:

* 石峻峰(1974—),男,汉族,吉林辽源人,博士,讲师。研究方向:智能建造、工程结构BIM技术。E-mail:
Design of T-beam Formwork Polishing Robot Based on Deep Learning
Ying-quan CHEN1 , Jun-feng SHI1, * , Lei ZHAI2, Chen LIU3, Hui-dong LI2
Affiliations
  • 1 School of Civil, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China
  • 2 National Key Laboratory of Intelligent and Green Bridge Construction, Wuhan 430050, China
  • 3 China Railway Bridge Science Research Institute Ltd. , Wuhan 430034, China
出版时间: 2025-06-08 doi: 10.12404/j.issn.1671-1815.2406880
文章导航
收藏切换

针对变截面T梁模板打磨工作难度大,费时长等问题,设计一种基于深度学习的T梁模板打磨机器人。首先提出一种自适应打磨结构,解决了现有打磨装置无法贴合T梁模板内侧变截面,打磨辊容易卡在T梁模板隔断内的技术问题;其次为实现打磨质量的量化监测,提出一种YOLOv8n-DSE算法,对模板上的混凝土痂块及污渍进行识别,引入DySample动态上采样模块,增强模型的抗干扰能力并加快计算速度;为提高小目标检测的精度,设计SOEP(small object enhance pyramid)模块,通过SPDConv(space to depth convolution)获得小目标信息特征给到CSP(cross stage partial)-Omni-Kernel进行特征整合提高小目标的检测检测性能;最后替换EMA(exponential moving average)-Slide Loss让模型更加关注困难目标,可以改善在难例检测上的效果。准确率、召回率和均值平均精度(mean average precision,mAP)值相较改进前分别提升了3.1%,9.7%和3.2%,将改进的模型部署至机器人并进行现场测试。结果表明:设备满足工厂对变截面T梁模板的打磨需求。

变截面T梁模板  /  深度学习  /  打磨机器人  /  自适应打磨  /  YOLOv8n

A deep learning based T-beam formwork polishing robot was designed for the problems of difficult and time consuming polishing of T-beam formwork for variable cross-section. Firstly, an adaptive polishing structure was proposed to solve the technical problem that the existing polishing device cannot fit the inner variable cross-section of the T-beam formwork, and the polishing roller was easy to get stuck in the T-beam formwork partition. Secondly, in order to realize the quantitative monitoring of the polishing quality, a YOLOv8n-DSE algorithm was proposed to identify concrete dirt and stains on the formwork, the DySample dynamic up-sampling module was introduced to enhance the anti-interference ability of the model and accelerate the calculation speed, to improve the accuracy of small target detection, the SOEP (small object enhance pyramid) module was designed to improve the detection performance of small target detection through the SPDConv(space to depth convolution) to obtain the information features of the small target and give them to the CSP(cross stage partial)-Omni-Kernel for the integration of the features. Finally, the EMA(exponential moving average)-SlideLoss was replaced to make the model more concerned with the quantitative monitoring of the concrete, allowed the model to focus more on difficult targets, which can improve the effect on difficult case detection. The accuracy, recall, and mAP(mean average precision) values are improved by 3.1%, 9.7%, and 3.2%, respectively, compared with those before the improvement. The improved model was deployed to the robot and tested in the field. The results show that the equipment meets the plant's needs for polishing variable-section T-beam formwork.

T-beam formwork for variable cross-section  /  deep learning  /  polishing robot  /  adaptive sanding  /  YOLOv8n
陈映全, 石峻峰, 翟磊, 刘郴, 李惠东. 基于深度学习的T梁模板打磨机器人设计. 科学技术与工程, 2025 , 25 (16) : 6733 -6741 . DOI: 10.12404/j.issn.1671-1815.2406880
Ying-quan CHEN, Jun-feng SHI, Lei ZHAI, Chen LIU, Hui-dong LI. Design of T-beam Formwork Polishing Robot Based on Deep Learning[J]. Science Technology and Engineering, 2025 , 25 (16) : 6733 -6741 . DOI: 10.12404/j.issn.1671-1815.2406880
当前,中国城市交通基础设施更新持续推进[1]。以城市高架桥梁建设为例,传统高架桥梁采用现浇作业施工,存在高污染、对城市干扰大、劳动密集、工效低、标准化程度低等缺点。为了尽可能减少桥梁建设对周边环境及交通通行的影响,结合先进智能传感和数字化技术,城市桥梁建设正在向全预制装配式结构[2]和快速施工技术方向[3]发展。桥梁全预制装配式构件的生产采用工厂流水化作业模式,在预制梁场利用待预制构件模板开展构件的大批量生产。以公路常用的T梁构件为例,当T梁在预制完成后,其模板常会残留混凝土等杂物,这需要打磨清理后才能进行下一片T梁的生产。现阶段打磨清理主要由人工完成,但由于模板表面积约为200 m2,工作量很大,需要消耗大量的人力成本和时间成本,影响梁场的生产效率。同时国家知识产权局官网上公开了一种铁路预制箱梁的钢模板用打磨机[4],该打磨机沿箱梁端面形状布置有数台打磨托辊组件,通过打磨机前后行走,配合托辊旋转打磨实现对箱梁模板上的清理。由于铁路箱梁沿纵桥方向是等截面形状,所以该打磨机只能适应等截面箱梁的打磨作业。
公路T梁的马蹄部位在设计时为了更好保证侧向刚度,增强两端抗弯能力,其梁肋一般为中间窄,顶底部宽,所以导致公路T梁模板截面形状不是等截面。现有技术无法适应变截面的公路T梁模板内侧壁的打磨作业。同时T梁梁肋一般还设置有数个横隔板,故T梁模板在相应位置处存有隔断,现有技术的打磨托辊在T梁模板打磨过程中,会出现因打磨托辊陷入隔断而发生卡住现象。
在实现打磨功能的基础上,通过简单测试发现,T梁模板上的混凝土块大小不一且顽固程度不同在打磨过程中可能会出现打磨不完全,残留混凝土等问题。而若无法对打磨进行质量检查,直接进行T梁浇筑易造成对预制T梁质量产生影响。由此可见,基础的功能已无法满足实际需要。为了克服现有技术的不足之处,针对残留混凝土的特点,设计一种基于YOLOv8[5]的混凝土残留检测算法,即在机器人各作业面加装相机,对质量实时监测,并对获取的图像进行实时处理,根据检测到的残留混凝土的综合情况调整打磨机器人的作业。设计一种基于深度学习的适用于变截面公路T梁模板的多功能作业设备,能够将打磨托辊组件始终紧贴在待打磨T梁模板表面,保证T梁模板表面残留混凝土可有效去除干净,显著降低打磨作业时间,提高预制T梁生产效率;同时设计一种自适应打磨结构,在打磨托辊碰到T梁模板隔断时,可对其进行锁紧,保证打磨托辊不会陷入T梁模板隔断中。通过设计一款用于预制梁场T梁模板打磨除尘喷涂油的多功能智能设备,可以解决预制梁场生产T梁时打磨模板耗时长,工作强度大的问题。
打磨吸尘涂油机器人是用于预制梁场T梁模板集打、吸尘喷涂一体化的施工装备,极大减轻了工人的劳动强度、提升了工作效率,表1为打磨机器人主要技术参数。
图1所示,主骨架为打磨吸尘涂油机器人主要载体,通过下部驱动支撑主骨架在T梁底模板上作往复直运动,同时磨筒电机驱动主骨架前端的磨筒转动打磨模板断面,吸尘器通附着与主骨架中间的一圈吸嘴吸走当前打磨产生的粉尘:打磨完成后,通过检测开关使磨筒旋刷驱动及走行驱动停止;启动电动推杆使筒与模板脱开然后走行电机驱动主骨架在T梁底模板上反向走行:空压机启动,通过附与主骨架后端的一圈喷嘴将脱模剂均匀喷涂在模板上。
为了实现T梁模板表面打磨全覆盖,本设备桁架采用沿T梁模板纵向方向上截面形状为T形的结构形式。如图2所示,主骨架为行走、打磨、吸尘、喷涂等所有执行机构的载体,其整体为矩形框架结构,主要采用方形钢管等轻钢结构焊接而成;主骨架上下均设置有平台,可放置电控柜、空压机等机电件,同时也可作为设备检修平台。主骨架底部有4根支撑脚可使其在地面站立。
行走驱动系统包括行走电机、减速箱、单边轨道轮、扫轨器及轮轴等构件。扫轨器布置在轨道轮前端,可以扫除T梁模板底模上的杂物,设备走行驱动系统安装固定在设备桁架底端,实现设备沿T梁模板底模纵向方向往复走行;行走驱动设置在主骨架下方,采用四轮驱动,电机可变频调速,保证整机行走平稳、无冲击。“三合一”减速机安装在主骨架下方的前、后端,驱动车轮在底模上行走,可实现在检测出打磨不完全的情况下主骨架在地模板上做往复直线运动。行走电机布局已考虑喷涂脱模剂对其的影响,将其中一侧布置在主骨架的内侧。
托辊打磨系统包括顶部托辊打磨组件、两侧上端托辊打磨组件、两侧下端托辊打磨组件及底部托辊打磨组件;顶部托辊打磨组件设置在设备桁架顶端,负责打磨T梁模板上端表面,其由两组打磨托辊、传动轴及驱动电机等组成;两侧上端托辊打磨组件设置在设备桁架两侧,负责打磨T梁模板两侧表面,其由两组打磨托辊、驱动电机等组成;两侧下端托辊打磨组件设置在设备桁架两侧,负责打磨T梁模板两侧马蹄部位表面,其由两组打磨托辊、驱动电机等组成;底部托辊打磨组件设置在设备桁架底端,负责打磨T梁模板底模表面,其由一组打磨托辊、驱动电机等组成;托辊调节锁定系统包括顶底部托辊调节锁定系统和两侧托辊调节锁定系统;顶底部托辊调节锁定系统包括提升门架、导向轮、电动推杆等;图3为托辊调节顶升结构,顶部托辊打磨组件和底部托辊打磨组件均安装固定在提升门架之上,电动推杆负责提升门架上下动作,使得顶部托辊打磨组件和底部托辊打磨组件对T梁模板待打磨表面正压力的自动调节。
为了自适应T梁模板横向距离变化,图4所示两侧托辊调节锁定系统包括电动推杆、钢丝绳、钢丝绳转向导轮、托辊旋转支架、配重块、电磁吸力装置等。配重块通过钢丝绳带动托辊旋转支架旋转,在设备行走过程中,可让图两侧上端打磨托辊始终贴紧在T梁模板两侧表面对其进行打磨作业。在设备行走至T梁模板隔断时,通过行程开关触发控制信号,使电磁铁通电吸住配重块,阻止因配重块继续下落造成打磨托辊陷入T梁模板隔断内;通过设置电磁装置与配重块,可以根据设备桁架在T梁模板内的行走路径,灵活控制上侧壁打磨组件的伸出位置,以适应公路T梁模板截面的变化,同时还可避免在打磨过程中,上侧壁打磨组件陷入T梁模板隔断内;通过设置加宽的下侧壁打磨组件,可以适应T梁模板内侧壁的宽度变化,以对T梁模板内侧壁底部加宽处进行打磨工作,顶底部打磨单元可以对T梁模板的顶面和底面进行打磨工作,且可以在打磨完成后上升并远离打磨处,方便对T梁模板进行后续养护工作。上下电动推杆则可通过改变对钢丝绳的牵引力以达到调节打磨压力的功能。
图5为打磨机器人控制流程图,打磨机器人主控制器为PLC控制,采用西门子S7-200SMART系列,是整机控制系统大脑,负责采集遥控器和触摸屏操作指令、各检测安全开关状态、变频器通讯运行数据、走行编码器运行坐标,再通过内部程序运算处理,输出各机构运行命令,以实现“一键”完成多机构智能协同作。
打磨机器人人机界面采用昆仑动态10 in(1 in=2.54 cm)触摸屏,可显示运行状态、运行位置、各机构故障,设置运行速度、运行时间、打磨压力,无线遥控器可远程遥控操作各机构运行。
目标检测检测领域主要分为单阶段检测[6]和双阶段[7]检测。单阶段检测由于无需生成候选区域,具有更快的运算速度,双阶段检测在候选区域进行分类和精细化定位具有更高的精度。在本文应用场景下需要更快的检测速度所以采用单阶段检测。
YOLOv8作为主流的单接段检测的第八代版本相对与其前代版本在多个方面进行了显著的改进,引入PANet(path aggregation network)模块,从原先的耦合头变成了解耦头,使用C2F模块代替C3模块等;使得其在目标检测任务中取得了更加出色的表现。
在处理小目标检测任务时,传统的P3、P4、P5检测层级略显不足。为了增强对小目标的识别能力,传统方法倾向于引入P2检测层作为补充[8],这种做法虽能带来一定的提升,却也伴随着计算负担显著增加、后处理流程耗时加剧等不利因素。为此,在经典的路径聚合特征金字塔网络(path aggregation feature pyramid network,PAFPN)[9]架构基础上进行了创新,设计出小目标增强金字塔(small object enhance pyramid,SOEP)。与直接增加P2检测层不同,本文方法利用P2特征层,通过细粒化卷积模块SPDConv[10]处理,将卷积细粒化后提取并强化蕴含小目标特征的信息,随后将这些增强后的特征融合到P3层级中。为了进一步优化特征整合过程,融合跨阶段部分连接(CSP)[11]思想和Omni-Kernel模块[12]。CSP技术有助于提升网络的学习效率与泛化能力,而Omni-Kernel则通过其独特的三分支结构(全局分支、大分支与局部分支)实现从全面到细节的多层次特征学习,这种设计使得网络能够更有效地捕捉并融合从小到大的各类特征信息,最终显著提升小目标的检测效果。
图6所示,通过CSP-Omni-Kernel模块构建针对小目标检测优化的特征金字塔结构,不仅解决了传统方法中的计算瓶颈与耗时问题,还大幅提升了小目标的检测性能。
在深度学习中,特征上采样是密集预测模型(如语义分割、目标检测等)中不可或缺的一环。这一过程旨在逐步恢复特征图的分辨率,以便更精细地捕捉图像中的细节信息。传统的上采样方法,虽然简单高效,但受限于固定的插值规则,难以适应复杂多变的图像内容。因此,引入DySample[13]。动态上采样模块通过绕过动态卷积操作,从点采样的角度重新定义了上采样过程。这种方法具有更少的参数、浮点数运算(FLOPs)、GPU内存占用和延迟,通过优化采样策略和减少不必要的计算开销,实现了显著的性能提升。通过动态调整采样点的位置来执行上采样,而不是依赖于静态的卷积核。初始采样点的位置分布和偏移量的范围被精心设计,以更好地捕捉输入特征的空间关系,同时减少重叠和混乱。为了减少偏移量的重叠和混乱,在局部范围内约束偏移量的范围,从而提高了采样点的分布质量。
图7(a)为基于动态范围因子的点采样器。在点采样器内首先获得比例系数sC×H×W(CHW分别为特征图的通道数、宽、高)的特征图X,通过输入C与输出2gs2的线性层后,其中2g代表xy的坐标,再通过Sigmoid函数,其表达式为
S(x)= 1 1 + e - x
此时的特征图与未通过Sigmoid函数的相乘生成一个2gs2×H×W的特征图,最后通过像素重组变换成2g×sH×sW的偏移量Δ,然后与原始采样网格O求和得到点采样器的输出D
D=O+Δ
图7(b)为Dysample动态上采样的流程图,最后通过grid_sample(双线性插值)对输入特征图进行可变形采样,将点采样器的输出与原特征图X重新采样为X',可表示为
X'=grid_sample(X,D)
YOLOv8模型在迭代的过程中使用CIOU来评估模型预测值与实际值之间的差异程度。在迭代过程中,模型的目标是寻找能够最小化损失函数的参数。损失函数用于指导模型的优化更新方向,损失函数的选择直接影响到模型的性能和学习效率。
针对大多数情况下简单样本数量远多于困难样本的问题,引入EMA-SlideLoss[14]损失函数来改善对困难例子的检测性能。首先计算所有预测框和真实框的交并比IoU(intersection over union)值的平均值作为阈值α。IoU值小于的样本被视为负样本(简单样本),而IoU值大于α的样本被视为正样本(困难样本)。为了更好地处理边界附近的样本,对于接近阈值α的样本(即IoU值接近α的样本),权重系数可表示为
f(x)= 1 , x α - 0.1 e 1 - α , α < x < α - 0.1 e 1 - x , x α
当训练图像IoU值越接近α时,权重越大,使困难样本在训练过程中受到更多的关注,这种方法减少了对超参数的依赖,使得模型更加灵活和自适应。
在slideloss中引入指数移动平均[15](exponential moving average,EMA),可定义为
αi=βi_iou+αi-1(1-β)
式(5)中:β∈(0,1)为指数移动平均系数,用来调整α迭代的权重分配;i_iou为在i时刻时的IoU值。
使用EMA指数移动平均进行参数更新,可以减少参数更新的噪声和波动,使模型在训练过程中更加稳定地收敛,并提高模型的泛化能力。同时,EMA还可以通过平滑模型参数来减少预测结果的波动,提高预测的稳定性和准确性。
数据集为自制数据集,采集于湖北省智慧梁场,拍摄T梁模板打磨前与打磨后的图像,拍摄设备为智能手机,在多种光线条件下拍摄,图像格式为JPG,经过简单筛选后选择600张图像。在训练模型之前,需要对采集的图像进行标注处理,采用Labelme进行数据标注,在标注后再将json文件转换为txt文件。将转换完成的图像数据按8∶1∶1的比例划分为训练集(480张)、验证集(60张)、和测试集(60张) 。
本实验运行的操作系统为windows, CPU为 Intel©i7-12700H,GPU为NVIDIA GeForce RTX3060,集成开发环境为VScode,PyTorch版本为2.3.1,加速计算框架为CUDA11.8.0,Python版本为3.8.16。训练过程中启用mosaic增强,训练轮次为500。
基于应用场景的需求,评价指标采用查全率R、查准率P、均值平均精度(mAP)标准度量指标对模型性能进行全面评估,其定义分别为
P= T P T P + F P
R= T P T P + F N
AP= 0 1 P(R)dR
mAP= 1 n i = 1 nAPi
式中:TP表示正样本识别为真;TN表示负样本识别为真;FP表示正样本识别为假;FN表示负样本识别为假;i为检测次数;AP为平均准确率;APi为第i次检测次数下的平均准确率;n为图像个数。
为了验证所提出改进算法的有效性,设置逐一增加改进模块的实验如表2所示。
为验证所改进算法在T梁模板上混凝土残留物检测方面的优越性与高效性,将改进后的算法与目前主流的YOLO算法进行了对比试验,得到PR、mAP50、mAP50-90,实验结果如表3所示。
表3可知,本文算法相对于改进前的YOLOv8n算法,R提升了9.7%,P提升了3.1%,mAP50提升了3.2%,mAP50-90提升了6.3%。相较于其他主流YOLO算法,只有YOLOv5n的mAP50较本文算法高0.3%,但其召回率低了6.8%,可知YOLOv5n在检测痂块及污渍时漏检的概率更高。由此可见,本文算法在T梁模板表面痂块及污渍检测任务中拥有更出色的检测性能。
图8为原YOLOv8n和改进算法的检测效果对比。从检测效果对比可见,当出现多个目标时,改进后的模型出现误检、漏检的次数更少,改进后的算法对目标的检测更加细致,改善了多检测框重叠的问题,并对于背景相似或者模糊不清的小目标的检测更加准确。
应公路预制T梁场生产需求,在梁场T梁模板上进行实地应用测试。首先如图9所示,将视觉检测系统装载至T梁模板打磨机器人的4个打磨工作面。 再将作业设备放置于T梁模板下端平面,使单边导轨轮与模板下端平面边缘贴合保证行走运动的稳定性,如图10所示,搭载YOLOv8n-DSE算法的视觉检测系统的T梁模板打磨机器人在实际生产场景下测试,打磨后效果满足梁场对T梁模板生产前的整备要求。
表4为经多次测试得出数据,该打磨机器人作业于T梁模板整备的效率为0.25~0.5 h/片,传统人工作业由于工作平面复杂且狭窄通常需要4人同时作业2 h,而打磨机器人只需要一人进行设备的操作,使用打磨机器人后梁场产出预制梁的效率提升约42.85%,缩短满足工程项目需要的生产周期。以本实验梁场为例,使用本打磨机器人后在一年的时间里可为企业节约人工成本及各项开支约50万元。
实验结果表明,T梁模板打磨机器人各项指标均满足企业实际生产的需求,提高了企业产能,减轻了作业人员的工作强度,缓解企业用人压力,同时推动了预制梁生产流程的自动化和智能化。
(1)针对变截面T梁模板打磨工作时打磨托辊容易陷入隔断的问题,设计一种变截面自适应打磨结构完成对4个复杂工作面的打磨,提升企业生产效率,降低T梁生产成本。但在泛用性上仍需持续改进,以满足其他不同尺寸T梁模板时的打磨需求。
(2)提出一种基于YOLOv8n的YOLOv8-DSE算法,对残余的混凝土痂块及污渍进行识别,在YOLOv8n的基础上引入改进小目标检测层,将富含小目标信息的特征给到检测头,提升对小目标的检测效果;应用Dysample动态上采样,增强模型的抗干扰性能,提升目标检测性能;替换EMA-SlideLoss,使模型在训练过程中更加关注困难样本,EMA提高损失值更新过程中的稳定性。
(3)实现了在打磨工作中对机器人工作表面实时监测,并对获取的图像进行实时处理,根据检测到的残留混凝土的综合情况实时调整打磨的力度及时间,并形成了打磨质量反馈的闭环控制体系。
  • 中国中铁股份有限公司科技研究开发计划课题(2023-重大-02)
  • 中铁大桥局集团有限公司科学技术研究与开发课题(GS.2023-01)
参考文献 引证文献
排序方式:
[1]
李兰冰, 高雪莲, 黄玖立. “十四五”时期中国新型城镇化发展重大问题展望[J]. 管理世界, 2020, 36(11): 7-22.
Li Lanbing, Gao Xuelian, Huang Jiuli. Prospects for major issues of China's new urbanization development during the “14th Five-Year Plan” period[J]. Management World, 2020, 36(11): 7-22.
[2]
Zhang R, Guo T, Li A. Experimental investigation into demountable dry connections for fully precast frame structures through shaking table tests[J]. Thin-Walled Structures, 2024, 201: 112014.
[3]
查道宏, 李军堂, 胡勇. 大跨度斜拉桥钢筋混凝土桥塔快速施工技术发展与展望[J]. 世界桥梁, 2022, 50(4): 32-40.
Cha Daohong, Li Juntang, Hu Yong. Development and prospects of rapid construction technology for reinforced concrete pylons of long-span cable-stayed bridges[J]. World Bridges, 2022, 50(4): 32-40.
[4]
李勇海, 余秀平, 刘宁, 等. 铁路预制箱梁的钢模板用打磨机: CN202121692391.X[P]. 2021-12-21.
Li Yonghai, Yu Xiuping, Liu Ning, et al. Grinding machine for steel mould plates for prefabricated box girders for railways: CN202121692391.X[P]. 2021-12-21.
[5]
Li Y, Fan Q, Huang H, et al. A modified YOLOv8 detection network for UAV aerial image recognition[J]. Drones, 2023, 7(5):304.
[6]
蒲玲玲, 杨柳. 改进YOLOv5的多车辆目标实时检测及跟踪算法[J]. 科学技术与工程, 2023, 23(28): 12159-12167.
Pu Lingling, Yang Liu. Improved real-time detection and tracking algorithm for multi vehicle targets in YOLOv5[J]. Science Technology and Engineering, 2023, 23(28): 12159-12167.
[7]
张世强, 史卫亚, 张绍文, 等. 基于改进YOLOv5算法的钢铁表面缺陷检测[J]. 科学技术与工程, 2023, 23(35): 15148-15157.
Zhang Shiqiang, Shi Weiya, Zhang Shaowen, et al. Steel surfacedefect detection based on improved YOLOv5 algorithm[J]. Science Technology and Engineering, 2023, 23(35): 15148-15157.
[8]
王俊岭, 王晨晨, 熊玉华. 基于YOLOv5l-Im的排水管道缺陷检测方法及效果分析[J]. 科学技术与工程, 2024, 24(18): 7833-7842.
Wang Junling, Wang Chenchen, Xiong Yuhua. Drainage pipe defect detection and effect based on the improved YOLOv5l-Im[J]. Science Technology and Engineering, 2024, 24(18): 7833-7842.
[9]
Liu S, Qi L, Qin H, et al. Path aggregation network for instance segmentation[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 8759-8768.
[10]
Sunkara R, Luo T. No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects[C]// Joint European Conference on Machine Learning and Knowledge Discovery in databases. Cham: Springer Nature Switzerland, 2022: 443-459.
[11]
Wang C Y, Yuan H, Wu Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Seattle: IEEE, 2020: 1571-1580.
[12]
Cui Y N, Ren W Q, Alois K. Omni-kernel network for image restoration[C]// Proceedings of the AAAI Conference on Artificial Intelligence. Vancouver: AAAI, 2024: 1426-1434.
[13]
Liu W, Lu H, Fu H, et al. Learning to upsample by learning to sample[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Paris: IEEE, 2023: 6027-6037.
[14]
Yu Z, Huang H, Chen W, et al. YOLO-facev2: a scale and occlusion aware face detector[J]. Pattern Recognition, 2024, 155: 110714.
[15]
吕君环, 许柯, 王东. 基于指数移动平均知识蒸馏的神经网络低比特量化方法[J]. 模式识别与人工智能, 2021, 34(12): 1143-1151.
Junhuan, Xu Ke, Wang Dong. Low-bit quantization of neural network based on exponential moving average knowledge distillation[J]. Pattern Recognition and Artificial Intelligence, 2021, 34(12): 1143-1151.
2025年第25卷第16期
PDF下载
323
122
引用本文
BibTeX
文章信息
doi: 10.12404/j.issn.1671-1815.2406880
  • 接收时间:2024-09-13
  • 首发时间:2025-07-09
  • 出版时间:2025-06-08
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-09-13
  • 修回日期:2025-03-20
基金
中国中铁股份有限公司科技研究开发计划课题(2023-重大-02)
中铁大桥局集团有限公司科学技术研究与开发课题(GS.2023-01)
作者信息
    1 湖北工业大学土木建筑与环境学院, 武汉 430068
    2 桥梁智能与绿色建造全国重点实验室, 武汉 430050
    3 中铁大桥科学研究院有限公司, 武汉 430034

通讯作者:

* 石峻峰(1974—),男,汉族,吉林辽源人,博士,讲师。研究方向:智能建造、工程结构BIM技术。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2406880
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