Article(id=1327011059782758978, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1327010748263411889, articleNumber=null, orderNo=null, doi=10.7544/issn1000-1239.202440034, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1705507200000, receivedDateStr=2024-01-18, revisedDate=1723392000000, revisedDateStr=2024-08-12, acceptedDate=1725292800000, acceptedDateStr=2024-09-03, onlineDate=1794313691325, onlineDateStr=2026-11-10, pubDate=1743436800000, pubDateStr=2025-04-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1794313691325, onlineIssueDateStr=2026-11-10, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1794313691325, creator=13701087609, updateTime=1794313691325, updator=13701087609, issue=Issue{id=1327010748263411889, tenantId=1146029695717560320, journalId=1324759914146529478, year='2025', volume='62', issue='4', pageStart='821', pageEnd='1089', issueExtLink='null', onlineDate='null', pubDate='1743436800000', pubDateStr='2025-04-01', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1794313617052, creator='13701087609', updateTime=1794376751260, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1327275552727217083, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1327010748263411889, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1327275552727217084, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1327010748263411889, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=950, endPage=962, ext={EN=ArticleExt(id=1327011060059583044, articleId=1327011059782758978, tenantId=1146029695717560320, journalId=1324759914146529478, language=EN, title=Control Logic Routing for Continuous-Flow Microfluidic Biochips Based on Deep Reinforcement Learning, columnId=1327010913338634647, journalTitle=Journal of Computer Research and Development, columnName=Architecture, runingTitle=null, highlight=null, articleAbstract=

With the advancement of electronic design automation, continuous-flow microfluidic biochips have become one of the most promising platforms for biochemical experiments. This chip manipulates fluid samples in milliliters or nanoliters by utilizing internal microvalves and microchannels, and thus automatically performs basic biochemical experiments, such as mixing and detection. To achieve the correct bioassay function, the microvalves deployed inside the chip are usually managed by a multiplexer-based control logic, and valves receive control signals from a core input through the control channel for accurate switching. Since biochemical reactions typically require high sensitivity, the length of control paths connecting each valve needs to be reduced to ensure immediate signal propagation, and thus to reduce the signal propagation delay. In addition, to reduce the fabrication cost of chips, a vital issue to be addressed in the logic architecture design is how to effectively reduce the total channel length within the control logic. To address the above issues, we propose a deep reinforcement learning-based control logic routing algorithm to minimize the signal propagation delay and total control channel length, thereby automatically constructing an efficient control channel network. The algorithm employs the dueling deep Q-network architecture as the agent of the deep reinforcement learning framework to evaluate the tradeoff between signal propagation delay and total channel length. Besides, the diagonal channel routing is implemented for the first time for control logic, thus fundamentally improving the efficiency of valve switching operations and reducing the fabrication cost of the chip. The experimental results demonstrate that the proposed algorithm can effectively construct a high-performance and low-cost control logic architecture.

, authors=Huayang Cai1, 2, 3, Xing Huang4, Genggeng Liu1, 2, 3, *, authorsList=Huayang Cai, Xing Huang, Genggeng Liu, authorCompany=null, correspAuthors=Genggeng Liu, authorNote=

born in 1997. PhD candidate. His main research interest includes design automation of microfluidic biochips

born in 1991. PhD, professor, PhD supervisor. His main research interest includes design automation for microfluidic biochips and integrated circuits

born in 1988. PhD, professor, PhD supervisor. Senior member of CCF. His main research interest includes design automation for microfluidic biochips and integrated circuits

, correspAuthorsNote=null, copyrightStatement=Copyright © 2025 Journal of Computer Research and Development. 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=1327011064358744667, articleId=1327011059782758978, tenantId=1146029695717560320, journalId=1324759914146529478, language=CN, title=基于深度强化学习的连续微流控生物芯片控制逻辑布线, columnId=1325489904427344387, journalTitle=计算机研究与发展, columnName=体系结构, runingTitle=null, highlight=null, articleAbstract=

随着电子设计自动化技术的迅速发展,连续微流控生物芯片成为了目前最具前景的生化实验平台之一. 该芯片通过采用内部的微阀门以及微通道来操纵体积仅为毫升或纳升的流体样品,从而自动执行混合和检测等基本的生化实验操作. 为了实现正确的生化测定功能,部署于芯片内部的微阀门通常需要由基于多路复用器的控制逻辑进行管控,其通过控制通道获得来自核心输入的控制信号以实现精确切换. 由于生化反应通常需要非常高的灵敏度,因此为了保证信号的即时传输,需要尽可能地减少连接每个阀门的控制路径长度,以降低信号传输的时延. 此外,为了降低芯片的制造成本,如何有效减少控制逻辑中通道的总长度也是逻辑架构设计需要解决的关键问题之一. 针对上述问题,提出了一种基于深度强化学习的控制逻辑布线算法以最小化信号传输时延以及控制通道总长度,从而自动构建高效的控制通道网络. 该算法采用竞争深度Q网络架构作为深度强化学习框架的智能体,从而对信号传输时延和通道总长度进行权衡评估. 此外,针对控制逻辑首次实现了对角型的通道布线,从根本上提高了阀门切换操作的效率并降低了芯片的制造成本. 实验结果表明,所提出的算法能够有效构建高性能、低成本的控制逻辑架构.

, authors=蔡华洋1, 2, 3, 黄兴4, 刘耿耿1, 2, 3, *, authorsList=蔡华洋, 黄兴, 刘耿耿, authorCompany=null, correspAuthors=刘耿耿, authorNote=

1997年生. 博士研究生. 主要研究方向为微流控生物芯片的设计自动化

1991年生. 博士,教授,博士生导师. 主要研究方向为微流控生物芯片和集成电路的设计自动化

1988年生. 博士,教授,博士生导师. CCF高级会员. 主要研究方向为微流控生物芯片和集成电路的设计自动化

, correspAuthorsNote=
刘耿耿()
, copyrightStatement=版权所有 © 2025 计算机研究与发展编辑部, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=MuZbNjAblOy6lmhTslH7fw==, magXml=0QKtjpB4lJwW/GFYfgVkRA==, pdfUrl=null, pdf=/5U4FnXqOmOaI9RcFLEclQ==, pdfFileSize=1233855, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=eoHUttSHMYVO2qxY5hoBng==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=P54eD/GlAgB5QmivXzGotg==, mapNumber=null, fund=null)}, authors=[Author(id=1327011065008861801, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=c_huayang@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327011065092747885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065008861801, language=EN, stringName=Huayang Cai, firstName=Huayang, middleName=null, lastName=Cai, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116
2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116
3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116, bio={"content":"

born in 1997. PhD candidate. His main research interest includes design automation of microfluidic biochips

"}, bioImg=null, bioContent=

born in 1997. PhD candidate. His main research interest includes design automation of microfluidic biochips

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327011065168245358, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065008861801, language=CN, stringName=蔡华洋, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1福州大学计算机与大数据学院 福州 350116
2大数据智能教育部工程研究中心(福州大学) 福州 350116
3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116, bio={"img":"M6mam6vc0YyZ+QHX9r2nxQ==","content":"

1997年生. 博士研究生. 主要研究方向为微流控生物芯片的设计自动化

"}, bioImg=M6mam6vc0YyZ+QHX9r2nxQ==, bioContent=

1997年生. 博士研究生. 主要研究方向为微流控生物芯片的设计自动化

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327011064618791516, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=1, ext=[AuthorCompanyExt(id=1327011064627180125, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116), AuthorCompanyExt(id=1327011064635568734, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1福州大学计算机与大数据学院 福州 350116)]), AuthorCompany(id=1327011064732037727, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=2, ext=[AuthorCompanyExt(id=1327011064740426336, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116), AuthorCompanyExt(id=1327011064765592161, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大数据智能教育部工程研究中心(福州大学) 福州 350116)]), AuthorCompany(id=1327011064815923810, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=3, ext=[AuthorCompanyExt(id=1327011064820118115, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116), AuthorCompanyExt(id=1327011064828506724, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116)])]), Author(id=1327011065243742832, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327011065323434610, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065243742832, language=EN, stringName=Xing Huang, firstName=Xing, middleName=null, lastName=Huang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4School of Computer Science, Northwestern Polytechnical University, Xi’an 710072, bio={"content":"

born in 1991. PhD, professor, PhD supervisor. His main research interest includes design automation for microfluidic biochips and integrated circuits

"}, bioImg=null, bioContent=

born in 1991. PhD, professor, PhD supervisor. His main research interest includes design automation for microfluidic biochips and integrated circuits

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327011065390543475, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065243742832, language=CN, stringName=黄兴, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4西北工业大学计算机学院 西安 710072, bio={"img":"sGKXo2bnSpDvvN8iQBDRWA==","content":"

1991年生. 博士,教授,博士生导师. 主要研究方向为微流控生物芯片和集成电路的设计自动化

"}, bioImg=sGKXo2bnSpDvvN8iQBDRWA==, bioContent=

1991年生. 博士,教授,博士生导师. 主要研究方向为微流控生物芯片和集成电路的设计自动化

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327011064895615589, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=4, ext=[AuthorCompanyExt(id=1327011064904004198, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064895615589, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4School of Computer Science, Northwestern Polytechnical University, Xi’an 710072), AuthorCompanyExt(id=1327011064912392807, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064895615589, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4西北工业大学计算机学院 西安 710072)])]), Author(id=1327011065457652341, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=liugenggeng@fzu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1327011065562509945, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065457652341, language=EN, stringName=Genggeng Liu, firstName=Genggeng, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, *, address=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116
2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116
3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116, bio={"content":"

born in 1988. PhD, professor, PhD supervisor. Senior member of CCF. His main research interest includes design automation for microfluidic biochips and integrated circuits

"}, bioImg=null, bioContent=

born in 1988. PhD, professor, PhD supervisor. Senior member of CCF. His main research interest includes design automation for microfluidic biochips and integrated circuits

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327011065667367546, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065457652341, language=CN, stringName=刘耿耿, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, *, address=1福州大学计算机与大数据学院 福州 350116
2大数据智能教育部工程研究中心(福州大学) 福州 350116
3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116, bio={"img":"3hRdH17G5rzUdQrx2Pintw==","content":"

1988年生. 博士,教授,博士生导师. CCF高级会员. 主要研究方向为微流控生物芯片和集成电路的设计自动化

"}, bioImg=3hRdH17G5rzUdQrx2Pintw==, bioContent=

1988年生. 博士,教授,博士生导师. CCF高级会员. 主要研究方向为微流控生物芯片和集成电路的设计自动化

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327011064618791516, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=1, ext=[AuthorCompanyExt(id=1327011064627180125, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116), AuthorCompanyExt(id=1327011064635568734, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1福州大学计算机与大数据学院 福州 350116)]), AuthorCompany(id=1327011064732037727, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=2, ext=[AuthorCompanyExt(id=1327011064740426336, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116), AuthorCompanyExt(id=1327011064765592161, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大数据智能教育部工程研究中心(福州大学) 福州 350116)]), AuthorCompany(id=1327011064815923810, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=3, ext=[AuthorCompanyExt(id=1327011064820118115, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116), AuthorCompanyExt(id=1327011064828506724, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116)])])], keywords=[Keyword(id=1327011065784808059, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, orderNo=1, keyword=continuous-flow microfluidic biochips), Keyword(id=1327011065847722620, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, orderNo=2, keyword=deep reinforcement learning), Keyword(id=1327011065931608701, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, orderNo=3, keyword=control logic), Keyword(id=1327011066007106174, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, orderNo=4, keyword=control channel network), Keyword(id=1327011066082603647, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, orderNo=5, keyword=diagonal channel routing), Keyword(id=1327011066141323904, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, orderNo=1, keyword=连续微流控生物芯片), Keyword(id=1327011066204238465, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, orderNo=2, keyword=深度强化学习), Keyword(id=1327011066267153026, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, orderNo=3, keyword=控制逻辑), Keyword(id=1327011066351039107, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, orderNo=4, keyword=控制通道网络), Keyword(id=1327011066451702404, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, orderNo=5, keyword=对角通道布线)], refs=[Reference(id=1327011073187754668, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, 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=Huang Xing, Ho T, Guo Wenzhong, et al. Computer-aided design techniques for flow-based microfluidic lab-on-a-chip systems[J]. ACM Computing Surveys, 2021, 54(5): 1−29, articleTitle=null, refAbstract=null), Reference(id=1327011073250669229, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Genggeng, Huang Hongbin, Chen Zhisheng, et al. Design automation for continuous-flow microfluidic biochips: A comprehensive review[J]. Integration, 2022, 82: 48−66, articleTitle=null, refAbstract=null), Reference(id=1327011073317778094, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Xu, Chen Zhen, Chen Zhisheng, et al. Architectural synthesis of continuous-flow microfluidic biochips with connection pair optimization[J]. Electronics, 2024, 13(2): 247, articleTitle=null, refAbstract=null), Reference(id=1327011073389081263, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=Einav S, Gerber D, Bryson P D, et al. Discovery of a hepatitis C target and its pharmacological inhibitors by microfluidic affinity analysis[J]. Nature Biotechnology, 2008, 26(9): 1019−1027, articleTitle=null, refAbstract=null), Reference(id=1327011073464578736, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=Hong J W, Chen Yan, Anderson W F, et al. Molecular biology on a microfluidic chip[J]. Journal of Physics: Condensed Matter, 2006, 18(18): 691−701, articleTitle=null, refAbstract=null), Reference(id=1327011073535881905, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Unger M A, Chou H P, Thorsen T, et al. Monolithic microfabricated valves and pumps by multilayer soft lithography[J]. Science, 2000, 288(5463): 113−116, articleTitle=null, refAbstract=null), Reference(id=1327011073619767986, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=Thorsen T, Maerkl S J, Quake S R. Microfluidic large-scale integration[J]. Science, 2002, 298(5593): 580−584, articleTitle=null, refAbstract=null), Reference(id=1327011073686876851, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Huang Xing, Ho T, Chakrabarty K, et al. Timing-driven flow-channel network construction for continuous-flow microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2019, 39(6): 1314−1327, articleTitle=null, refAbstract=null), Reference(id=1327011073762374324, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=Araci I E, Quake S R. Microfluidic very large scale integration (mVLSI) with integrated micromechanical valves[J]. Lab on a Chip, 2012, 12(16): 2803−2806, articleTitle=null, refAbstract=null), Reference(id=1327011073846260405, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Fidalgo L M, Maerkl S J. A software-programmable microfluidic device for automated biology[J]. Lab on a Chip, 2011, 11(9): 1612−1619, articleTitle=null, refAbstract=null), Reference(id=1327011073955312310, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=Lim Y C, Kouzani A Z, Duan W. Lab-on-a-chip: A component view[J]. Microsystem Technologies, 2010, 16: 1995−2015, articleTitle=null, refAbstract=null), Reference(id=1327011074030809783, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Tseng T M, Li Mengchu, Freitas D N, et al. Columba 2.0: A co-layout synthesis tool for continuous-flow microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2017, 37(8): 1588−1601, articleTitle=null, refAbstract=null), Reference(id=1327011074097918648, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=Yang Kailin, Yao Hailong, Ho T, et al. AARF: Any-angle routing for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2018, 37(12): 3042−3055, articleTitle=null, refAbstract=null), Reference(id=1327011074165027513, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Qin, Zuo Shiliang, Yao Hailong, et al. Hamming-distance-based valve switching optimization for control-layer multiplexing in flow-based microfluidic biochips[C]//Proc of the 22nd Asia and South Pacific Design Automation Conf. Piscataway, NJ: IEEE, 2017: 524−529, articleTitle=null, refAbstract=null), Reference(id=1327011074244719290, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Qin, Xu Yue, Zuo Shiliang, et al. Pressure-aware control layer optimization for flow-based microfluidic biochips[J]. IEEE Transactions on Biomedical Circuits and Systems, 2017, 11(6): 1488−1499, articleTitle=null, refAbstract=null), Reference(id=1327011074307633851, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhu Ying, Huang Xing, Li Bing, et al. Multicontrol: Advanced control-logic synthesis for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2019, 39(10): 2489−2502, articleTitle=null, refAbstract=null), Reference(id=1327011074399908540, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=Huang Xing, Cai Huayang, Guo Wenzhong, et al. Control-logic synthesis of fully programmable valve array using reinforcement learning[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2024, 43(1): 277−290, articleTitle=null, refAbstract=null), Reference(id=1327011074542514877, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=Liang Siyuan, Li Mengchu, Tseng T M, et al. CoMUX: Combinatorial-coding-based high-performance microfluidic control multiplexer design[C/OL]//Proc of the 41st ACM Int Conf on Computer-Aided Design. New York: ACM, 2022 [2024-06-19]. https://dl.acm.org/doi/abs/10.1145/3508352.3549353, articleTitle=null, refAbstract=null), Reference(id=1327011074609623742, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen Zhisheng, Huang Xing, Guo Wenzhong, et al. Physical synthesis of flow-based microfluidic biochips considering distributed channel storage[C]//Proc of Design, Automation & Test in Europe Conf & Exhibition. Piscataway, NJ: IEEE, 2019: 1525−1530, articleTitle=null, refAbstract=null), Reference(id=1327011074701898431, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Kai, Dinh T A, Ho T, et al. Control-layer routing and control-pin minimization for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2016, 36(1): 55−68, articleTitle=null, refAbstract=null), Reference(id=1327011074802561728, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=曾俊杰,秦龙,徐浩添,等,基于内在动机的深度强化学习探索方法综述[J]. 计算机研究与发展,2023,60(10):2359−2382, articleTitle=null, refAbstract=null), Reference(id=1327011074873864897, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=Zeng Junjie, Qin Long, Xu Haotian, et al. Exploration approaches in deep reinforcement learning based on intrinsic motivation: A review[J]. Journal of Computer Research and Development, 2023, 60(10): 2359−2382 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1327011074945168067, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=Sutton R S, Barto A G. Reinforcement learning: An introduction[M]//Adaptive Computation and Machine Learning. Cambridge, MA: MIT, 2018, articleTitle=null, refAbstract=null), Reference(id=1327011075008082628, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=唐楚哲,王肇国,陈海波. 机器学习方法赋能系统软件:挑战、实践与展望[J]. 计算机研究与发展,2023,60(5):964−973, articleTitle=null, refAbstract=null), Reference(id=1327011075112940229, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=Tang Chuzhe, Wang Zhaoguo, Chen Haibo. Empowering system software with machine learning methods: Challenges, practice, and prospects[J]. Journal of Computer Research and Development, 2023, 60(5): 964−973 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1327011075192632006, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=Kober J, Bagnell J A, Peters J. Reinforcement learning in robotics: A survey[J]. The International Journal of Robotics Research, 2013, 32(11): 1238−1274, articleTitle=null, refAbstract=null), Reference(id=1327011075272323783, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=Mnih V, Kavukcuoglu K, Silver D, et al. Human-level control through deep reinforcement learning[J]. Nature, 2015, 518(7540): 529−533, articleTitle=null, refAbstract=null), Reference(id=1327011075326849736, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Ziyu, Schaul T, Hessel M, et al. Dueling network architectures for deep reinforcement learning[C]//Proc of the 33rd Int Conf on Machine Learning. Cambridge, MA: MIT, 2016: 1995−2003, articleTitle=null, refAbstract=null), Reference(id=1327011075414930121, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Chunfeng, Huang Xing, Li Bing, et al. DCSA: Distributed channel-storage architecture for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2020, 40(1): 115−128, articleTitle=null, refAbstract=null), Reference(id=1327011075482038986, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=Marcus J S, Anderson W F, Quake S R. Microfluidic single-cell mRNA isolation and analysis[J]. Analytical Chemistry, 2006, 78(9): 3084−3089, articleTitle=null, refAbstract=null), Reference(id=1327011077147177675, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=Van Hasselt H, Guez A, Silver D. Deep reinforcement learning with double q-learning[C]//Proc of the 30th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2016: 2094−2100, articleTitle=null, refAbstract=null), Reference(id=1327011077218480844, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=Mnih V, Kavukcuoglu K, Silver D, et al. Playing Atari with deep reinforcement learning[J]. arXiv preprint, arXiv:1312.560, 2013, articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1327011064618791516, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=1, ext=[AuthorCompanyExt(id=1327011064627180125, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116), AuthorCompanyExt(id=1327011064635568734, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1福州大学计算机与大数据学院 福州 350116)]), AuthorCompany(id=1327011064732037727, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=2, ext=[AuthorCompanyExt(id=1327011064740426336, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116), AuthorCompanyExt(id=1327011064765592161, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大数据智能教育部工程研究中心(福州大学) 福州 350116)]), AuthorCompany(id=1327011064815923810, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=3, ext=[AuthorCompanyExt(id=1327011064820118115, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116), AuthorCompanyExt(id=1327011064828506724, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116)]), AuthorCompany(id=1327011064895615589, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=4, ext=[AuthorCompanyExt(id=1327011064904004198, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064895615589, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4School of Computer Science, Northwestern Polytechnical University, Xi’an 710072), AuthorCompanyExt(id=1327011064912392807, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064895615589, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4西北工业大学计算机学院 西安 710072)])], figs=[ArticleFig(id=1327011068196532869, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 1, caption=Schematic diagram of the basic structure of CFMBs, figureFileSmall=OQkEvthHYCDD5oEY205ZFw==, figureFileBig=eoHUttSHMYVO2qxY5hoBng==, tableContent=null), ArticleFig(id=1327011068263641734, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图1, caption=CFMBs基本结构示意图, figureFileSmall=OQkEvthHYCDD5oEY205ZFw==, figureFileBig=eoHUttSHMYVO2qxY5hoBng==, tableContent=null), ArticleFig(id=1327011068532077191, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 2, caption=General platform of a complete biochip, figureFileSmall=4+0yfQxwZIwUMQDHaKO+Eg==, figureFileBig=rFaY+hrV1RTLtc40YfFsxA==, tableContent=null), ArticleFig(id=1327011068594991752, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图2, caption=完整的生物芯片通用平台, figureFileSmall=4+0yfQxwZIwUMQDHaKO+Eg==, figureFileBig=rFaY+hrV1RTLtc40YfFsxA==, tableContent=null), ArticleFig(id=1327011068653712009, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 3, caption=Control logic architecture for driving three flow valves, figureFileSmall=iSEC+myC6xg7wNN+8UqkUQ==, figureFileBig=+R84VRZcgfIsAjPGwlyQsA==, tableContent=null), ArticleFig(id=1327011068771152522, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图3, caption=驱动3个流阀门的控制逻辑架构, figureFileSmall=iSEC+myC6xg7wNN+8UqkUQ==, figureFileBig=+R84VRZcgfIsAjPGwlyQsA==, tableContent=null), ArticleFig(id=1327011068850844299, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 4, caption=Flow chart of DRL-based control logic routing algorithm, figureFileSmall=2gcY9wIKBcNIVJZ9chDgyg==, figureFileBig=CaYyhYho1z94zURyZwrYAA==, tableContent=null), ArticleFig(id=1327011068922147468, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图4, caption=基于DRL的控制逻辑布线算法流程图, figureFileSmall=2gcY9wIKBcNIVJZ9chDgyg==, figureFileBig=CaYyhYho1z94zURyZwrYAA==, tableContent=null), ArticleFig(id=1327011068989256333, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 5, caption=Pattern sequences of flow valves $ {f_1} $ to $ {f_5} $ and initialized routing grid, figureFileSmall=lESM2cAxeUDXC8unFvJgaw==, figureFileBig=fKCh1fB9zGsycKGA6EtIlQ==, tableContent=null), ArticleFig(id=1327011069064753806, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图5, caption=流阀门$ {f_1} $至$ {f_5} $的模式序列以及初始化的布线网格, figureFileSmall=lESM2cAxeUDXC8unFvJgaw==, figureFileBig=fKCh1fB9zGsycKGA6EtIlQ==, tableContent=null), ArticleFig(id=1327011069136056975, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 6, caption=Schematic diagram of routing angles for control logic, figureFileSmall=Q89vGeDkN+HdcNcN/TVZ0w==, figureFileBig=Jy4aeDdhlSoYApIuqC+DnQ==, tableContent=null), ArticleFig(id=1327011069207360144, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图6, caption=控制逻辑布线角度示意图, figureFileSmall=Q89vGeDkN+HdcNcN/TVZ0w==, figureFileBig=Jy4aeDdhlSoYApIuqC+DnQ==, tableContent=null), ArticleFig(id=1327011069320606353, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 7, caption=Schematic diagram of control paths of flow valves $ {f_1} $to$ {f_5} $, figureFileSmall=+I+dOvIUdE3yb2S50Z7qEg==, figureFileBig=zL9LNB3P4URtyqs/Rfgc6A==, tableContent=null), ArticleFig(id=1327011069383520914, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图7, caption=流阀门$ {f_1} $至$ {f_5} $的控制路径示意图, figureFileSmall=+I+dOvIUdE3yb2S50Z7qEg==, figureFileBig=zL9LNB3P4URtyqs/Rfgc6A==, tableContent=null), ArticleFig(id=1327011069463212691, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Figure 8, caption=Routing results generated by our proposed algorithm and DRA for RC02, figureFileSmall=zgL820gscxnXDcUTCRxvag==, figureFileBig=2AP4P4I0PNxuvk9z4D8O+Q==, tableContent=null), ArticleFig(id=1327011069547098772, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=图8, caption=本文算法与DRA针对RC02生成的布线结果, figureFileSmall=zgL820gscxnXDcUTCRxvag==, figureFileBig=2AP4P4I0PNxuvk9z4D8O+Q==, tableContent=null), ArticleFig(id=1327011069610013333, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 1, caption=Configuration of DuelingDQN Architecture, figureFileSmall=null, figureFileBig=null, tableContent=
网络层数网络类型神经元数量激活函数
1输入层7
2隐层(全连接)64ReLu
3隐层(全连接)128ReLu
4隐层(全连接)128ReLu
5输出层5
), ArticleFig(id=1327011069698093718, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表1, caption=

DuelingDQN架构配置

, figureFileSmall=null, figureFileBig=null, tableContent=
网络层数网络类型神经元数量激活函数
1输入层7
2隐层(全连接)64ReLu
3隐层(全连接)128ReLu
4隐层(全连接)128ReLu
5输出层5
), ArticleFig(id=1327011069773591191, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 2, caption=Details of Test Cases Used in Our Paper, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例网格大小流阀门数量控制阀门数量
RC0118×342075
RC0218×432692
RC0318×5331123
RC0418×5928139
RC0518×8250229
RC0619×9852327
), ArticleFig(id=1327011069849088664, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表2, caption=

本文所用测试用例的详细信息

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例网格大小流阀门数量控制阀门数量
RC0118×342075
RC0218×432692
RC0318×5331123
RC0418×5928139
RC0518×8250229
RC0619×9852327
), ArticleFig(id=1327011069920391833, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 3, caption=Setting of Parameters Used in Our Paper, figureFileSmall=null, figureFileBig=null, tableContent=
参数取值参数取值
$ c_{\text{p}}^1 $20$ c_{\text{n}}^4 $−10
$ c_{\text{p}}^2 $15$ {B_{\text{S}}} $128
$ c_{\text{p}}^3 $8$ {E_{\text{P}}} $45000
$ c_{\text{n}}^1 $−40$ \varepsilon $0.9
$ c_{\text{n}}^2 $−2$ {U_{\text{c}}} $200
$ c_{\text{n}}^3 $−5$ \partial $10-3
), ArticleFig(id=1327011070025249434, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表3, caption=

本文使用的参数设置

, figureFileSmall=null, figureFileBig=null, tableContent=
参数取值参数取值
$ c_{\text{p}}^1 $20$ c_{\text{n}}^4 $−10
$ c_{\text{p}}^2 $15$ {B_{\text{S}}} $128
$ c_{\text{p}}^3 $8$ {E_{\text{P}}} $45000
$ c_{\text{n}}^1 $−40$ \varepsilon $0.9
$ c_{\text{n}}^2 $−2$ {U_{\text{c}}} $200
$ c_{\text{n}}^3 $−5$ \partial $10-3
), ArticleFig(id=1327011070104941211, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 4, caption=Comparison of the Total Control Channel Length, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%
DRA本文算法DRA
RC01268.2231.613.6
RC02332.8296.610.9
RC03455.8405.212.5
RC04483.0415.813.9
RC05699.6618.411.6
RC06903.4804.011.0
平均12.3
), ArticleFig(id=1327011070193021596, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表4, caption=

控制通道总长度的比较

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%
DRA本文算法DRA
RC01268.2231.613.6
RC02332.8296.610.9
RC03455.8405.212.5
RC04483.0415.813.9
RC05699.6618.411.6
RC06903.4804.011.0
平均12.3
), ArticleFig(id=1327011070272713373, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 5, caption=Comparison of the Signal Transmission Delay, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例时延/s优化率/%
DRA本文算法DRA
RC012.01.620.0
RC021.61.412.5
RC031.81.422.2
RC041.81.611.1
RC052.21.818.2
RC062.01.620.0
平均17.3
), ArticleFig(id=1327011070364988062, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表5, caption=

信号传输时延的比较

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例时延/s优化率/%
DRA本文算法DRA
RC012.01.620.0
RC021.61.412.5
RC031.81.422.2
RC041.81.611.1
RC052.21.818.2
RC062.01.620.0
平均17.3
), ArticleFig(id=1327011070440485535, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 6, caption=Generalization Verification of Our Proposed Algorithm, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%时延/s优化率/%
DRA本文算法DRADRA本文算法DRA
RC05699.6625.810.52.21.818.2
RC06903.4821.09.12.01.620.0
), ArticleFig(id=1327011070515983008, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表6, caption=

本文算法的泛化性验证

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%时延/s优化率/%
DRA本文算法DRADRA本文算法DRA
RC05699.6625.810.52.21.818.2
RC06903.4821.09.12.01.620.0
), ArticleFig(id=1327011070587286177, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 7, caption=Effectiveness Validation of DuelingDQN with Respect to the Total Control Channel Length, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%
DQNDoubleDQNDuelingDQNDQNDoubleDQN
RC01249.8241.4231.67.34.1
RC02328.7315.4296.69.86.0
RC03456.8426.2405.211.34.9
RC04480.2448.0415.813.47.2
RC05687.2642.2618.410.03.7
RC06890.4844.6804.09.74.8
平均10.35.1
), ArticleFig(id=1327011070683755170, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表7, caption=

DuelingDQN在控制通道总长度方面的有效性验证

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%
DQNDoubleDQNDuelingDQNDQNDoubleDQN
RC01249.8241.4231.67.34.1
RC02328.7315.4296.69.86.0
RC03456.8426.2405.211.34.9
RC04480.2448.0415.813.47.2
RC05687.2642.2618.410.03.7
RC06890.4844.6804.09.74.8
平均10.35.1
), ArticleFig(id=1327011070759252644, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 8, caption=Effectiveness Validation of DuelingDQN with Respect to the Signal Transmission Delay, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例时延/s优化率/%
DQNDoubleDQNDuelingDQNDQNDoubleDQN
RC011.81.81.611.111.1
RC021.61.41.412.50.0
RC031.81.61.422.212.5
RC041.81.61.611.10.0
RC052.21.81.818.20.0
RC062.01.81.620.011.1
平均15.65.8
), ArticleFig(id=1327011070889276069, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表8, caption=

DuelingDQN在信号传输时延方面的有效性验证

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例时延/s优化率/%
DQNDoubleDQNDuelingDQNDQNDoubleDQN
RC011.81.81.611.111.1
RC021.61.41.412.50.0
RC031.81.61.422.212.5
RC041.81.61.611.10.0
RC052.21.81.818.20.0
RC062.01.81.620.011.1
平均15.65.8
), ArticleFig(id=1327011070981550758, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 9, caption=Effectiveness Validation of Diagonal Routing with Respect to the Total Control Channel Length, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%
MHR本文算法MHR
RC01292.0231.620.7
RC02383.0296.622.6
RC03509.0405.220.4
RC04534.0415.822.2
RC05768.0618.419.5
RC06996.0804.019.3
平均20.8
), ArticleFig(id=1327011072659272359, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表9, caption=

对角型布线在控制通道总长度方面的有效性验证

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例通道总长度/mm优化率/%
MHR本文算法MHR
RC01292.0231.620.7
RC02383.0296.622.6
RC03509.0405.220.4
RC04534.0415.822.2
RC05768.0618.419.5
RC06996.0804.019.3
平均20.8
), ArticleFig(id=1327011072743158440, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 10, caption=Effectiveness Validation of Diagonal Routing with Respect to the Signal Transmission Delay, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例时延/s优化率/%
MHR本文算法MHR
RC012.11.623.8
RC021.81.422.2
RC032.11.433.3
RC042.21.627.3
RC052.61.830.8
RC062.51.636.0
平均28.9
), ArticleFig(id=1327011072806073001, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表10, caption=

对角型布线在信号传输时延方面的有效性验证

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例时延/s优化率/%
MHR本文算法MHR
RC012.11.623.8
RC021.81.422.2
RC032.11.433.3
RC042.21.627.3
RC052.61.830.8
RC062.51.636.0
平均28.9
), ArticleFig(id=1327011072898347690, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=EN, label=Table 11, caption=Training Time and Inference Time of Our Proposed Algorithms and Running Time of DRA and MHR, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例训练时间/min推理时间/s运行时间/s
本文算法本文算法DRAMHR
RC0190.30.10.10.1
RC02112.50.10.20.1
RC03150.20.10.30.1
RC04172.50.10.30.2
RC05225.20.20.30.2
RC06337.50.20.40.2
), ArticleFig(id=1327011072994816683, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, language=CN, label=表11, caption=

本文算法的训练和推理时间以及DRA和MHR的运行时间

, figureFileSmall=null, figureFileBig=null, tableContent=
测试用例训练时间/min推理时间/s运行时间/s
本文算法本文算法DRAMHR
RC0190.30.10.10.1
RC02112.50.10.20.1
RC03150.20.10.30.1
RC04172.50.10.30.2
RC05225.20.20.30.2
RC06337.50.20.40.2
)], attaches=null, journal=Journal(id=1324759606947315908, delFlag=0, nameCn=计算机研究与发展, nameEn=Journal of Computer Research and Development, nameHistory1=null, nameHistory2=null, issn=1000-1239, eissn=null, cn=11-1777/TP, 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=k2l8zgqwstuxxevFX2qxVA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Computer Research and Development, journalRemark=null, publicationField=null, createdTime=1793776903141, updatedTime=1793850698612, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Electronic Information, subjectName=null, subjectCodeEn=Electronic Information, subjectNameEn=null, picCn=k2l8zgqwstuxxevFX2qxVA==, picEn=r6k0Ry+RD+c/VXj5QarmDw==, jcr=null, cjcr=null, exts=[JournalExt(id=1325069127710286688, language=CN, name=计算机研究与发展, 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=1793850698640, updatedTime=1793850698640, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1325069127777395553, language=EN, name=Journal of Computer Research and Development, 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=1793850698656, updatedTime=1793850698656, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1324759914146529478, websiteList=[Website(id=1325070300139892771, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, 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/jsjyjyfz/CN, language=CN, createTime=1793850978169, createBy=18614031015, updateTime=1793851040259, updateBy=18614031015, name=计算机研究与发展-中文, tplId=1146099689490845704, title=计算机研究与发展, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070893751349360, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=articleTextType, value=kx, createTime=1793851119697, updateTime=1793851119697, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893730377837, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=banner, value=null, createTime=1793851119692, updateTime=1793851119692, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893772320883, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=grayFlag, value=0, createTime=1793851119702, updateTime=1793851119702, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893721989228, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851119690, updateTime=1793851119690, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893784903797, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=minRunFlag, value=0, createTime=1793851119705, updateTime=1793851119705, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893742960751, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic, createTime=1793851119695, updateTime=1793851119695, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893780709492, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=silenceFlag, value=0, createTime=1793851119704, updateTime=1793851119704, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893738766446, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1793851119694, updateTime=1793851119694, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893755543665, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeColor, value=null, createTime=1793851119698, updateTime=1793851119698, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893763932274, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeStyle, value=null, createTime=1793851119700, updateTime=1793851119700, creator=18614031015, updator=18614031015)]), Website(id=1325070300580294698, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, 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/jsjyjyfz/EN, language=EN, createTime=1793850978274, createBy=18614031015, updateTime=1793851036638, updateBy=18614031015, name=计算机研究与发展-英文, tplId=1146101810881728533, title=Journal of Computer Research and Development, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070925288321148, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=articleTextType, value=kx, createTime=1793851127216, updateTime=1793851127216, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925263155321, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=banner, value=null, createTime=1793851127210, updateTime=1793851127210, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925309292671, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=grayFlag, value=0, createTime=1793851127221, updateTime=1793851127221, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925254766712, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851127208, updateTime=1793851127208, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925326069889, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=minRunFlag, value=0, createTime=1793851127225, updateTime=1793851127225, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925279932539, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic, createTime=1793851127214, updateTime=1793851127214, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925317681280, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=silenceFlag, value=0, createTime=1793851127223, updateTime=1793851127223, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925271543930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1793851127212, updateTime=1793851127212, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925292515453, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeColor, value=null, createTime=1793851127217, updateTime=1793851127217, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925300904062, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeStyle, value=null, createTime=1793851127219, updateTime=1793851127219, creator=18614031015, updator=18614031015)])], journalTitle=计算机研究与发展, weixinUrl=null, journalUrl=https://crad.ict.ac.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Computer Research and Development, journalPhotoCn=k2l8zgqwstuxxevFX2qxVA==, journalPhotoEn=r6k0Ry+RD+c/VXj5QarmDw==, journalFirstLetter=J, 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/jsjyjyfz/CN/10.7544/issn1000-1239.202440034, detailUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/10.7544/issn1000-1239.202440034, pdfUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/PDF/10.7544/issn1000-1239.202440034, pdfUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/PDF/10.7544/issn1000-1239.202440034, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1743436800000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于深度强化学习的连续微流控生物芯片控制逻辑布线
蔡华洋 1, 2, 3 , 黄兴 4 , 刘耿耿 1, 2, 3, *
计算机研究与发展 | 体系结构 2025,62(4): 950-962
收起
收藏切换
计算机研究与发展 |体系结构 2025 , 62 (4) : 950 -962
基于深度强化学习的连续微流控生物芯片控制逻辑布线
全屏
[Author(id=1327011065008861801, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=c_huayang@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327011065092747885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065008861801, language=EN, stringName=Huayang Cai, firstName=Huayang, middleName=null, lastName=Cai, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116
2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116
3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116, bio={"content":"

born in 1997. PhD candidate. His main research interest includes design automation of microfluidic biochips

"}, bioImg=null, bioContent=

born in 1997. PhD candidate. His main research interest includes design automation of microfluidic biochips

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327011065168245358, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065008861801, language=CN, stringName=蔡华洋, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1福州大学计算机与大数据学院 福州 350116
2大数据智能教育部工程研究中心(福州大学) 福州 350116
3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116, bio={"img":"M6mam6vc0YyZ+QHX9r2nxQ==","content":"

1997年生. 博士研究生. 主要研究方向为微流控生物芯片的设计自动化

"}, bioImg=M6mam6vc0YyZ+QHX9r2nxQ==, bioContent=

1997年生. 博士研究生. 主要研究方向为微流控生物芯片的设计自动化

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327011064618791516, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=1, ext=[AuthorCompanyExt(id=1327011064627180125, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116), AuthorCompanyExt(id=1327011064635568734, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1福州大学计算机与大数据学院 福州 350116)]), AuthorCompany(id=1327011064732037727, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=2, ext=[AuthorCompanyExt(id=1327011064740426336, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116), AuthorCompanyExt(id=1327011064765592161, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大数据智能教育部工程研究中心(福州大学) 福州 350116)]), AuthorCompany(id=1327011064815923810, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=3, ext=[AuthorCompanyExt(id=1327011064820118115, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116), AuthorCompanyExt(id=1327011064828506724, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116)])]), Author(id=1327011065243742832, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327011065323434610, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065243742832, language=EN, stringName=Xing Huang, firstName=Xing, middleName=null, lastName=Huang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4School of Computer Science, Northwestern Polytechnical University, Xi’an 710072, bio={"content":"

born in 1991. PhD, professor, PhD supervisor. His main research interest includes design automation for microfluidic biochips and integrated circuits

"}, bioImg=null, bioContent=

born in 1991. PhD, professor, PhD supervisor. His main research interest includes design automation for microfluidic biochips and integrated circuits

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327011065390543475, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065243742832, language=CN, stringName=黄兴, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4西北工业大学计算机学院 西安 710072, bio={"img":"sGKXo2bnSpDvvN8iQBDRWA==","content":"

1991年生. 博士,教授,博士生导师. 主要研究方向为微流控生物芯片和集成电路的设计自动化

"}, bioImg=sGKXo2bnSpDvvN8iQBDRWA==, bioContent=

1991年生. 博士,教授,博士生导师. 主要研究方向为微流控生物芯片和集成电路的设计自动化

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327011064895615589, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=4, ext=[AuthorCompanyExt(id=1327011064904004198, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064895615589, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4School of Computer Science, Northwestern Polytechnical University, Xi’an 710072), AuthorCompanyExt(id=1327011064912392807, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064895615589, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4西北工业大学计算机学院 西安 710072)])]), Author(id=1327011065457652341, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=liugenggeng@fzu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1327011065562509945, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065457652341, language=EN, stringName=Genggeng Liu, firstName=Genggeng, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, *, address=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116
2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116
3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116, bio={"content":"

born in 1988. PhD, professor, PhD supervisor. Senior member of CCF. His main research interest includes design automation for microfluidic biochips and integrated circuits

"}, bioImg=null, bioContent=

born in 1988. PhD, professor, PhD supervisor. Senior member of CCF. His main research interest includes design automation for microfluidic biochips and integrated circuits

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327011065667367546, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, authorId=1327011065457652341, language=CN, stringName=刘耿耿, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, *, address=1福州大学计算机与大数据学院 福州 350116
2大数据智能教育部工程研究中心(福州大学) 福州 350116
3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116, bio={"img":"3hRdH17G5rzUdQrx2Pintw==","content":"

1988年生. 博士,教授,博士生导师. CCF高级会员. 主要研究方向为微流控生物芯片和集成电路的设计自动化

"}, bioImg=3hRdH17G5rzUdQrx2Pintw==, bioContent=

1988年生. 博士,教授,博士生导师. CCF高级会员. 主要研究方向为微流控生物芯片和集成电路的设计自动化

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327011064618791516, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=1, ext=[AuthorCompanyExt(id=1327011064627180125, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer and Data Science, Fuzhou University, Fuzhou 350116), AuthorCompanyExt(id=1327011064635568734, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064618791516, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1福州大学计算机与大数据学院 福州 350116)]), AuthorCompany(id=1327011064732037727, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=2, ext=[AuthorCompanyExt(id=1327011064740426336, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116), AuthorCompanyExt(id=1327011064765592161, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064732037727, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2大数据智能教育部工程研究中心(福州大学) 福州 350116)]), AuthorCompany(id=1327011064815923810, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, xref=3, ext=[AuthorCompanyExt(id=1327011064820118115, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116), AuthorCompanyExt(id=1327011064828506724, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327011059782758978, companyId=1327011064815923810, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116)])])]
蔡华洋1, 2, 3 , 黄兴4, 刘耿耿1, 2, 3, *
作者信息
  • 1福州大学计算机与大数据学院 福州 350116
  • 2大数据智能教育部工程研究中心(福州大学) 福州 350116
  • 3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116
  • 4西北工业大学计算机学院 西安 710072
通讯作者:
刘耿耿()
作者简介:

1997年生. 博士研究生. 主要研究方向为微流控生物芯片的设计自动化

1991年生. 博士,教授,博士生导师. 主要研究方向为微流控生物芯片和集成电路的设计自动化

1988年生. 博士,教授,博士生导师. CCF高级会员. 主要研究方向为微流控生物芯片和集成电路的设计自动化

Control Logic Routing for Continuous-Flow Microfluidic Biochips Based on Deep Reinforcement Learning
Huayang Cai1, 2, 3 , Xing Huang4, Genggeng Liu1, 2, 3, *
Affiliations
  • 1College of Computer and Data Science, Fuzhou University, Fuzhou 350116
  • 2Engineering Research Center of Big Data Intelligence (Fuzhou University), Ministry of Education, Fuzhou 350116
  • 3Fujian Key Laboratory of Network Computing and Intelligent Information Processing (Fuzhou University), Fuzhou 350116
  • 4School of Computer Science, Northwestern Polytechnical University, Xi’an 710072
  • born in 1997. PhD candidate. His main research interest includes design automation of microfluidic biochips

    born in 1991. PhD, professor, PhD supervisor. His main research interest includes design automation for microfluidic biochips and integrated circuits

    born in 1988. PhD, professor, PhD supervisor. Senior member of CCF. His main research interest includes design automation for microfluidic biochips and integrated circuits

出版时间: 2025-04-01 doi: 10.7544/issn1000-1239.202440034
文章导航
收藏切换

随着电子设计自动化技术的迅速发展,连续微流控生物芯片成为了目前最具前景的生化实验平台之一. 该芯片通过采用内部的微阀门以及微通道来操纵体积仅为毫升或纳升的流体样品,从而自动执行混合和检测等基本的生化实验操作. 为了实现正确的生化测定功能,部署于芯片内部的微阀门通常需要由基于多路复用器的控制逻辑进行管控,其通过控制通道获得来自核心输入的控制信号以实现精确切换. 由于生化反应通常需要非常高的灵敏度,因此为了保证信号的即时传输,需要尽可能地减少连接每个阀门的控制路径长度,以降低信号传输的时延. 此外,为了降低芯片的制造成本,如何有效减少控制逻辑中通道的总长度也是逻辑架构设计需要解决的关键问题之一. 针对上述问题,提出了一种基于深度强化学习的控制逻辑布线算法以最小化信号传输时延以及控制通道总长度,从而自动构建高效的控制通道网络. 该算法采用竞争深度Q网络架构作为深度强化学习框架的智能体,从而对信号传输时延和通道总长度进行权衡评估. 此外,针对控制逻辑首次实现了对角型的通道布线,从根本上提高了阀门切换操作的效率并降低了芯片的制造成本. 实验结果表明,所提出的算法能够有效构建高性能、低成本的控制逻辑架构.

连续微流控生物芯片  /  深度强化学习  /  控制逻辑  /  控制通道网络  /  对角通道布线

With the advancement of electronic design automation, continuous-flow microfluidic biochips have become one of the most promising platforms for biochemical experiments. This chip manipulates fluid samples in milliliters or nanoliters by utilizing internal microvalves and microchannels, and thus automatically performs basic biochemical experiments, such as mixing and detection. To achieve the correct bioassay function, the microvalves deployed inside the chip are usually managed by a multiplexer-based control logic, and valves receive control signals from a core input through the control channel for accurate switching. Since biochemical reactions typically require high sensitivity, the length of control paths connecting each valve needs to be reduced to ensure immediate signal propagation, and thus to reduce the signal propagation delay. In addition, to reduce the fabrication cost of chips, a vital issue to be addressed in the logic architecture design is how to effectively reduce the total channel length within the control logic. To address the above issues, we propose a deep reinforcement learning-based control logic routing algorithm to minimize the signal propagation delay and total control channel length, thereby automatically constructing an efficient control channel network. The algorithm employs the dueling deep Q-network architecture as the agent of the deep reinforcement learning framework to evaluate the tradeoff between signal propagation delay and total channel length. Besides, the diagonal channel routing is implemented for the first time for control logic, thus fundamentally improving the efficiency of valve switching operations and reducing the fabrication cost of the chip. The experimental results demonstrate that the proposed algorithm can effectively construct a high-performance and low-cost control logic architecture.

continuous-flow microfluidic biochips  /  deep reinforcement learning  /  control logic  /  control channel network  /  diagonal channel routing
蔡华洋, 黄兴, 刘耿耿. 基于深度强化学习的连续微流控生物芯片控制逻辑布线. 计算机研究与发展, 2025 , 62 (4) : 950 -962 . DOI: 10.7544/issn1000-1239.202440034
Huayang Cai, Xing Huang, Genggeng Liu. Control Logic Routing for Continuous-Flow Microfluidic Biochips Based on Deep Reinforcement Learning[J]. Journal of Computer Research and Development, 2025 , 62 (4) : 950 -962 . DOI: 10.7544/issn1000-1239.202440034
连续微流控生物芯片(continuous-flow microfluidic biochips,CFMBs),也称为片上实验设备,由于其具有高效率、高精度、低成本等优点,在过去的20年里已经成为了一个越来越有吸引力的生物化学实验平台[1-2]. 微流控技术通过将流体样品制备、反应、分离、检测等生物化学和医学分析过程的基本操作单元集成到微米尺度的芯片上,从而自动执行生化测定. 与需要人为干预的传统生化实验流程相比,避免了由于手动干预而产生的实验误差,显著地提升了实验结果的准确率和成功率. 如今,这种片上实验设备已经应用于生物化学与生物医学的许多领域[3],如药物研发[4]和DNA分析[5]等.
CFMBs的基本结构如图1所示,其由2层逻辑层构成,分别为流层和控制层,其中这2个层都由弹性体材料,即聚二甲基硅氧烷(polydimethylsiloxane,PDMS)[6]采用多层软光刻(multilayer soft lithography,MSL)技术制造而成. 每一个逻辑层都包含了一组属于该层的微通道,分别为控制层中的控制通道以及流层中的流通道,其中控制通道构建于流通道的顶部. 芯片上的设备,如混合器和检测器,通过流通道进行连接以实现设备间流体样品的输送. 如图1(a)所示,阀门是建立在流通道和控制通道交叉处的微小开关,且需要由外部的空气或流体压力模式驱动以改变其状态. 图1(b)展示了CFMBs的横截面示意图. 具体来说,外部的空气在施压下将从控制端口注入到控制通道进行传输,从而驱使阀门向下挤压流通道以阻止流动样品的移动. 当释放控制通道中的空气/流体压力时,阀门将恢复原样,同时来自于流端口的外部压力将促使流体继续沿着流通道进行传输. 换句话说,阀门的工作原理如同微型开关. 以阀门作为基本的控制单元,可以构建具有复杂功能的CFMBs[7-8].
随着制造技术的发展,成千上万的阀门已经可以被集成到一个只有几平方厘米的单一芯片中,密度可以达到每平方厘米100万个[9]. 然而,由于受到芯片面积以及组件尺寸的限制,为每个阀门分配1个独立的控制端口是不切实际的,其将会导致芯片的制造复杂性以及成本大大增加. 为了克服上述挑战,采用基于多路复用器的控制逻辑来控制芯片中的阀门已经成为了一种必不可少的选择.
图2展示了一个具有控制逻辑的完整生物芯片平台[10],其中平台的中心为包含了116个阀门的连续微流控生物芯片,底部的核心输入则提供指导阀门打开或关闭的控制信号. 围绕在芯片周围的控制通道、多路复用器、核心输入以及右侧的控制端口共同形成一个控制逻辑,从而生成用于切换芯片中阀门状态的控制模式. 当执行生化测定时,该平台不是直接使用116个控制端口,而是部署了一个仅包含14个控制端口的多路复用器来生成控制模式以实现对所有阀门的状态切换. 通常来说,$ {2^{n/2}} $个阀门仅需要被具有n个控制端口的多路复用器进行控制[10]. 因此,基于多路复用器的控制逻辑在一定程度上减少了连续微流控生物芯片中控制端口的使用数量. 由于在执行生化测定的过程中,芯片内部的阀门需要频繁切换,因此在控制逻辑的设计过程中需要仔细规划和构建大量的控制路径. 这些路径在微小的区域内相互交叉以形成一个高度集成的控制通道网络,且主导着生物芯片的性能.
为了从根本上提高生物芯片的性能,同时降低芯片的制造成本,针对控制逻辑的通道布线设计应当考虑以下问题:1)应尽量减少连接芯片内每个阀门的控制路径长度以降低信号传输时延,从而缩短切换阀门的响应时间. 由于控制通道是采用PDMS材料制造而成的,受到此材料物理特性的影响,压力信号在通道中的传播速度较低[11]. 然而,CFMBs中的生化反应通常需要非常高的灵敏度,因此在一些生化实验操作中,阀门需要做出快速响应. 通常来说,连接阀门的控制路径增长将会导致该路径下信号的传输时延增大,那么阀门对信号的响应速度也会相应降低,且较慢的响应速度将增加最终生化反应的完成时间. 因此,信号传输时延是评估控制逻辑架构设计的一个重要指标. 2)需要尽可能减少控制逻辑中所需的控制通道总长度以进一步提高CFMBs的执行效率,同时降低芯片的制造成本. 由于现有的控制逻辑设计仍采用仅有水平和垂直布线方向的曼哈顿布线方式来实现控制通道网络构建,因此在一定程度上增加了控制逻辑所需的通道总长度. 此外,控制通道总长度的增加同样也导致了信号传输时延的增大,从而降低了生物芯片的执行效率. 因此,本文在原有曼哈顿布线方式的基础上引入对角型的布线方式以实现控制逻辑中通道网络的构建. 需要注意的是,当前的制造技术已经支持对PDMS微通道实现对角型布线[12-13].
目前为止,已经出现了一些针对基于多路复用器控制逻辑的相关工作. 此类控制逻辑架构首先是由Thorsen等人[7]所提出的,该架构由二进制阀门的组合阵列所构成,且通过2N个控制通道以分别驱动2N个流通道的状态. Fidalgo等人[10]进一步提出采用控制通道来对阀门进行定位的控制逻辑架构,该架构通过部署2N个基于二进制编码的控制端口以实现对2N个阀门的单独控制. Wang等人[14]提出了一种基于汉明距离的阀门切换优化方法,通过调整控制逻辑内部阀门所需的模式顺序来优化阀门的切换顺序,从而提高控制逻辑的可靠性. 在此基础上,Wang等人[15]提出一种基于异或的压力刷新方法,从而克服控制逻辑内部阀门切换时产生的压力偏差问题. Zhu等人[16]提出了一种称为多通道控制的合成方法来生成特定的控制逻辑,该控制逻辑可以同时对一些控制通道进行寻址,然后利用剩余的编码容量引入备用通道,从而提高控制逻辑的容错能力. Huang等人[17]提出了一种基于强化学习的控制逻辑合成方法以进一步提高阀门的切换效率并减少了架构设计的成本. 此外,Liang等人[18]提出了一种基于斯珀纳定理的组合编码策略以减少控制逻辑的阀门和控制通道的数量,从而进一步提高了控制逻辑的可靠性.
尽管上述方法在一定程度上优化了现有的控制逻辑架构,但尚未对控制逻辑通道布线方面存在的问题进行考虑,且这些方法中逻辑架构的控制通道仍采用曼哈顿的布线方式,因此一定程度上限制了控制通道网络构建的灵活性. 为了从根本上解决上述问题,本文研究了非曼哈顿布线方式在CFMBs控制逻辑布线中的应用,并提出了一种基于深度强化学习的控制逻辑布线算法.
本文的主要贡献包括4个方面:
1)为CFMBs的控制逻辑提出了第1个同时考虑信号传输时延和通道总长度的控制通道网络构建问题模型,这对于提高控制逻辑的布线设计自动化以及性能优化具有重要意义.
2)针对CFMBs提出了一种基于深度强化学习(deep reinforcement learning,DRL)的控制逻辑布线算法. 该算法采用竞争深度Q网络(dueling deep Q-network,DuelingDQN)架构对信号传输时延和通道总线长进行权衡评估,通过减少阀门的控制路径长度以实现有效的通道复用,从而同时降低信号传输时延和控制设计成本.
3)首次将具有对角型通道的控制通道网络集成到生物芯片的控制逻辑中,从而进一步提高了阀门切换操作的效率,同时降低了逻辑架构的整体成本.
4)使用6个测试用例来评估所提出的布线算法. 实验结果表明,所提出的算法能够为CFMBs生成具有较短信号传输时延以及较低芯片设计成本的优化控制逻辑架构.
如前所述,在CFMBs上执行生化测定时,需要一个控制逻辑生成控制模式,从而更新流层中的阀门(后续称为流阀门)的状态,以便生化操作能够根据用户预先制定的测定计划有效执行[19]. 如图3所示,本文以一个用于驱动3个流阀门的逻辑架构示意图为例来说明控制逻辑的工作原理. 在该架构中,核心输入部分和各流阀门由控制通道进行连接,从而传输驱动流阀门状态切换的控制信号. 4个连接到外部压力源的控制端口$ {x_1} $,$ {\bar x_1} $,$ {x_2} $,$ {\bar x_2} $被部署在右侧以引导构建在控制通道上方的控制阀门状态. 需要注意的是,控制端口的压力值通常是互补的[10],这意味着一对互补的控制端口,例如$ {x_1} $和$ {\bar x_1} $,仅有1个端口可被设置为高压. 因此,通过控制端口的组合可以生成指定各控制通道连接情况的控制模式,从而实现对应流阀门的状态更新.
例如,图3的控制逻辑中流阀门1连接了1条控制通道,而流阀门2,3则分别连接了2条控制通道,其中连接流阀门1的通道被分配了控制模式$ {x_1}{x_2} $. 一旦控制端口$ {x_1} $和$ {x_2} $被设置为高压时,表示为逻辑值1,那么这2个端口所连接的控制阀门被打开,且控制逻辑左侧的通道将被连通,使得流阀门1的状态能够更新为与核心输入的值相同. 相反,流阀门2,3的状态则保持不变,因为剩余2个控制端口$ {\bar x_1} $和$ {\bar x_2} $都为低压,表示为逻辑值0. 此时连接流阀门2,3的控制通道都未被连通,那么来自于核心输入的控制信号则未能传输至这2个流阀门以实现状态切换. 需要注意的是,连接流阀门2的2条控制通道被分别分配了控制模式$ {\bar x_1}{x_2} $和$ {\bar x_1}{\bar x_2} $,其中控制模式$ {\bar x_1}{x_2} $同样也分配给了连接流阀门3的控制通道. 一旦控制模式$ {\bar x_1}{x_2} $被激活,流阀门2和流阀门3将同时进行状态切换. 此外,该控制逻辑仍然保留了流阀门2和流阀门3单独切换的功能. 这2个流阀门能够通过分别激活控制模式$ {\bar x_1}{\bar x_2} $和$ {x_1}{\bar x_2} $以实现独立的状态切换. 基于上述方式,所有流阀门的状态都需要激活对应的控制模式进行更新,从而保证用户能够编程一系列由多个控制模式组成的模式序列以自动执行生化测定.
与传统电子系统中的布线问题不同,例如在印刷电路和超大规模集成电路设计中,信号传输是电子化的,而CFMBs中的控制信号是由外部压力产生的,该信号受到PDMS材料的影响,在控制通道中的传输速度相对较低(通常为10 mm/s)[20]. 当控制信号从核心输入处注入时,CFMBs中的流阀门不能立即关闭,因此信号传输延迟的现象在流阀门的切换过程中普遍发生. 为了评估延迟的程度,需要对信号传输时延Tp进行计算,其定义为:
$ T_{\text{p}}=L/v_{\text{p}}, $
其中L为控制路径的长度,$ {v_{\text{p}}} $为控制信号在控制通道中的传输速度(在本文实验中$ {v_{\text{p}}} $同样设置为10 mm/s).
然而,较长的信号传输时延将导致2个问题:1)流阀门切换的响应时间较长. 由于压力信号在控制通道中存在传输延迟,阀门不能对信号做出即时反应,缓慢的阀门切换不仅增加了生化反应的完成时间,而且对于时间灵敏度要求高的生化反应而言,较慢的阀门响应可能会影响最终实验结果的质量. 2)较长的控制模式设置时间. 为了确保所有的阀门完全切换到所需的状态,在执行2个连续的阀门控制模式之间必须插入1个模式设置时间,该时间必须大于最长控制路径所造成的延迟,那么控制模式引导的所有阀门才能被设置到合适的状态. 由于该时间间隙发生在激活每个控制模式之前,因此同样需要通过优化信号传输时延以减少控制模式设置时间. 综上所述,如何有效降低控制信号的传输时延已成为CFMBs控制逻辑设计中的一个关键问题.
另一方面,在CFMBs控制逻辑的架构设计中,应合理地规划和构建控制通道网络,以便所有的流阀门能够按实际需求连接至控制逻辑,从而实现控制信号的有效传输. 此外,由于控制通道网络的构建直接决定了最终的控制逻辑架构,因此在通道布线阶段还应优化控制逻辑的通道总长度,从而有效降低芯片的制造成本并进一步提高其执行效率. 为此,可以在原有逻辑架构中加入对角型的通道布线方式,同时通过有效规划控制通道以实现不同控制路径之间的资源共享,从而为CFMBs控制逻辑构建一个高效的控制通道网络.
需要注意的是,虽然通过减少控制通道总长度能够一定程度上优化信号传输时延,但并不能仅从该方面进行优化以实现传输时延的减小. 实际上信号传输时延还受到核心输入与流阀门之间控制路径长度的影响. 此外,所有控制路径的长度总和并不等同于控制通道的总长度. 这是因为不同的控制路径会存在复用同一段控制通道的情况,复用的控制通道越多,那么所需的控制通道总长度也会相应地减少. 然而,如果某条控制路径仅通过尽可能复用其他路径使用的控制通道段(即,减少控制通道总长度)以优化信号传输时延,那么可能会导致该路径并不会采用最短路径的布线方式连接到流阀门,而是为了实现更多的通道复用采取绕路的布线方式,从而导致该路径的信号传输时延增大. 因此本文算法需要从信号传输时延和通道总线长2个方面进行权衡评估,以生成一个合适的控制逻辑布线方案.
基于上述分析,针对CFMBs控制逻辑的控制通道网络构建问题模型表述如下:
输入:控制逻辑中控制阀门的位置,CFMBs中流阀门的位置,驱动所有流阀门的模式序列.
输出:考虑信号传输时延的优化控制逻辑架构.
目标:最小化控制信号传输时延,最小化控制通道总长度.
DRL是一种将强化学习与深度学习相结合的机器学习方法. 在深度强化学习中,智能体通过观察环境的状态,实现动作与环境的交互,从而获得奖励信号[21]. 智能体的目标是通过与环境的不断交互来学习如何在给定任务中做出正确的决策以最大化累积奖励.
上述过程通常被表述成一个马尔可夫决策过程(Markov decision process,MDP)[22],通常定义为(S,A,T,R),其中S是一个包含了智能体所有可感知状态的集合,A是一个包含了智能体所有可采取动作的集合,S和A分别称为状态空间和动作空间;T是一个状态转移概率,表示在给定状态和动作下,智能体在下一时刻转移到某个状态的概率;R是一个奖励函数,该函数定义了智能体在特定状态下采取特定动作后所获得的奖励. DRL的基本流程如下所述:在时间步t下,智能体首先从环境中感知当前状态$ {s_t} $并从A中选取一个动作$ {a_t} $. 然后,智能体通过执行动作$ {a_t} $后从环境中获得奖励$ {r_t} $,并转移到下一个状态$ {s_{t + 1}} $. 之后,智能体再根据新状态$ {s_{t + 1}} $选择新的动作. 这个过程会不断地迭代,直至智能体学习到一个有效的策略. 到目前为止,DRL在许多领域都取得了重要的突破[23],如机器人控制[24]和自动驾驶[25]等.
为了构建一个高效的控制通道网络,本文提出了一种基于DRL的连续微流控生物芯片控制逻辑布线算法. 图4展示了该算法的整体设计流程,其主要由DRL框架构建以及网络架构训练2部分组成. 本文将所有控制阀门、流阀门的位置以及模式序列作为输入,并根据输入信息初始化布线网格. 图5(a)(b)分别展示了流阀门$ {f_1} $至$ {f_5} $对应的模式序列以及根据阀门位置初始化后的布线网格,其中图5(a)中从顶部核心输入沿着箭头方向到达底部某个流阀门所经过的中间节点组成该流阀门的控制模式. 例如,流阀门$ {f_1} $的控制模式为$ {x_1}{x_2}{\bar x_3} $.
然后,将CFMBs控制逻辑布线问题构建成符合MDP求解的DRL框架. 具体来说,首先为DRL智能体生成一个可感知控制逻辑布线环境的状态空间. 其次根据可用的控制通道布线角度构建动作空间,从而为智能体提供布线方向. 当动作空间构建完成后,制定能够引导智能体布线行为的奖励函数. 最后根据构建完成的DRL框架,对基于DuelingDQN的DRL智能体进行训练以获得一个有效的控制逻辑布线方案,从而输出优化的控制逻辑架构. 本文算法所涉及的更多技术细节,如上述的状态空间、动作空间、奖励函数设计、网络架构训练将会在后续进行介绍.
为了使DRL智能体能够准确感知到环境的信息,本文将时刻t下的状态表示为$ {s_t} $,该状态不仅取决于当前控制路径到达的位置,而且需要对各条控制路径进行区分. 原因是针对处于相同位置的不同控制路径,智能体选择的合理动作也可能不同. 如果不对控制路径进行区分,那么智能体将无法识别控制路径所适用的布线策略,从而导致最终布线策略失效. 因此,整个状态空间可以表述为
$ \begin{split}S =& \left\{ {{s_t}|{s_t} = [p_t^{\text{n}},p_t^{\text{v}},p_t^{\text{e}},p_t^{\text{i}}],p_t^{\text{v}} = } \right. \\& \left. {(v_t^x,v_t^y),p_t^{\text{e}} = (e_t^x,e_t^y),p_t^{\text{i}} = (i_t^x,i_t^y)} \right\}, \end{split} $
其中$ p_t^{\text{n}} $表示时刻t下正在规划的控制路径序号,$ p_t^{\text{v}} $表示时刻t下控制路径到达的位置,$ p_t^{\text{e}} $为时刻t下控制路径需要到达的目标点位置,$ p_t^{\text{i}} $为时刻t下控制路径需要到达的必经点位置. 上述位置信息由布线网格中各点的x,y坐标构成.
根据给定的模式序列,各条控制路径将会依次经过对应控制模式指定的必经点,并最终到达目标点,其中,必经点为控制逻辑中的控制阀门,目标点为CFMBs中的流阀门. 需要注意的是,状态$ {s_t} $中的必经点位置$ p_t^{\text{i}} $将会根据实际的布线情况进行更新. 如果智能体在时刻t下采取动作到达当前需要经过的必经点,那么在下一时刻,即时刻t+1,状态$ {s_{t + 1}} $中的$ p_{t + 1}^{\text{i}} $将更新为下一个需要经过的必经点位置. 最终,当前规划的控制路径在到达流阀门后结束布线,同时智能体开始为下一条控制路径的布线做准备,直到控制逻辑中所有的控制路径都完成布线.
本文将时刻t下所选择并执行的动作表示为$ {a_t} $,其中动作$ {a_t} $表示控制路径的布线行为. 针对控制逻辑的控制通道布线阶段,由于本文在原有曼哈顿布线方式的基础上引入了对角型的布线方式,因此动作空间可以表述成
$ A = \left\{ {{a_t}{\text{|}}{a_t} \in \{ {a_{\text{r}}},{a_{{\text{dr}}}},{a_{\text{d}}},{a_{{\text{dl}}}},{a_{\text{l}}}\} } \right\}, $
其中$ {a_{\text{r}}} $,$ {a_{{\text{dr}}}} $,$ {a_{\text{d}}} $,$ {a_{{\text{dl}}}} $,$ {a_{\text{l}}} $则分别表示智能体的5个布线方向,分别为0°,45°,90°,135°,180°,如图6所示. 当智能体在时刻t执行动作$ {a_t} $后,控制路径将依据动作$ {a_t} $所表示的布线方向通过1个网格. 此外,如果布线方向为非对角型,那么控制路径的长度将增加1个单位长度,即1 mm,否则控制路径长度将增加1.4 mm.
奖励是评价智能体所选择动作质量的直接信号. 为了引导DRL智能体探索并学习到有效的布线策略,本文的奖励函数将基于智能体执行动作后控制路径所到达的点的类型进行构建. 具体来说,控制路径到达的点主要分为5类,分别为上述的必经点$ {p^{\text{i}}} $、目标点$ {p^{\text{e}}} $、障碍点$ {p^{\text{o}}} $、共享点$ {p^{\text{s}}} $、普通点$ {p^{\text{c}}} $. 这5类节点的定义分别如下:
1)必经点. 控制路径根据给定的模式序列在布线过程中必须经过的特定位置,这类点通常为控制阀门.
2)目标点. 控制路径在布线过程中需要到达的终点,这类点为CFMBs中的流阀门.
3)障碍点. 控制路径在布线过程中需要避开的位置,这类点通常为其他路径的必经点以及目标点.
4)共享点. 多条控制路径汇聚或相交的位置,且该位置不能存在控制阀门.
5)普通点. 为布线网格内除了必经点、目标点、障碍点等特殊点之外的其他普通节点.
本文将时刻t下智能体在状态$ {s_t} $执行动作$ {a_t} $后获得的即时奖励表示为$ {r_t} $,因此总体的奖励函数可以表示为
$ r_t=\left\{\begin{aligned} & c_{\text{p}}^1,p_{t+1}^{\text{v}}=\text{ }p^{\text{i}}\text{,} \\ & c_{\text{n}}^1,p_{t+1}^{\text{v}}=\text{ }p^{\text{o}}\text{,} \\ & c_{\text{p}}^2,p_{t+1}^{\text{v}}=\text{ }p^{\text{e}}, \\ & R_{\text{p}}(p_{t+1}^{\text{v}}),p_{t+1}^{\text{v}}=\text{ }p^{\text{c}}, \\ & R_{\text{p}}(p_{t+1}^{\text{v}})\text{ }+\text{ }c_{\text{p}}^3,其他,\end{aligned}\right. $
其中$ p_{t + 1}^{\text{v}} $为时刻t下智能体执行动作$ {a_t} $后控制路径到达的位置,$ c_{\text{p}}^1 $,$ c_{\text{p}}^2 $,$ c_{\text{p}}^3 $分别为3个正的常数,$ c_{\text{n}}^1 $为负的常数,$ {R}_{\text{p}}({p}_{t+1}^{\text{v}})$是$ p_{t + 1}^{\text{v}} $为普通点$ {p^{\text{c}}} $的奖励函数.
图7展示了图5(b)中流阀门$ {f_1} $至$ {f_5} $对应的控制路径,本文以图中流阀门$ {f_5} $的控制路径(虚线)为例来说明奖励的计算,其中该路径的序号为50,且对应的控制模式为$ {\bar x_1}{\bar x_2}{\bar x_3} $. 在该例子中,假设时刻t下路径50的必经点为控制阀门$ {\bar x_2} $(上对角线网纹方格). 如果执行$ {a_t} $后$ p_{t + 1}^{\text{v}} $到达控制阀门$ {\bar x_2} $,正向奖励$ c_{\text{p}}^1 $将被反馈给智能体以鼓励其经过该必经点.
同时,除了当前属于必经点的控制阀门$ {\bar x_2} $外,网格中其余的控制阀门以及其他路径需要经过的流阀门都不可被控制路径50经过,因此这些点将被视为障碍点$ {p^{\text{o}}} $(灰底方格). 如果执行$ {a_t} $后$ p_{t + 1}^{\text{v}} $为障碍点$ {p^{\text{o}}} $,那么当前状态$ {s_t} $将在下一时刻转移成终止状态,且智能体将获得惩罚值$ c_{\text{n}}^1 $以引导其在布线过程中避开这类点.
此外,如果智能体在状态$ {s_t} $执行$ {a_t} $后可以到达目标点$ {p^{\text{e}}} $(灰底十字网纹方格),其将会获得一个正向奖励$ c_{\text{p}}^2 $以鼓励到达目标点的行为,即完成该条路径的构建. 需要注意的是,$ c_{\text{p}}^2 $小于上述的$ c_{\text{p}}^1 $,这是因为$c_{\text{p}}^2 $设置方式能够引导控制路径根据控制模式优先经过属于必经点的控制阀门,到达属于目标点的流阀门,从而保证路径正确的控制功能.
另一方面,针对路径50中$ p_{t + 1}^{\text{v}} $为普通点(白底方格)的情况,引入关键点$ {p^{\text{k}}} $以实现奖励计算,其中关键点${p^{\text{k}}} $包括了必经点$ {p^{\text{i}}} $以及目标点$ {p^{\text{e}}} $. 相关的奖励函数表示为
$ R_{\text{p}}(p_{t+1}^{\text{v}})=\left\{\begin{aligned} & c_{\text{n}}^2,md(p_{t+1}^{\text{v}},p_t^{\text{k}}) < md(p_t^{\text{v}},p_t^{\text{k}}), \\ & c_{\text{n}}^3,md(p_{t+1}^{\text{v}},p_t^{\text{k}})=md(p_t^{\text{v}},p_t^{\text{k}})\text{,} \\ & c_{\text{n}}^4,其他,\end{aligned}\right. $
其中$ c_{\text{n}}^2 $,$ c_{\text{n}}^3 $,$ c_{\text{n}}^4 $为3个负的常数,且$ c_{\text{n}}^2 $>$ c_{\text{n}}^3 $>$ c_{\text{n}}^4 $,$ md(p_{t + 1}^{\text{v}}, p_t^{\text{k}}) $和$ md(p_t^{\text{v}},p_t^{\text{k}}) $分别为智能体在时刻t执行动作前后控制路径到达的位置与当前关键点之间的曼哈顿距离. 根据布线情况,关键点$ {p^{\text{k}}} $将随着必经点$ {p^{\text{e}}} $同步更新. 需要注意的是,当控制路径到达最后一个必经点时,关键点$ {p^{\text{k}}} $将更新为目标点$ {p^{\text{e}}} $.
从上述奖励函数可以看出,智能体在执行$ {a_t} $后$ p_{t + 1}^{\text{v}} $到达普通点$ {p^{\text{c}}} $时存在3种情况:1)如果执行$ {a_t} $后使得控制路径更接近关键点,那么智能体将会获得一个较小的惩罚值$ c_{\text{n}}^2 $;2)如果执行$ {a_t} $后,控制路径到达的位置与关键点的曼哈顿距离保持不变,此时智能体会获得一个较大的惩罚值$ c_{\text{n}}^3 $,因为该行为没有带来更好的布线结果;3)如果执行$ {a_t} $后使得控制路径更远离关键点,此时智能体将会获得一个更大的惩罚值$ c_{\text{n}}^4 $,因为该行为导致了更差的布线结果. 在控制逻辑的布线过程中,该惩罚机制将为智能体提供有效信号以指引控制路径靠近关键点,且保证了每条控制路径以最短的布线方式连接对应的流阀门,从而有效降低了控制信号的传输时延.
由于布线网格上的2个控制阀门之间可能存在多条路径,为了减少控制逻辑的控制通道总长度,本文基于式(4)附加一个正向奖励$ c_{\text{p}}^3 $,从而鼓励智能体引导控制路径在靠近目标点的同时尽可能地使用网格上先前其他控制路径所占用的区域,即共享点$ {p^{\text{s}}} $. 例如,控制路径50的共享点为图7中的右下对角线灰底网纹方格. 通过该方式,智能体将有效规划控制通道以实现不同控制路径之间的资源共享,从而降低控制逻辑的总成本. 此外,从该奖励函数可能获得的奖励值可以看出,如果智能体选择了一个远离关键点的共享点,那么其获得的惩罚(“−4”)甚至大于智能体选择一个靠近关键点的普通点所获得的惩罚(“−2”),因此智能体在整个奖励函数的引导下更偏向于首先确保控制路径长度的减少,然后再考虑通道资源的复用. 换句话说,优化信号传输时延的优先级整体上是大于优化通道总长度的.
基于上述所设计的状态空间、动作空间以及奖励函数,本文采用了DRL方法中的DuelingDQN网络架构[26]作为智能体对信号传输时延和通道总长度进行权衡评估,从而实现控制通道的自动化布线. 为了学习到一种有效的布线策略,智能体需要在训练过程中计算时刻t下状态动作对$ ({s_t},{a_t}) $的Q值,并对所有状态动作对的Q值进行迭代更新. 最终,智能体对各时刻下所感知的状态分别选择具有最大Q值的动作来组成布线策略.
表1展示了本文所采用DuelingDQN架构的基本配置,包括网络层数、网络类型、神经元数量以及各网络层使用的激活函数,其中输入层与输出层的神经元数量即为各自对应的维度. 输入层的7个维度分别对应智能体状态中包含的7个元素. 此外,输出层的5个维度则分别对应智能体可选择的5个布线方向. DuelingDQN通过改变架构中称为目标网络的神经网络结构来优化算法,它将用于计算时刻t下状态动作对$ ({s_t},{a_t}) $的Q值动作值函数拆分成了与状态相关的价值函数V,以及与动作相关的优势函数D. 具体来说,其主要是在目标网络的全连接层使用2个分离流,从而对状态值函数和优势函数进行单独计算,最后将2个分离流进行汇总以输出单一Q值. 因此针对Q值的计算方式可以表示成
$ Q(s_t,a_t,\theta_t,\alpha_t,\beta_t)=V(s_t,\theta_t,\beta_t)+D(s_t,a_t,\theta_t,\alpha_t), $
其中$ V({s_t},{\theta _t},{\beta _t}) $和$ D({s_t},{a_t},{\theta _t},{\alpha _t}) $分别是价值函数和优势函数的输出. $ {\theta _t} $指的是时刻t下目标网络的参数,$ {\beta _t} $是价值函数部分的参数,$ {\alpha _t} $是优势函数部分的参数. 可以看出$ V({s_t},{\theta _t},{\beta _t}) $仅与状态有关,而$ D({s_t},{a_t},{\theta _t},{\alpha _t}) $与状态和动作都有关. 但是式(6)的Q值计算方式无法辨识最终输出里面$ V({s_t},{\theta _t},{\beta _t}) $和$ D({s_t},{a_t},{\theta _t},{\alpha _t}) $各自的作用,从而严重地降低了神经网络的性能.
为了解决上述不可辨识性的问题,将优势函数部分引入平均操作,因此最终Q值的计算方式为
$ \begin{split} & Q(s_t,a_t,\theta_t,\alpha_t,\beta_t)=V(s_t,\theta_t,\beta_t)+ \\ & \quad\left(D(s_t,a_t,\theta_t,\alpha_t)-\sum\limits_{a\in A}^{ }D\left(s,a,\theta_t,\alpha_t\right)/|A|\right),\end{split} $
其中$ |A| $表示动作空间A中的动作总数,且该Q值也称为目标网络的预测Q值.
然后,目标网络利用动作$ {a_t} $以及下一个时刻状态$ {s_{t + 1}} $来计算状态动作对$ ({s_t},{a_t}) $的目标Q值$ {G_t} $. 需要注意的是,智能体在状态$ {s_t} $下采取动作$ {a_t} $就能够获得下一时刻的状态$ {s_{t + 1}} $,即使当前状态$ {s_t} $并未进行转移. 因此,状态动作对$ ({s_t},{a_t}) $的目标Q值计算方式为
$ G_t=r_t+\gamma Q(s_{t+1},a_t,\theta_t,\alpha_t,\beta_t), $
其中$ {r_t} $是在状态$ {s_t} $下采取动作$ {a_t} $后获得的奖励,$ \gamma $表示折扣因子,$ {G_t} $用于评估所采取动作$ {a_t} $的质量. 通常来说,$ \gamma $用于权衡当前和未来获得的奖励. 当$ \gamma $接近1时,智能体更关注长期回报;否则,智能体更关注即时奖励[22].
为了使智能体在学习过程中不断向正确的策略调整,需要基于上述预测Q值以及目标Q值对目标网络的参数进行更新,因此需要为神经网络构建损失函数$ L(\theta ) $为:
$ L(\theta ) = E[{({G_t} - Q({s_t},{a_t},{\theta _t},{\alpha _t},{\beta _t}))^2}]. $
然后,通过采用梯度下降法来最小化该损失函数[25],从而实现目标网络中参数的更新.
基于上述技术细节,具体的算法如算法1所示.
算法1. 基于深度强化学习的控制逻辑布线算法.
输入:控制阀门位置,流阀门位置,模式序列;
输出:最小化信号传输时延以及通道总长度的控制逻辑布线方案.
① 根据输入信息构建并初始化布线网格$ {G_{\text{r}}} $;
② 初始化目标网络参数$ \theta $;
③ 初始化训练批次$ {B_{\text{S}}} $,训练回合$ {E_{\text{P}}} $,学习率 $ \partial $, 动作选择概率$ \varepsilon $以及目标网络更新周期$ {U_{\text{c}}} $;
④ for 每一个训练回合 $ {E_{\text{P}}} $
⑤ 重置环境参数;
⑥ if 状态$ {s_t} $不是终止状态
⑦  随机生成概率值$ {V_{\text{P}}} $;
⑧  if $ {V_{\text{P}}} $>$ \varepsilon $
⑨   从动作空间A中随机选择动作$ {a_t} $;
⑩  else
⑪   选择Q值最大的动作$ {a_t} $;
⑫  end if
⑬  执行$ {a_t} $并根据式(4)计算对应奖励$ {r_t} $;
⑭  获取下一时刻状态$ {s_{t + 1}} $;
⑮  元组$ ({s_t},{\text{ }}{a_t},{\text{ }}{r_t},{\text{ }}{s_{t + 1}}) $存入经验回放区;
⑯  if $ {E_{\text{P}}} $mod$ {U_{\text{c}}} $= 0
⑰   从经验回放区随机选取$ {B_{\text{S}}} $个元组;
⑱   采用$ \partial $以最小化式(9)来更新$ \theta $;
⑲  end if
⑳  将状态$ {s_t} $更新为下一时刻状态$ {s_{t + 1}} $;
㉑ end if
㉒ end for
从上述算法步骤中可知,首先需要根据模式序列以及所有阀门的位置来构建并初始化布线网格$ {G_{\text{r}}} $. 其次,需要对DRL布线方法中的网络参数以及使用的超参数进行初始化. 然后,智能体感知状态$ {s_t} $,并采用$ \varepsilon $贪婪策略来选取动作. 具体来说,智能体将以$ \varepsilon $的概率选择具有最大Q值的布线方向,否则将从动作空间A随机选择1个布线方向. 利用该策略,智能体能够更好地权衡动作的开发和探索,从而更有机会探索到最优策略. 当执行动作$ {a_t} $后,智能体将根据式(4)获得对应的奖励信号,并相应获得下一时刻状态$ {s_{t + 1}} $. 基于上述结果,元组$ ({s_t},{\text{ }}{a_t},{\text{ }}{r_t},{\text{ }}{s_{t + 1}}) $将被存入经验回放区. 当智能体需要学习新策略时,即此时目标网络的参数需要更新,那么将从经验回放区中随机选取$ {B_{\text{S}}} $个元组作为学习样本以最小化式(9)表示的损失函数. 之后,智能体感知下一时刻状态$ {s_{t + 1}} $以开始下一轮的环境交互. 当完成所有指定的学习迭代后,智能体将生成一个最小化信号传输时延以及通道总长度的控制逻辑布线方案.
本文所提出的基于深度强化学习的连续微流控生物芯片控制逻辑布线算法采用TensorFlow 1.13.1框架实现,并在具有2.40 GHz和8 GB内存的设备上进行测试. 6个测试用例用于验证本文算法的有效性,其中RC01是用于实现RA30芯片中比色蛋白分析[27]的布局测试用例,RC02是用于实现CPA芯片中体外诊断分析[27]的布局测试用例,RC03是用于实现mRNA芯片中体外诊断分析[28]的布局测试用例,而RC04,RC05,RC06则分别对应3种随机合成分析的布局测试用例. 表2列出了测试用例的相关信息,其中从左到右分别给出了控制逻辑占用的网格大小、CFMBs中流阀门数量、控制逻辑中控制阀门数量. 此外,表3给出了本文算法提及的相关参数取值.
由于当前还未有针对CFMBs控制逻辑通道布线的相关工作,因此本文实现了另一种控制逻辑布线算法,称为DRA,以验证本文算法的有效性. 具体来说,DRA是基于A*算法来实现逻辑架构中控制通道网络的构建,且该算法同样以控制信号传输时延以及控制通道总长度作为优化目标. 需要注意的是,后续实验所评估的控制信号传输时延是由最长控制路径引起的延迟,即最大信号传输时延. 本文在上述测试用例上分别运行这2种算法,表4和表5分别展示了本文算法与DRA算法在控制通道总长度与信号传输时延指标上的对比结果,其中“优化率”栏展示了本文算法相对于DRA算法在各指标上的优化程度.
从表4看出,相对于DRA算法,本文算法在控制通道总长度方面实现了10.9%~13.9%的长度缩减,平均减少了12.3%. 此外,从表5可以看出,本文算法相对于DRA算法具有更好的时延优化效果,在各测试用例上本文算法实现了11.1%~22.2%的时延降低,平均降低了17.3%. 另外,图8展示了本文算法与DRA针对测试用例RC02分别生成的局部控制逻辑布线结果,从而更直观地表现2种方法在布线效果上的差别. 上述对比实验结果表明了本文所提布线算法的有效性,这主要得益于本文算法中的DRL智能体倾向于以长远的眼光在布线过程中自适应地调整布线策略,并根据环境的奖励反馈以有效地规划各控制路径的长度,从而降低了控制信号的传输时延. 同时该智能体能够通过先前的布线经验来最大程度地对不同控制路径的通道资源进行资源共享,从而有效地减少了控制通道的总长度,一定程度上保证了最终布线方案的质量.
为了验证本文算法具有一定的泛化能力,我们选择测试用例RC01~RC04对单个模型进行训练,并使用训练完成的模型对RC05和RC06进行测试,然后将生成的布线结果与DRA在控制通道总长度以及信号传输时延方面进行对比,对比结果如表6所示. 从表6看出,相对于DRA算法,本文算法在这2个测试用例上分别实现了10.5%以及9.1%的通道长度缩减. 此外,本文算法相对于DRA算法仍具有更好的时延优化效果,在这2个测试用例上分别实现了18.2%以及20.0%的时延降低. 上述实验结果表明了本文算法具有一定的泛化能力,其针对未知的测试用例仍能够生成有效的布线方案.
在所提出的基于DRL的控制逻辑布线算法中,采用了一种对末层全连接层进行拆分的DuelingDQN架构作为DRL智能体,从而高效地实现智能体的学习过程. 为了验证DuelingDQN架构的优越性,本文引入了另外2种DRL架构以代替DuelingDQN,即深度双Q网络(double deep Q-network,DoubleDQN)[29]和深度Q网络(deep Q-network,DQN)[30],其中DoubleDQN是在DQN的基础上提出的一种改进模型. 具体来说,DoubleDQN通过使用2个神经网络来减少Q值的过高估计问题,其中策略网络用于选择动作,而目标网络则用于评价所选动作的质量. 本文分别比较了3种网络架构在控制通道总长度和信号传输时延的实验结果,并在表7和表8进行展示,其中“优化率”栏分别展示了DuelingDQN架构相对于其他2种DRL架构在各指标上的优化程度.
从表7可以看出,DuelingDQN优于基础的DQN架构,在控制通道总长度方面实现了7.3%~13.4%的长度缩减,平均缩减了10.3%. 此外,与改进后的网络架构DoubleDQN相比,DuelingDQN在控制通道总长度方面仍然实现了平均5.1%的优化. 另一方面,如表8所示,DuelingDQN在信号传输时延方面同样优于其余2个DRL架构,尤其是与基础的DQN架构相比,本文算法所采用的DuelingDQN架构能够实现11.1%~22.2%的信号传输时延降低,平均降低了15.6%. 而与DoubleDQN相比,DuelingDQN同样实现了信号传输时延的优化,且在所有测试用例上能够实现平均5.8%的传输时延降低. 上述结果表明了所采用的DuelingDQN架构具有更优异的性能,这主要是因为DuelingDQN架构相比于另外2种架构能够分别对状态价值和动作价值进行估计以计算更为准确的Q值,使得DRL智能体的学习过程更加稳定高效,从而获得高质量的布线方案.
为了验证在控制逻辑中采用对角型通道布线的有效性,本文基于所提出的DRL框架实现了另一种未采用对角型布线方式的算法MHR(仅采用曼哈顿布线方式),并与本文算法(采用对角型布线方式的算法)分别在控制通道总长度以及信号传输时延2方面进行对比. 对比结果如表9和表10所示,其中“优化率”栏展示了采用对角型布线方式相对于未采用对角型布线方式在上述各指标上的优化程度.
从表9可以看出,在相同的DRL框架下,采用对角型布线方式相对于未采用对角型布线方式在各测试用例上能够实现19.3%~22.6%的控制通道总长度缩减,平均缩减了20.8%. 此外,如表10所示,采用对角型的布线方式在信号传输时延方面仍优于未采用对角型的布线方式,实现了22.2%~36.0%的传输时延降低,平均降低了28.9%. 上述实验结果表明了在控制逻辑中采用对角型布线方式的有效性. 需要注意的是,虽然对角型布线方式能够有效地优化控制通道总长度以及信号传输时延,但是如何指导该布线方式的实施本质上是基于本文所提出的深度强化学习框架,从而能够生成高质量的控制逻辑布线方案. 此外,在3.1节也证明了所提出的深度强化学习方法的有效性. 因此本文中所采用的深度强化学习方法以及对角型布线方式都是使得最终实验结果质量提升的关键因素,二者缺一不可,前者基于后者以进一步提升实验结果的质量,而后者需要依赖于前者才能够有效发挥其自身作用.
另一方面,表11列出了本文算法针对每个测试用例所需的训练时间和推理时间,以及2种基线算法DRA和MHR的运行时间,其中推理时间指的是训练完成后的模型对各测试用例生成有效布线方案所需的时间. 需要注意的是,表11中基线算法MHR的运行时间为训练完成后模型的推理时间. 从表11中可以看出,虽然本文算法在模型训练阶段需要花费较多的时间,但训练好的模型能够在较短的计算时间内针对各测试用例生成高质量的布线方案,从而体现了本文算法在执行效率方面的有效性.
本文针对CFMBs控制逻辑研究了同时考虑信号传输时延和通道总长度的控制通道网络构建问题,并提出了第1个基于深度强化学习的连续微流控生物芯片控制逻辑布线算法. 该算法采用DuelingDQN架构实现了对信号传输时延和通道总线长的权衡评估,旨在自动构建一个高效的控制通道网络. 此外,通过引入对角型的控制通道布线,为控制逻辑的通道布线阶段提供了巨大的灵活性,从而进一步降低了芯片的制造成本并提高其执行效率. 6个测试用例被用于验证本文算法的有效性. 实验结果表明,本文算法能够有效构建高性能、低成本的控制逻辑架构.
作者贡献声明:蔡华洋提出初步方案、实验设计,撰写和修改论文;黄兴负责论文思路构建、理论指导和论文修改;刘耿耿负责论文方案分析、理论指导、论文润色和修改.

参考文献 引证文献
排序方式:
1
Huang Xing, Ho T, Guo Wenzhong, et al. Computer-aided design techniques for flow-based microfluidic lab-on-a-chip systems[J]. ACM Computing Surveys, 2021, 54(5): 1−29
2
Liu Genggeng, Huang Hongbin, Chen Zhisheng, et al. Design automation for continuous-flow microfluidic biochips: A comprehensive review[J]. Integration, 2022, 82: 48−66
3
Hu Xu, Chen Zhen, Chen Zhisheng, et al. Architectural synthesis of continuous-flow microfluidic biochips with connection pair optimization[J]. Electronics, 2024, 13(2): 247
4
Einav S, Gerber D, Bryson P D, et al. Discovery of a hepatitis C target and its pharmacological inhibitors by microfluidic affinity analysis[J]. Nature Biotechnology, 2008, 26(9): 1019−1027
5
Hong J W, Chen Yan, Anderson W F, et al. Molecular biology on a microfluidic chip[J]. Journal of Physics: Condensed Matter, 2006, 18(18): 691−701
6
Unger M A, Chou H P, Thorsen T, et al. Monolithic microfabricated valves and pumps by multilayer soft lithography[J]. Science, 2000, 288(5463): 113−116
7
Thorsen T, Maerkl S J, Quake S R. Microfluidic large-scale integration[J]. Science, 2002, 298(5593): 580−584
8
Huang Xing, Ho T, Chakrabarty K, et al. Timing-driven flow-channel network construction for continuous-flow microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2019, 39(6): 1314−1327
9
Araci I E, Quake S R. Microfluidic very large scale integration (mVLSI) with integrated micromechanical valves[J]. Lab on a Chip, 2012, 12(16): 2803−2806
10
Fidalgo L M, Maerkl S J. A software-programmable microfluidic device for automated biology[J]. Lab on a Chip, 2011, 11(9): 1612−1619
11
Lim Y C, Kouzani A Z, Duan W. Lab-on-a-chip: A component view[J]. Microsystem Technologies, 2010, 16: 1995−2015
12
Tseng T M, Li Mengchu, Freitas D N, et al. Columba 2.0: A co-layout synthesis tool for continuous-flow microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2017, 37(8): 1588−1601
13
Yang Kailin, Yao Hailong, Ho T, et al. AARF: Any-angle routing for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2018, 37(12): 3042−3055
14
Wang Qin, Zuo Shiliang, Yao Hailong, et al. Hamming-distance-based valve switching optimization for control-layer multiplexing in flow-based microfluidic biochips[C]//Proc of the 22nd Asia and South Pacific Design Automation Conf. Piscataway, NJ: IEEE, 2017: 524−529
15
Wang Qin, Xu Yue, Zuo Shiliang, et al. Pressure-aware control layer optimization for flow-based microfluidic biochips[J]. IEEE Transactions on Biomedical Circuits and Systems, 2017, 11(6): 1488−1499
16
Zhu Ying, Huang Xing, Li Bing, et al. Multicontrol: Advanced control-logic synthesis for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2019, 39(10): 2489−2502
17
Huang Xing, Cai Huayang, Guo Wenzhong, et al. Control-logic synthesis of fully programmable valve array using reinforcement learning[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2024, 43(1): 277−290
18
Liang Siyuan, Li Mengchu, Tseng T M, et al. CoMUX: Combinatorial-coding-based high-performance microfluidic control multiplexer design[C/OL]//Proc of the 41st ACM Int Conf on Computer-Aided Design. New York: ACM, 2022 [2024-06-19]. https://dl.acm.org/doi/abs/10.1145/3508352.3549353
19
Chen Zhisheng, Huang Xing, Guo Wenzhong, et al. Physical synthesis of flow-based microfluidic biochips considering distributed channel storage[C]//Proc of Design, Automation &amp; Test in Europe Conf &amp; Exhibition. Piscataway, NJ: IEEE, 2019: 1525−1530
20
Hu Kai, Dinh T A, Ho T, et al. Control-layer routing and control-pin minimization for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2016, 36(1): 55−68
21
曾俊杰,秦龙,徐浩添,等,基于内在动机的深度强化学习探索方法综述[J]. 计算机研究与发展,2023,60(10):2359−2382
Zeng Junjie, Qin Long, Xu Haotian, et al. Exploration approaches in deep reinforcement learning based on intrinsic motivation: A review[J]. Journal of Computer Research and Development, 2023, 60(10): 2359−2382 (in Chinese)
22
Sutton R S, Barto A G. Reinforcement learning: An introduction[M]//Adaptive Computation and Machine Learning. Cambridge, MA: MIT, 2018
23
唐楚哲,王肇国,陈海波. 机器学习方法赋能系统软件:挑战、实践与展望[J]. 计算机研究与发展,2023,60(5):964−973
Tang Chuzhe, Wang Zhaoguo, Chen Haibo. Empowering system software with machine learning methods: Challenges, practice, and prospects[J]. Journal of Computer Research and Development, 2023, 60(5): 964−973 (in Chinese)
24
Kober J, Bagnell J A, Peters J. Reinforcement learning in robotics: A survey[J]. The International Journal of Robotics Research, 2013, 32(11): 1238−1274
25
Mnih V, Kavukcuoglu K, Silver D, et al. Human-level control through deep reinforcement learning[J]. Nature, 2015, 518(7540): 529−533
26
Wang Ziyu, Schaul T, Hessel M, et al. Dueling network architectures for deep reinforcement learning[C]//Proc of the 33rd Int Conf on Machine Learning. Cambridge, MA: MIT, 2016: 1995−2003
27
Liu Chunfeng, Huang Xing, Li Bing, et al. DCSA: Distributed channel-storage architecture for flow-based microfluidic biochips[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2020, 40(1): 115−128
28
Marcus J S, Anderson W F, Quake S R. Microfluidic single-cell mRNA isolation and analysis[J]. Analytical Chemistry, 2006, 78(9): 3084−3089
29
Van Hasselt H, Guez A, Silver D. Deep reinforcement learning with double q-learning[C]//Proc of the 30th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2016: 2094−2100
30
Mnih V, Kavukcuoglu K, Silver D, et al. Playing Atari with deep reinforcement learning[J]. arXiv preprint, arXiv:1312.560, 2013
2025年第62卷第4期
124
53
0
引用本文
BibTeX
文章信息
doi: 10.7544/issn1000-1239.202440034
  • 接收时间:2024-01-18
  • 首发时间:2026-11-10
  • 出版时间:2025-04-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-01-18
  • 修回日期:2024-08-12
  • 录用日期:2024-09-03
基金
作者信息
    1福州大学计算机与大数据学院 福州 350116
    2大数据智能教育部工程研究中心(福州大学) 福州 350116
    3福建省网络计算与智能信息处理重点实验室(福州大学) 福州 350116
    4西北工业大学计算机学院 西安 710072

通讯作者:

刘耿耿()
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202440034
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
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
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