Article(id=1274300228095205417, tenantId=1146029695717560320, journalId=1272208980697911299, issueId=1274300092707266809, articleNumber=null, orderNo=null, doi=10.3724/1000-6915.jrme.2025.0477, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1751472000000, receivedDateStr=2025-07-03, revisedDate=1758902400000, revisedDateStr=2025-09-27, acceptedDate=null, acceptedDateStr=null, onlineDate=1781746449731, onlineDateStr=2026-06-18, pubDate=1769875200000, pubDateStr=2026-02-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1781746449731, onlineIssueDateStr=2026-06-18, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1781746449731, creator=13701087609, updateTime=1781746449731, updator=13701087609, issue=Issue{id=1274300092707266809, tenantId=1146029695717560320, journalId=1272208980697911299, year='2026', volume='45', issue='2', pageStart='321', pageEnd='638', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1781746417452, creator=13701087609, updateTime=1781746463571, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1274300286466335306, tenantId=1146029695717560320, journalId=1272208980697911299, issueId=1274300092707266809, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1274300286466335307, tenantId=1146029695717560320, journalId=1272208980697911299, issueId=1274300092707266809, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=553, endPage=577, ext={EN=ArticleExt(id=1274300228304920619, articleId=1274300228095205417, tenantId=1146029695717560320, journalId=1272208980697911299, language=EN, title=Dynamic prediction model of tunnel surrounding rock deformation based on Bayesian network, columnId=null, journalTitle=Chinese Journal of Rock Mechanics and Engineering, columnName=null, runingTitle=null, highlight=null, articleAbstract=

Significant limitations and hysteresis are presented in dynamic prediction methods driven by on-site monitored displacement data for tunnel surrounding rock deformation. By comprehensively utilizing the physical information contained in tunnel construction project documents and the mathematical information from displacement time-series curves, a modelling method based on the dynamic Bayesian network (DBN) was developed using the concept of physical information machine learning (PIML) to achieve dynamic predictions of surrounding rock deformation. Through discretization processing and reconstruction of displacement time-series curves, a static sample database was established by combining physical information data with ultimate displacement data, while a dynamic sample database was created by integrating physical information data with displacement time-series curve data. Based on the characteristics of the static samples, the K2-score algorithm was improved to construct a static Bayesian network (BN) model for ultimate displacement prediction. Utilizing the static BN model and the characteristics of the dynamic samples, physical-data dual-drive modelling methods for the Markov DBN were derived by incorporating prior information, including the constraints of steady-state random processes and Markov process constraints. By integrating prior information for constraint-enhanced optimization, the optimized Markov DBN model was established. Five-fold cross-validation tests revealed that the prediction capability of the Markov DBN model decreased rapidly over time and that the network transition direction significantly affected this capability. In contrast, the prediction ability of the optimized Markov DBN model remained robust over time, was unaffected by the network transition direction, and significantly exceeded that of the Markov DBN model, as the optimized model enhanced constraint connections between target nodes and influencing factor nodes throughout the entire timeframe. Through engineering case analysis, it was concluded that before and during the early stages of tunnel construction, the optimized Markov DBN model could effectively predict displacement time-series curves, overcoming the limitations and hysteresis inherent in traditional methods. Furthermore, during construction, self-updating of the optimized Markov DBN model and dynamic predictions of surrounding rock deformation could be achieved by inputting the on-site monitored displacement data.

, correspAuthors=Chao ZHANG, authorNote=null, correspAuthorsNote=
* ZHANG Chao (1978–), research fellow, is engaged in geotechnical engineering disaster prevention and control. E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Hongxing WANG, Keyao LI, Chao ZHANG, Junhao RUAN, Liping WANG, Wei LIU, Shangwei WU), CN=ArticleExt(id=1274300232662802505, articleId=1274300228095205417, tenantId=1146029695717560320, journalId=1272208980697911299, language=CN, title=基于贝叶斯网络的隧道围岩变形动态预测模型, columnId=1274300130443420099, journalTitle=岩石力学与工程学报, columnName=数值模拟与人工智能, runingTitle=null, highlight=null, articleAbstract=

基于现场监测位移数据驱动的隧道围岩变形动态预测方法,存在着较大的局限性和滞后性。综合利用隧道工程建设资料蕴含的物理信息和位移时序曲线蕴含的数学信息,通过物理信息机器学习的思想推导动态贝叶斯网络(dynamic Bayesian network,DBN)模型的构建方法,实现围岩变形的动态预测。通过离散化处理、位移时序曲线重构等方式,构建包含物理信息数据和极限位移数据的静态样本库,及包含物理信息数据和位移时序曲线数据的动态样本库。基于静态样本特征,改进K2-score算法建立极限位移预测的静态BN模型。基于静态BN模型和动态样本特征,融合平稳随机过程约束和Markov过程约束等先验信息,推导Markov-DBN的物理数据双驱动建模方法。融合约束增强优化的先验信息,构建优化Markov-DBN模型。5折交叉验证试验表明:Markov-DBN模型的预测能力随时间增加而快速降低,网络转移方向对其影响很大;优化Markov-DBN模型建立了所有时刻目标节点与变形影响因素节点的增强约束联系,其预测能力不随时间发生明显弱化,不受网络转移方向的影响,远高于Markov-DBN模型。案例分析表明:隧道工程施工前和施工初期,优化Markov-DBN模型即可实现位移时序曲线的有效预测,克服了传统方法的局限性和滞后性;隧道施工过程中,实时输入现场监测位移数据,可实现优化Markov-DBN模型的自我更新和围岩变形的动态预测。

, correspAuthors=张超, authorNote=null, correspAuthorsNote=
* 张超(1978–),现任研究员,主要从事岩土工程灾害防控方面的研究工作。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=YihmhKKIbJmd6NzSIM2R4w==, magXml=0uX/az9KUKgPoGnno+50pg==, pdfUrl=null, pdf=rK4YO47b4XlbgJAFbdy6dQ==, pdfFileSize=843840, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=HgpLD/OBpoeikMSDb4nW4Q==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=2ZNWSY8bDE0SGHrGZzqBwQ==, mapNumber=null, authorCompany=null, fund=null, authors=

WANG Hongxing (1983–), associate professor, is engaged in engineering disaster prediction and prevention based on machine learning algorithms. E-mail:

汪洪星(1983–),现任副教授,主要从事基于机器学习算法的工程灾害预测和防控方面的研究工作。E-mail:

, authorsList=汪洪星, 李珂瑶, 张超, 阮俊浩, 王丽萍, 刘伟, 巫尚蔚)}, authors=[Author(id=1274369016906965577, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=2021078@cqust.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1274369017007628875, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369016906965577, language=EN, stringName=Hongxing WANG, firstName=Hongxing, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369017401893452, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369016906965577, language=CN, stringName=汪洪星, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.重庆科技大学 安全科学与工程学院,重庆 400041, bio={"content":"

WANG Hongxing (1983–), associate professor, is engaged in engineering disaster prediction and prevention based on machine learning algorithms. E-mail:

汪洪星(1983–),现任副教授,主要从事基于机器学习算法的工程灾害预测和防控方面的研究工作。E-mail:

"}, bioImg=null, bioContent=

WANG Hongxing (1983–), associate professor, is engaged in engineering disaster prediction and prevention based on machine learning algorithms. E-mail:

汪洪星(1983–),现任副教授,主要从事基于机器学习算法的工程灾害预测和防控方面的研究工作。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016286208573, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=1., ext=[AuthorCompanyExt(id=1274369016294597182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China), AuthorCompanyExt(id=1274369016302985791, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.重庆科技大学 安全科学与工程学院,重庆 400041)])]), Author(id=1274369017859072590, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, 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=1274369018702127696, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369017859072590, language=EN, stringName=Keyao LI, firstName=Keyao, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369019083809361, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369017859072590, language=CN, stringName=李珂瑶, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.重庆科技大学 安全科学与工程学院,重庆 400041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016286208573, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=1., ext=[AuthorCompanyExt(id=1274369016294597182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China), AuthorCompanyExt(id=1274369016302985791, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.重庆科技大学 安全科学与工程学院,重庆 400041)])]), Author(id=1274369019872338515, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=czhang@whrsm.ac.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1274369020644090453, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369019872338515, language=EN, stringName=Chao ZHANG, firstName=Chao, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2.State Key Laboratory of Geomechanics and Geotechnical Engineering Safety, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan, Hubei 430071, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369021159989846, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369019872338515, language=CN, stringName=张超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2.中国科学院武汉岩土力学研究所 岩土力学与工程安全全国重点实验室,湖北 武汉 430071, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016357511744, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=2., ext=[AuthorCompanyExt(id=1274369016378483265, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016357511744, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.State Key Laboratory of Geomechanics and Geotechnical Engineering Safety, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan, Hubei 430071, China), AuthorCompanyExt(id=1274369016391066178, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016357511744, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中国科学院武汉岩土力学研究所 岩土力学与工程安全全国重点实验室,湖北 武汉 430071)])]), Author(id=1274369022036599384, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1274369023173255770, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369022036599384, language=EN, stringName=Junhao RUAN, firstName=Junhao, middleName=null, lastName=RUAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369023454274139, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369022036599384, language=CN, stringName=阮俊浩, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.重庆科技大学 安全科学与工程学院,重庆 400041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016286208573, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=1., ext=[AuthorCompanyExt(id=1274369016294597182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China), AuthorCompanyExt(id=1274369016302985791, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.重庆科技大学 安全科学与工程学院,重庆 400041)])]), Author(id=1274369023802401373, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1274369024557376095, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369023802401373, language=EN, stringName=Liping WANG, firstName=Liping, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369024649650784, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369023802401373, language=CN, stringName=王丽萍, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.重庆科技大学 安全科学与工程学院,重庆 400041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016286208573, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=1., ext=[AuthorCompanyExt(id=1274369016294597182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China), AuthorCompanyExt(id=1274369016302985791, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.重庆科技大学 安全科学与工程学院,重庆 400041)])]), Author(id=1274369025295573602, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, orderNo=5, 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=1274369026176377444, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369025295573602, language=EN, stringName=Wei LIU, firstName=Wei, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3.School of Resource and Safety Engineering, Chongqing University, Chongqing 400044, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369026516116069, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369025295573602, language=CN, stringName=刘伟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3.重庆大学 资源与安全学院,重庆 400044, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016592392772, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=3., ext=[AuthorCompanyExt(id=1274369016609169989, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016592392772, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.School of Resource and Safety Engineering, Chongqing University, Chongqing 400044, China), AuthorCompanyExt(id=1274369016621752902, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016592392772, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.重庆大学 资源与安全学院,重庆 400044)])]), Author(id=1274369026629362279, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, orderNo=6, 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=1274369027875070569, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369026629362279, language=EN, stringName=Shangwei WU, firstName=Shangwei, middleName=null, lastName=WU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1274369028256752234, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, authorId=1274369026629362279, language=CN, stringName=巫尚蔚, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.重庆科技大学 安全科学与工程学院,重庆 400041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1274369016286208573, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=1., ext=[AuthorCompanyExt(id=1274369016294597182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China), AuthorCompanyExt(id=1274369016302985791, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.重庆科技大学 安全科学与工程学院,重庆 400041)])])], keywords=[Keyword(id=1274369029494071915, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, orderNo=1, keyword=tunnel engineering), Keyword(id=1274369030290989676, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, orderNo=2, keyword=dynamic prediction of surrounding rock deformation), Keyword(id=1274369030379070061, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, orderNo=3, keyword=multi-source information fusion), Keyword(id=1274369030756557422, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, orderNo=4, keyword=dynamic Bayesian network), Keyword(id=1274369031163404911, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, orderNo=5, keyword=physical-data hybrid modelling), Keyword(id=1274369031545086576, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, orderNo=6, keyword=model self-update), Keyword(id=1274369032069374577, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, orderNo=1, keyword=隧道工程), Keyword(id=1274369032807572082, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, orderNo=2, keyword=围岩变形动态预测), Keyword(id=1274369033206030963, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, orderNo=3, keyword=多源信息融合), Keyword(id=1274369033675793012, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, orderNo=4, keyword=动态贝叶斯网络), Keyword(id=1274369034061668981, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, orderNo=5, keyword=物理数据混合建模), Keyword(id=1274369034502070902, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, orderNo=6, keyword=模型自我更新)], refs=[Reference(id=1274369044396434089, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=46, issue=增1, pageStart=833, pageEnd=836, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=郑艾辰, 黄锋, 林志, journalName=施工技术, refType=null, unstructuredReference=郑艾辰,黄锋,林志,. 2008年至2016年我国隧道工程施工安全事故统计与分析[J]. 施工技术201746(增1):833–836., articleTitle=2008年至2016年我国隧道工程施工安全事故统计与分析, refAbstract=null), Reference(id=1274369044648092330, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=46, issue=Supp.1, pageStart=833, pageEnd=836, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=ZHENG Aichen, HUANG Feng, LIN Zhi, journalName=Construction Technology, refType=null, unstructuredReference=(ZHENG AichenHUANG FengLIN Zhiet al. The statistics and analysis of safety accident during tunnel engineering construction in China from 2008 to 2016[J]. Construction Technology201746(Supp.1):833–836.(in Chinese)), articleTitle=The statistics and analysis of safety accident during tunnel engineering construction in China from 2008 to 2016, refAbstract=null), Reference(id=1274369044954276523, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=31, issue=6, pageStart=1 590, pageEnd=1 600, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=吕擎峰, 李钰, 牛荣, journalName=应用基础与工程科学学报, refType=null, unstructuredReference=吕擎峰,李钰,牛荣,. 基于深度学习的特殊岩土隧道围岩变形预测研究[J]. 应用基础与工程科学学报202331(6):1 590–1 600., articleTitle=基于深度学习的特殊岩土隧道围岩变形预测研究, refAbstract=null), Reference(id=1274369045029773996, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=31, issue=6, pageStart=1 590, pageEnd=1 600, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=LÜ Qingfeng, LI Yu, NIU Rong, journalName=Journal of Fundamental and Engineering Science, refType=null, unstructuredReference=( QingfengLI YuNIU Ronget al. Research on deformation prediction of surrounding rock in special geotechnical tunnels based on deep learning[J]. Journal of Fundamental and Engineering Science202331(6):1 590–1 600.(in Chinese)), articleTitle=Research on deformation prediction of surrounding rock in special geotechnical tunnels based on deep learning, refAbstract=null), Reference(id=1274369045101077165, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=43, issue=7, pageStart=12, pageEnd=20, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=孙振宇, 张顶立, 侯艳娟, journalName=岩土工程学报, refType=null, unstructuredReference=孙振宇,张顶立,侯艳娟,. 基于现场实测数据统计的隧道围岩全过程变形规律及稳定性判据确定[J]. 岩土工程学报202143(7):12–20., articleTitle=基于现场实测数据统计的隧道围岩全过程变形规律及稳定性判据确定, refAbstract=null), Reference(id=1274369045163991726, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=43, issue=7, pageStart=12, pageEnd=20, url=null, language=null, rfNumber=[3], rfOrder=5, authorNames=SUN Zhenyu, ZHANG Dingli, HOU Yanjuan, journalName=Chinese Journal of Geotechnical Engineering, refType=null, unstructuredReference=(SUN ZhenyuZHANG DingliHOU Yanjuanet al. Whole-process deformation laws and determination of stability criterion of surrounding rock of tunnels based on statistics of field measured data[J]. Chinese Journal of Geotechnical Engineering202143(7):12–20.(in Chinese)), articleTitle=Whole-process deformation laws and determination of stability criterion of surrounding rock of tunnels based on statistics of field measured data, refAbstract=null), Reference(id=1274369045231100591, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=5, pageStart=1 024, pageEnd=1 035, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=王述红, 董福瑞, journalName=岩土工程学报, refType=null, unstructuredReference=王述红,董福瑞. 基于变形预测和参数反演的山岭隧道围岩稳定性分析[J]. 岩土工程学报202345(5):1 024–1 035., articleTitle=基于变形预测和参数反演的山岭隧道围岩稳定性分析, refAbstract=null), Reference(id=1274369045373706928, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=5, pageStart=1 024, pageEnd=1 035, url=null, language=null, rfNumber=[4], rfOrder=7, authorNames=WANG Shuhong, DONG Furui, journalName=Chinese Journal of Geotechnical Engineering, refType=null, unstructuredReference=(WANG ShuhongDONG Furui. Stability analysis of surrounding rock of mountain tunnels based on deformation prediction and parameter inversion[J]. Chinese Journal of Geotechnical Engineering202345(5):1 024–1 035.(in Chinese)), articleTitle=Stability analysis of surrounding rock of mountain tunnels based on deformation prediction and parameter inversion, refAbstract=null), Reference(id=1274369045491147441, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=144, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=8, authorNames=WANG Y, LIU Y, WANG Z F, journalName=Engineering Failure Analysis, refType=null, unstructuredReference=WANG YLIU YWANG Z Fet al. Investigation on progressive failure process of tunnel lining induced by creep effect of surrounding rock:A case study[J]. Engineering Failure Analysis2023144:106946., articleTitle=Investigation on progressive failure process of tunnel lining induced by creep effect of surrounding rock:A case study, refAbstract=null), Reference(id=1274369045608587954, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2019, volume=90, issue=null, pageStart=1, pageEnd=11, url=null, language=null, rfNumber=[6], rfOrder=9, authorNames=SHI S, ZHAO R, LI S, journalName=Tunnelling and Underground Space Technology, refType=null, unstructuredReference=SHI SZHAO RLI Set al. Intelligent prediction of surrounding rock deformation of shallow buried highway tunnel and its engineering application[J]. Tunnelling and Underground Space Technology201990:1–11., articleTitle=Intelligent prediction of surrounding rock deformation of shallow buried highway tunnel and its engineering application, refAbstract=null), Reference(id=1274369045826691763, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=1993, volume=26, issue=2, pageStart=137, pageEnd=163, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=AYDAN O, AKAGI T, KAWAMOTO T, journalName=Rock Mechanics and Rock Engineering, refType=null, unstructuredReference=AYDAN OAKAGI TKAWAMOTO T. The squeezing potential of rocks around tunnels:Theory and prediction[J]. Rock Mechanics and Rock Engineering199326(2):137–163., articleTitle=The squeezing potential of rocks around tunnels:Theory and prediction, refAbstract=null), Reference(id=1274369046090932916, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2019, volume=78, issue=1, pageStart=431, pageEnd=444, url=null, language=null, rfNumber=[8], rfOrder=11, authorNames=BIAN K, LIU J, LIU Z, journalName=Bulletin of Engineering Geology and the Environment, refType=null, unstructuredReference=BIAN KLIU JLIU Zet al. Mechanisms of large deformation in soft rock tunnels a case study of Huangjiazhai Tunnel[J]. Bulletin of Engineering Geology and the Environment201978(1):431–444., articleTitle=Mechanisms of large deformation in soft rock tunnels a case study of Huangjiazhai Tunnel, refAbstract=null), Reference(id=1274369046279676597, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2020, volume=42, issue=1, pageStart=81, pageEnd=90, url=null, language=null, rfNumber=[9], rfOrder=12, authorNames=王明年, 王志龙, 张霄, journalName=岩土工程学报, refType=null, unstructuredReference=王明年,王志龙,张霄,. 深埋隧道围岩形变压力计算方法研究[J]. 岩土工程学报202042(1):81–90., articleTitle=深埋隧道围岩形变压力计算方法研究, refAbstract=null), Reference(id=1274369046434865846, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2020, volume=42, issue=1, pageStart=81, pageEnd=90, url=null, language=null, rfNumber=[9], rfOrder=13, authorNames=WANG Mingnian, WANG Zhilong, ZHANG Xiao, journalName=Chinese Journal of Geotechnical Engineering, refType=null, unstructuredReference=(WANG MingnianWANG ZhilongZHANG Xiaoet al. Method for calculating deformation pressure of surrounding rock of deep-buried tunnels[J]. Chinese Journal of Geotechnical Engineering202042(1):81–90.(in Chinese)), articleTitle=Method for calculating deformation pressure of surrounding rock of deep-buried tunnels, refAbstract=null), Reference(id=1274369046514557623, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=8, issue=2, pageStart=156, pageEnd=165, url=null, language=null, rfNumber=[10], rfOrder=14, authorNames=马鑫民, 陈攀, 陈晨, journalName=矿业科学学报, refType=null, unstructuredReference=马鑫民,陈攀,陈晨,. 基于机器学习的煤巷围岩稳定性预测与应用[J]. 矿业科学学报20238(2):156–165., articleTitle=基于机器学习的煤巷围岩稳定性预测与应用, refAbstract=null), Reference(id=1274369046732661432, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=8, issue=2, pageStart=156, pageEnd=165, url=null, language=null, rfNumber=[10], rfOrder=15, authorNames=MA Xinmin, CHEN Pan, CHEN Chen, journalName=Journal of Mining Science and Technology, refType=null, unstructuredReference=(MA XinminCHEN PanCHEN Chenet al. Prediction of surrounding rock stability of coal roadway based on machine learning and its application[J]. Journal of Mining Science and Technology20238(2):156–165.(in Chinese)), articleTitle=Prediction of surrounding rock stability of coal roadway based on machine learning and its application, refAbstract=null), Reference(id=1274369046837519033, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2024, volume=166, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[11], rfOrder=16, authorNames=WANG L, PAN Q, WANG S, journalName=Computers and Geotechnics, refType=null, unstructuredReference=WANG LPAN QWANG S. Data-driven predictions of shield attitudes using Bayesian machine learning[J]. Computers and Geotechnics2024166:106002., articleTitle=Data-driven predictions of shield attitudes using Bayesian machine learning, refAbstract=null), Reference(id=1274369046900433594, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2015, volume=195, issue=null, pageStart=214, pageEnd=224, url=null, language=null, rfNumber=[12], rfOrder=17, authorNames=FENG X D, JIMENEZ R, journalName=Engineering Geology, refType=null, unstructuredReference=FENG X DJIMENEZ R. Predicting tunnel squeezing with incomplete data using Bayesian networks[J]. Engineering Geology2015195:214–224., articleTitle=Predicting tunnel squeezing with incomplete data using Bayesian networks, refAbstract=null), Reference(id=1274369046959153851, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2015, volume=29, issue=5, pageStart=1 447, pageEnd=1 461, url=null, language=null, rfNumber=[13], rfOrder=18, authorNames=WU X G, JIANG Z, ZHANG L M, journalName=Stochastic Environmental Research and Risk Assessment, refType=null, unstructuredReference=WU X GJIANG ZZHANG L Met al. Dynamic risk analysis for adjacent buildings in tunneling environments:a Bayesian network based approach[J]. Stochastic Environmental Research and Risk Assessment201529(5):1 447–1 461., articleTitle=Dynamic risk analysis for adjacent buildings in tunneling environments:a Bayesian network based approach, refAbstract=null), Reference(id=1274369047026262716, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2022, volume=17, issue=4, pageStart=1 343, pageEnd=1 366, url=null, language=null, rfNumber=[14], rfOrder=19, authorNames=ZHOU J, ZHU S L, QIU Y G, journalName=Acta Geotechnica, refType=null, unstructuredReference=ZHOU JZHU S LQIU Y Get al. Predicting tunnel squeezing using support vector machine optimized by whale optimization algorithm[J]. Acta Geotechnica202217(4):1 343–1 366., articleTitle=Predicting tunnel squeezing using support vector machine optimized by whale optimization algorithm, refAbstract=null), Reference(id=1274369047135314621, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=36, issue=11, pageStart=2 763, pageEnd=2 770, url=null, language=null, rfNumber=[15], rfOrder=20, authorNames=梁文添, 陈劲慧, journalName=岩石力学与工程学报, refType=null, unstructuredReference=梁文添,陈劲慧. 新奥法在三连拱特大断面隧道施工中的应用[J]. 岩石力学与工程学报201736(11):2 763–2 770., articleTitle=新奥法在三连拱特大断面隧道施工中的应用, refAbstract=null), Reference(id=1274369047227589310, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=36, issue=11, pageStart=2 763, pageEnd=2 770, url=null, language=null, rfNumber=[15], rfOrder=21, authorNames=LIANG Wentian, CHEN Jinhui, journalName=Chinese Journal of Rock Mechanics and Engineering, refType=null, unstructuredReference=(LIANG WentianCHEN Jinhui. Application of new Austrian tunnelling method to a large span trinocular cross section tunnel[J]. Chinese Journal of Rock Mechanics and Engineering201736(11):2 763–2 770.(in Chinese)), articleTitle=Application of new Austrian tunnelling method to a large span trinocular cross section tunnel, refAbstract=null), Reference(id=1274369047303086783, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2022, volume=7, issue=4, pageStart=514, pageEnd=527, url=null, language=null, rfNumber=[16], rfOrder=22, authorNames=FENG T G, WANG C R, ZHANG J, journalName=Underground Space, refType=null, unstructuredReference=FENG T GWANG C RZHANG Jet al. An improved artificial bee colony-random forest(IABC-RF) model for predicting the tunnel deformation due to an adjacent foundation pit excavation[J]. Underground Space20227(4):514–527., articleTitle=An improved artificial bee colony-random forest(IABC-RF) model for predicting the tunnel deformation due to an adjacent foundation pit excavation, refAbstract=null), Reference(id=1274369047378584256, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2015, volume=52, issue=3, pageStart=75, pageEnd=81, url=null, language=null, rfNumber=[17], rfOrder=23, authorNames=CHEN S, ZHU Y, HUANG L, journalName=Modern Tunnelling Technology, refType=null, unstructuredReference=CHEN SZHU YHUANG L. Application of chaotic time series analysis to the prediction of tunnel surrounding rock displacement[J]. Modern Tunnelling Technology201552(3):75–81., articleTitle=Application of chaotic time series analysis to the prediction of tunnel surrounding rock displacement, refAbstract=null), Reference(id=1274369047454081729, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2024, volume=294, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[18], rfOrder=24, authorNames=LI Z H, MA E L, LAI J X, journalName=Computers and Structures, refType=null, unstructuredReference=LI Z HMA E LLAI J Xet al. Tunnel deformation prediction during construction:An explainable hybrid model considering temporal and static factors[J]. Computers and Structures2024294:107276., articleTitle=Tunnel deformation prediction during construction:An explainable hybrid model considering temporal and static factors, refAbstract=null), Reference(id=1274369047592493762, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=116, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=25, authorNames=BAI C H, XUE Y G, QIU D H, journalName=Tunnelling and Underground Space Technology, refType=null, unstructuredReference=BAI C HXUE Y GQIU D Het al. Analysis of factors affecting the deformation of soft rock tunnels by data envelopment analysis and a risk assessment model[J]. Tunnelling and Underground Space Technology2021116:104111., articleTitle=Analysis of factors affecting the deformation of soft rock tunnels by data envelopment analysis and a risk assessment model, refAbstract=null), Reference(id=1274369047672185539, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=9, pageStart=1 459, pageEnd=1 465, url=null, language=null, rfNumber=[20], rfOrder=26, authorNames=张振, 苏欣荣, 袁新, journalName=动力工程学报, refType=null, unstructuredReference=张振,苏欣荣,袁新. 基于物理信息神经网络的气膜冷却湍流模型反演学习[J]. 动力工程学报202444(9):1 459–1 465., articleTitle=基于物理信息神经网络的气膜冷却湍流模型反演学习, refAbstract=null), Reference(id=1274369047743488708, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=9, pageStart=1 459, pageEnd=1 465, url=null, language=null, rfNumber=[20], rfOrder=27, authorNames=ZHANG Zhen, SU Xinrong, YUAN Xin, journalName=Journal of Chinese Society of Power Engineering, refType=null, unstructuredReference=(ZHANG ZhenSU XinrongYUAN Xin. PINN-based inversion learning of turbulence model for film cooling[J]. Journal of Chinese Society of Power Engineering202444(9):1 459–1 465.(in Chinese)), articleTitle=PINN-based inversion learning of turbulence model for film cooling, refAbstract=null), Reference(id=1274369047823180485, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2022, volume=44, issue=3, pageStart=492, pageEnd=501, url=null, language=null, rfNumber=[21], rfOrder=28, authorNames=卢鑫月, 许成顺, 侯本伟, journalName=岩土工程学报, refType=null, unstructuredReference=卢鑫月,许成顺,侯本伟,. 基于动态贝叶斯网络的地铁隧道施工风险评估[J]. 岩土工程学报202244(3):492–501., articleTitle=基于动态贝叶斯网络的地铁隧道施工风险评估, refAbstract=null), Reference(id=1274369047957398214, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2022, volume=44, issue=3, pageStart=492, pageEnd=501, url=null, language=null, rfNumber=[21], rfOrder=29, authorNames=LU Xinyue, XU Chengshun, HOU Benwei, journalName=Chinese Journal of Geotechnical Engineering, refType=null, unstructuredReference=(LU XinyueXU ChengshunHOU Benweiet al. Risk assessment of metro construction based on dynamic Bayesian network[J]. Chinese Journal of Geotechnical Engineering202244(3):492–501.(in Chinese)), articleTitle=Risk assessment of metro construction based on dynamic Bayesian network, refAbstract=null), Reference(id=1274369048083227335, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=13, pageEnd=27, url=null, language=null, rfNumber=[22], rfOrder=30, authorNames=CHING W K, 程伟琪, 黄曦敏, journalName=马尔可夫链:模型、算法与应用, refType=null, unstructuredReference=CHING W K,程伟琪,黄曦敏,. 马尔可夫链:模型、算法与应用[M]. 北京:清华大学出版社,2015:13–27., articleTitle=null, refAbstract=null), Reference(id=1274369048158724808, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=13, pageEnd=27, url=null, language=null, rfNumber=[22], rfOrder=31, authorNames=CHING W K, CHENG Weiqi, HUANG Ximin, journalName=Markov Chains:models,algorithms and applications second edition, refType=null, unstructuredReference=(CHING W KCHENG WeiqiHUANG Ximinet al. Markov Chains:models,algorithms and applications second edition[M]. Beijing:Tsinghua University Press,2015:13–27.(in Chinese)), articleTitle=null, refAbstract=null), Reference(id=1274369048376828617, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2008, volume=172, issue=1, pageStart=42, pageEnd=68, url=null, language=null, rfNumber=[23], rfOrder=32, authorNames=SMITH J Q, ANDERSON P E, journalName=Artificial Intelligence, refType=null, unstructuredReference=SMITH J QANDERSON P E. Conditional independence and chain event graphs[J]. Artificial Intelligence2008172(1):42–68., articleTitle=Conditional independence and chain event graphs, refAbstract=null), Reference(id=1274369048502657738, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=33, issue=1, pageStart=5, pageEnd=9, url=null, language=null, rfNumber=[24], rfOrder=33, authorNames=杨悦, 陈孝国, 高霞, journalName=齐齐哈尔大学学报:自然科学版, refType=null, unstructuredReference=杨悦,陈孝国,高霞,. 基于神经网络–蒙特卡罗法的隧道初衬可靠度研究[J]. 齐齐哈尔大学学报:自然科学版201733(1):5–9., articleTitle=基于神经网络–蒙特卡罗法的隧道初衬可靠度研究, refAbstract=null), Reference(id=1274369048720761547, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=33, issue=1, pageStart=5, pageEnd=9, url=null, language=null, rfNumber=[24], rfOrder=34, authorNames=YANG Yue, CHEN Xiaoguo, GAO Xia, journalName=Journal of Qiqihar University:Natural Science, refType=null, unstructuredReference=(YANG YueCHEN XiaoguoGAO Xiaet al. Study on reliability of tunneling initial lining based on radial basis function neural network-Monte Carlo method[J]. Journal of Qiqihar University:Natural Science201733(1):5–9.(in Chinese)), articleTitle=Study on reliability of tunneling initial lining based on radial basis function neural network-Monte Carlo method, refAbstract=null), Reference(id=1274369048896922316, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=53, issue=1, pageStart=1, pageEnd=12, url=null, language=null, rfNumber=[25], rfOrder=35, authorNames=张广泽, 邓建辉, 王栋, journalName=工程科学与技术, refType=null, unstructuredReference=张广泽,邓建辉,王栋,. 隧道围岩构造软岩大变形发生机理及分级方法[J]. 工程科学与技术202153(1):1–12., articleTitle=隧道围岩构造软岩大变形发生机理及分级方法, refAbstract=null), Reference(id=1274369049014362829, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=53, issue=1, pageStart=1, pageEnd=12, url=null, language=null, rfNumber=[25], rfOrder=36, authorNames=ZHANG Guangze, DENG Jianhui, WANG Dong, journalName=Engineering Science and Technology, refType=null, unstructuredReference=(ZHANG GuangzeDENG JianhuiWANG Donget al. Mechanism and classification of tectonic-induced large deformation of soft rock tunnels[J]. Engineering Science and Technology202153(1):1–12.(in Chinese)), articleTitle=Mechanism and classification of tectonic-induced large deformation of soft rock tunnels, refAbstract=null), Reference(id=1274369049307964110, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[26], rfOrder=37, authorNames=中华人民共和国国家标准编写组, journalName=null, refType=null, unstructuredReference=中华人民共和国国家标准编写组. GB/T 10003—2016铁路隧道设计规范[S]. 北京:中国铁道出版社,2016., articleTitle=null, refAbstract=null), Reference(id=1274369049433793231, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[26], rfOrder=38, authorNames=The National Standards Compilation Group of People's Republic of China, journalName=null, refType=null, unstructuredReference=(The National Standards Compilation Group of People's Republic of China. GB/T 10003—2016 Code for design of railway tunnel[S]. Beijing:China Railway Publishing House,2016.(in Chinese)), articleTitle=null, refAbstract=null), Reference(id=1274369049618342608, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[27], rfOrder=39, authorNames=中华人民共和国行业标准编写组, journalName=null, refType=null, unstructuredReference=中华人民共和国行业标准编写组. JTG 3370.1—2018公路隧道设计规范[S]. 北京:人民交通出版社,2018., articleTitle=null, refAbstract=null), Reference(id=1274369049723200209, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[27], rfOrder=40, authorNames=The Professional Standards Compilation Group of People's Republic of China, journalName=null, refType=null, unstructuredReference=(The Professional Standards Compilation Group of People's Republic of China. JTG 3370.1—2018 Specification for design of highway tunnel[S]. Beijing:China Communications Press,2018.(in Chinese)), articleTitle=null, refAbstract=null), Reference(id=1274369049790309074, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=1992, volume=9, issue=4, pageStart=309, pageEnd=347, url=null, language=null, rfNumber=[28], rfOrder=41, authorNames=COOPER G, HERSKOVITS E, journalName=Machine Learning, refType=null, unstructuredReference=COOPER GHERSKOVITS E. A Bayesian method for the induction of probabilistic networks from data[J]. Machine Learning19929(4):309–347., articleTitle=A Bayesian method for the induction of probabilistic networks from data, refAbstract=null), Reference(id=1274369049861612243, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=21, pageEnd=30, url=null, language=null, rfNumber=[29], rfOrder=42, authorNames=朱永全, 宋玉香, journalName=隧道工程, refType=null, unstructuredReference=朱永全,宋玉香. 隧道工程[M]. 北京:中国铁道出版社,2018:21–30., articleTitle=null, refAbstract=null), Reference(id=1274369049937109716, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=21, pageEnd=30, url=null, language=null, rfNumber=[29], rfOrder=43, authorNames=ZHU Yongquan, SONG Yuxiang, journalName=Tunnel engineering, refType=null, unstructuredReference=(ZHU YongquanSONG Yuxiang. Tunnel engineering[M]. Beijing:China Railway Publishing House,2018:21–30.(in Chinese)), articleTitle=null, refAbstract=null), Reference(id=1274369050029384405, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=39, issue=6, pageStart=44, pageEnd=51, url=null, language=null, rfNumber=[30], rfOrder=44, authorNames=孙景来, 刘保国, 储昭飞, journalName=中国铁道科学, refType=null, unstructuredReference=孙景来,刘保国,储昭飞,. 隧道坍塌事故类型划分及其主要特征[J]. 中国铁道科学201839(6):44–51., articleTitle=隧道坍塌事故类型划分及其主要特征, refAbstract=null), Reference(id=1274369050176185046, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=39, issue=6, pageStart=44, pageEnd=51, url=null, language=null, rfNumber=[30], rfOrder=45, authorNames=SUN Jinglai, LIU Baoguo, CHU Zhaofei, journalName=China Railway Science, refType=null, unstructuredReference=(SUN JinglaiLIU BaoguoCHU Zhaofeiet al. Type classification and main characteristics of tunnel collapses[J]. China Railway Science201839(6):44–51.(in Chinese)), articleTitle=Type classification and main characteristics of tunnel collapses, refAbstract=null), Reference(id=1274369050260071127, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=53, issue=6, pageStart=1 237, pageEnd=1 244, url=null, language=null, rfNumber=[31], rfOrder=46, authorNames=陈子全, 何川, 吴迪, journalName=西南交通大学学报, refType=null, unstructuredReference=陈子全,何川,吴迪,. 高地应力层状软岩隧道大变形预测分级研究[J]. 西南交通大学学报201853(6):1 237–1 244., articleTitle=高地应力层状软岩隧道大变形预测分级研究, refAbstract=null), Reference(id=1274369050318791384, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=53, issue=6, pageStart=1 237, pageEnd=1 244, url=null, language=null, rfNumber=[31], rfOrder=47, authorNames=CHEN Ziquan, HE Chuan, WU Di, journalName=Journal of Southwest Jiaotong University, refType=null, unstructuredReference=(CHEN ZiquanHE ChuanWU Diet al. Study of large deformation classification criterion for layered soft rock tunnels under high geostress[J]. Journal of Southwest Jiaotong University201853(6):1 237–1 244.(in Chinese)), articleTitle=Study of large deformation classification criterion for layered soft rock tunnels under high geostress, refAbstract=null), Reference(id=1274369050385900249, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=1988, volume=13, issue=3, pageStart=13, pageEnd=19, url=null, language=null, rfNumber=[32], rfOrder=48, authorNames=任德惠, 张平, journalName=煤炭学报, refType=null, unstructuredReference=任德惠,张平. 不同倾角结构面对巷道稳定性的影响[J]. 煤炭学报198813(3):13–19., articleTitle=不同倾角结构面对巷道稳定性的影响, refAbstract=null), Reference(id=1274369050469786330, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=1988, volume=13, issue=3, pageStart=13, pageEnd=19, url=null, language=null, rfNumber=[32], rfOrder=49, authorNames=REN Dehui, ZHANG Ping, journalName=Journal of China Coal Society, refType=null, unstructuredReference=(REN DehuiZHANG Ping. Influence of structural planes with different dip angles on roadway stability[J]. Journal of China Coal Society198813(3):13–19.(in Chinese)), articleTitle=Influence of structural planes with different dip angles on roadway stability, refAbstract=null), Reference(id=1274369050574643931, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[33], rfOrder=50, authorNames=吴秋军, journalName=null, refType=null, unstructuredReference=吴秋军. 铁路隧道围岩变形机理与分级方法研究[博士学位论文][D]. 成都:西南交通大学,2017., articleTitle=铁路隧道围岩变形机理与分级方法研究, refAbstract=null), Reference(id=1274369050645947100, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[33], rfOrder=51, authorNames=WU Qiujun, journalName=null, refType=null, unstructuredReference=(WU Qiujun. Deformation mechanism and classification methods for surrounding rock of railway tunnels[Ph. D. Thesis][D]. Chengdu:Southwest Jiaotong University,2017.(in Chinese)), articleTitle=Deformation mechanism and classification methods for surrounding rock of railway tunnels, refAbstract=null), Reference(id=1274369050708861661, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=13, pageStart=234, pageEnd=235, url=null, language=null, rfNumber=[34], rfOrder=52, authorNames=王冠勇, journalName=交通世界(建养・机械), refType=null, unstructuredReference=王冠勇. 分析影响隧道围岩稳定性因素[J]. 交通世界(建养・机械)2012,(13):234–235., articleTitle=分析影响隧道围岩稳定性因素, refAbstract=null), Reference(id=1274369050775970526, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=13, pageStart=234, pageEnd=235, url=null, language=null, rfNumber=[34], rfOrder=53, authorNames=WANG Guanyong, journalName=Traffic World (Construction and Maintenance・Machinery), refType=null, unstructuredReference=(WANG Guanyong. Analysis of factors affecting the stability of tunnel surrounding rock[J]. Traffic World (Construction and Maintenance・Machinery)2012,(13):234–235.(in Chinese)), articleTitle=Analysis of factors affecting the stability of tunnel surrounding rock, refAbstract=null), Reference(id=1274369050893411039, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2005, volume=24, issue=10, pageStart=1 697, pageEnd=1 702, url=null, language=null, rfNumber=[35], rfOrder=54, authorNames=吉小明, journalName=岩石力学与工程学报, refType=null, unstructuredReference=吉小明. 隧道开挖的围岩损伤扰动带分析[J]. 岩石力学与工程学报200524(10):1 697–1 702., articleTitle=隧道开挖的围岩损伤扰动带分析, refAbstract=null), Reference(id=1274369050989880032, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2005, volume=24, issue=10, pageStart=1 697, pageEnd=1 702, url=null, language=null, rfNumber=[35], rfOrder=55, authorNames=JI Xiaoming, journalName=Chinese Journal of Rock Mechanics and Engineering, refType=null, unstructuredReference=(JI Xiaoming. Study on mechanical and hydraulic behavior of tunnel surrounding rock masses in excavation-disturbed zone[J]. Chinese Journal of Rock Mechanics and Engineering200524(10):1 697–1 702.(in Chinese)), articleTitle=Study on mechanical and hydraulic behavior of tunnel surrounding rock masses in excavation-disturbed zone, refAbstract=null), Reference(id=1274369051048600289, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2009, volume=31, issue=3, pageStart=390, pageEnd=395, url=null, language=null, rfNumber=[36], rfOrder=56, authorNames=来弘鹏, 林永贵, 谢永利, journalName=岩土工程学报, refType=null, unstructuredReference=来弘鹏,林永贵,谢永利,. 支护时机对软弱围岩公路隧道力学特征影响的试验研究[J]. 岩土工程学报200931(3):390–395., articleTitle=支护时机对软弱围岩公路隧道力学特征影响的试验研究, refAbstract=null), Reference(id=1274369051098931938, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2009, volume=31, issue=3, pageStart=390, pageEnd=395, url=null, language=null, rfNumber=[36], rfOrder=57, authorNames=LAI Hongpeng, LIN Yonggui, XIE Yongli, journalName=Chinese Journal of Geotechnical Engineering, refType=null, unstructuredReference=(LAI HongpengLIN YongguiXIE Yongliet al. Influence of supporting opportunity on stress characteristics of soft-weak surrounding rocks in highway tunnels[J]. Chinese Journal of Geotechnical Engineering200931(3):390–395.(in Chinese)), articleTitle=Influence of supporting opportunity on stress characteristics of soft-weak surrounding rocks in highway tunnels, refAbstract=null), Reference(id=1274369051157652195, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2013, volume=28, issue=1, pageStart=1, pageEnd=21, url=null, language=null, rfNumber=[37], rfOrder=58, authorNames=ŠPAČKOVÁ O, STRAUB D, journalName=Computer-Aided Civil and Infrastructure Engineering, refType=null, unstructuredReference=ŠPAČKOVÁ OSTRAUB D. Dynamic Bayesian network for probabilistic modeling of tunnel excavation processes[J]. Computer-Aided Civil and Infrastructure Engineering201328(1):1–21., articleTitle=Dynamic Bayesian network for probabilistic modeling of tunnel excavation processes, refAbstract=null), Reference(id=1274369051237343972, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=5, pageStart=164, pageEnd=169, url=null, language=null, rfNumber=[38], rfOrder=59, authorNames=李海斌, 翟秋柱, 张优, journalName=路基工程, refType=null, unstructuredReference=李海斌,翟秋柱,张优,. PSO-BP神经网络在隧道围岩变形预测中的应用[J]. 路基工程2017,(5):164–169., articleTitle=PSO-BP神经网络在隧道围岩变形预测中的应用, refAbstract=null), Reference(id=1274369051367367397, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=5, pageStart=164, pageEnd=169, url=null, language=null, rfNumber=[38], rfOrder=60, authorNames=LI Haibin, ZHAI Qiuzhu, ZHANG You, journalName=Subgrade Engineering, refType=null, unstructuredReference=(LI HaibinZHAI QiuzhuZHANG Youet al. Application of PSO-BP neural network in prediction of tunnel rock deformation[J]. Subgrade Engineering2017,(5):164–169.(in Chinese)), articleTitle=Application of PSO-BP neural network in prediction of tunnel rock deformation, refAbstract=null), Reference(id=1274369051447059174, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2008, volume=29, issue=2, pageStart=82, pageEnd=87, url=null, language=null, rfNumber=[39], rfOrder=61, authorNames=王树栋, 刘开云, journalName=中国铁道科学, refType=null, unstructuredReference=王树栋,刘开云. 长大隧道软弱围岩施工大变形智能预测方法[J]. 中国铁道科学200829(2):82–87., articleTitle=长大隧道软弱围岩施工大变形智能预测方法, refAbstract=null), Reference(id=1274369051509973735, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2008, volume=29, issue=2, pageStart=82, pageEnd=87, url=null, language=null, rfNumber=[39], rfOrder=62, authorNames=WANG Shudong, LIU Kaiyun, journalName=China Railway Science, refType=null, unstructuredReference=(WANG ShudongLIU Kaiyun. Intelligence method for large deformation prediction of weak surrounding rock construction in long and large tunnel[J]. China Railway Science200829(2):82–87.(in Chinese)), articleTitle=Intelligence method for large deformation prediction of weak surrounding rock construction in long and large tunnel, refAbstract=null), Reference(id=1274369051564499688, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[40], rfOrder=63, authorNames=解志刚, journalName=null, refType=null, unstructuredReference=解志刚. 隧道围岩变形预测及其在工程中的应用[硕士学位论文][D]. 兰州:兰州交通大学,2012., articleTitle=隧道围岩变形预测及其在工程中的应用, refAbstract=null), Reference(id=1274369051681940201, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2012, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[40], rfOrder=64, authorNames=XIE Zhigang, journalName=null, refType=null, unstructuredReference=(XIE Zhigang. Tunnel surrounding rock deformation and prediction and its application in engineering[M. S. Thesis][D]. Lanzhou:Lanzhou Jiaotong University,2012.(in Chinese)), articleTitle=Tunnel surrounding rock deformation and prediction and its application in engineering, refAbstract=null), Reference(id=1274369051824546538, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=10, issue=null, pageStart=1, pageEnd=15, url=null, language=null, rfNumber=[41], rfOrder=65, authorNames=NIU Z L, WANG Y Q, FAN S Y, journalName=Frontiers in Earth Science, refType=null, unstructuredReference=NIU Z LWANG Y QFAN S Y. Research on the deformation control of surrounding rock about large-section tunnel in strong-medium weathered slate[J]. Frontiers in Earth Science202310:1–15., articleTitle=Research on the deformation control of surrounding rock about large-section tunnel in strong-medium weathered slate, refAbstract=null), Reference(id=1274369051891655403, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=41, issue=6, pageStart=126, pageEnd=128, url=null, language=null, rfNumber=[42], rfOrder=66, authorNames=汪学清, 刘爽, 李秋燕, journalName=矿冶工程, refType=null, unstructuredReference=汪学清,刘爽,李秋燕,. 基于K折交叉验证的SVM隧道围岩分级判别[J]. 矿冶工程202141(6):126–128., articleTitle=基于K折交叉验证的SVM隧道围岩分级判别, refAbstract=null), Reference(id=1274369051979735788, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=41, issue=6, pageStart=126, pageEnd=128, url=null, language=null, rfNumber=[42], rfOrder=67, authorNames=WANG Xueqing, LIU Shuang, LI Qiuyan, journalName=Mining and Metallurgical Engineering, refType=null, unstructuredReference=(WANG XueqingLIU ShuangLI Qiuyanet al. Classification and discrimination of surrounding rock of tunnel based on SVM of K-Fold cross validation[J]. Mining and Metallurgical Engineering202141(6):126–128.(in Chinese)), articleTitle=Classification and discrimination of surrounding rock of tunnel based on SVM of K-Fold cross validation, refAbstract=null), Reference(id=1274369052055233261, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=55, issue=3, pageStart=458, pageEnd=466, url=null, language=null, rfNumber=[43], rfOrder=68, authorNames=毛奕喆, 龚国芳, 周星海, journalName=浙江大学学报:工学版, refType=null, unstructuredReference=毛奕喆,龚国芳,周星海,. 基于马尔可夫过程和深度神经网络的TBM围岩识别[J]. 浙江大学学报:工学版202155(3):458–466., articleTitle=基于马尔可夫过程和深度神经网络的TBM围岩识别, refAbstract=null), Reference(id=1274369052210422510, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2021, volume=55, issue=3, pageStart=458, pageEnd=466, url=null, language=null, rfNumber=[43], rfOrder=69, authorNames=MAO Yizhe, GONG Guofang, ZHOU Xinghai, journalName=Journal of Zhejiang University:Engineering Science, refType=null, unstructuredReference=(MAO YizheGONG GuofangZHOU Xinghaiet al. Identification of TBM surrounding rock based on Markov process and deep neural network[J]. Journal of Zhejiang University:Engineering Science202155(3):458–466.(in Chinese)), articleTitle=Identification of TBM surrounding rock based on Markov process and deep neural network, refAbstract=null), Reference(id=1274369052298502895, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2001, volume=15, issue=1, pageStart=9, pageEnd=42, url=null, language=null, rfNumber=[44], rfOrder=70, authorNames=GHAHRAMANI Z, journalName=International Journal of Pattern Recognition and Artificial Intelligence, refType=null, unstructuredReference=GHAHRAMANI Z. An introduction to hidden Markov models and Bayesian Networks[J]. International Journal of Pattern Recognition and Artificial Intelligence200115(1):9–42., articleTitle=An introduction to hidden Markov models and Bayesian Networks, refAbstract=null), Reference(id=1274369052365611760, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2004, volume=null, issue=null, pageStart=1, pageEnd=2, url=null, language=null, rfNumber=[45], rfOrder=71, authorNames=刘次华, journalName=随机过程及其应用, refType=null, unstructuredReference=刘次华. 随机过程及其应用[M]. 北京:清华大学出版社,2004:1–2., articleTitle=null, refAbstract=null), Reference(id=1274369052575326961, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2004, volume=null, issue=null, pageStart=1, pageEnd=2, url=null, language=null, rfNumber=[45], rfOrder=72, authorNames=LIU Cihua, journalName=Stochastic processes and their applications, refType=null, unstructuredReference=(LIU Cihua. Stochastic processes and their applications[M]. Beijing:Tsinghua University Press,2004:1–2.(in Chinese)), articleTitle=null, refAbstract=null), Reference(id=1274369052684378866, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2023, volume=41, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[46], rfOrder=73, authorNames=YE X W, ZHANG X L, ZHANG H Q, journalName=Transportation Geotechnics, refType=null, unstructuredReference=YE X WZHANG X LZHANG H Qet al. Prediction of lining upward movement during shield tunneling using machine learning algorithms and field monitoring data[J]. Transportation Geotechnics202341:101002., articleTitle=Prediction of lining upward movement during shield tunneling using machine learning algorithms and field monitoring data, refAbstract=null), Reference(id=1274369052768264947, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=37, issue=10, pageStart=2 010, pageEnd=2 018, url=null, language=null, rfNumber=[47], rfOrder=74, authorNames=杨背背, 殷坤龙, 杜娟, journalName=岩石力学与工程学报, refType=null, unstructuredReference=杨背背,殷坤龙,杜娟. 基于时间序列与长短时记忆网络的滑坡位移动态预测模型[J]. 岩石力学与工程学报201837(10):2 010–2 018., articleTitle=基于时间序列与长短时记忆网络的滑坡位移动态预测模型, refAbstract=null), Reference(id=1274369052835373812, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, doi=null, pmid=null, pmcid=null, year=2018, volume=37, issue=10, pageStart=2 010, pageEnd=2 018, url=null, language=null, rfNumber=[47], rfOrder=75, authorNames=YANG Beibei, YIN Kunlong, DU Juan, journalName=Chinese Journal of Rock Mechanics and Engineering, refType=null, unstructuredReference=(YANG BeibeiYIN KunlongDU Juan. A model for predicting landslide displacement based on time series and long and short term memory neural network[J]. Chinese Journal of Rock Mechanics and Engineering201837(10):2 010–2 018.(in Chinese)), articleTitle=A model for predicting landslide displacement based on time series and long and short term memory neural network, refAbstract=null)], funds=[Fund(id=1274369043398189731, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, awardId=KJQN202301530, language=EN, fundingSource=Science and Technology Research Program of Chongqing Municipal Education Commission(KJQN202301530), fundOrder=null, country=null), Fund(id=1274369043465298596, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, awardId=KJQN202301530, language=CN, fundingSource=重庆市教育委员会科学技术研究项目(KJQN202301530), fundOrder=null, country=null), Fund(id=1274369043528213157, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, awardId=52304125, language=EN, fundingSource=National Natural Science Foundation for Young Scholars of China(52304125), fundOrder=null, country=null), Fund(id=1274369043763094182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, awardId=52304125, language=CN, fundingSource=国家自然科学基金青年基金资助项目(52304125), fundOrder=null, country=null), Fund(id=1274369043867951783, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, awardId=CSTB2023NSCQ-MSX0828, language=EN, fundingSource=Chongqing Natural Science Foundation Project(CSTB2023NSCQ-MSX0828), fundOrder=null, country=null), Fund(id=1274369044216079016, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, awardId=CSTB2023NSCQ-MSX0828, language=CN, fundingSource=重庆市自然科学基金面上项目(CSTB2023NSCQ-MSX0828), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1274369016286208573, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=1., ext=[AuthorCompanyExt(id=1274369016294597182, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China), AuthorCompanyExt(id=1274369016302985791, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016286208573, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.重庆科技大学 安全科学与工程学院,重庆 400041)]), AuthorCompany(id=1274369016357511744, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=2., ext=[AuthorCompanyExt(id=1274369016378483265, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016357511744, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.State Key Laboratory of Geomechanics and Geotechnical Engineering Safety, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan, Hubei 430071, China), AuthorCompanyExt(id=1274369016391066178, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016357511744, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中国科学院武汉岩土力学研究所 岩土力学与工程安全全国重点实验室,湖北 武汉 430071)]), AuthorCompany(id=1274369016592392772, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, xref=3., ext=[AuthorCompanyExt(id=1274369016609169989, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016592392772, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.School of Resource and Safety Engineering, Chongqing University, Chongqing 400044, China), AuthorCompanyExt(id=1274369016621752902, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, companyId=1274369016592392772, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.重庆大学 资源与安全学院,重庆 400044)])], figs=[ArticleFig(id=1274369035399651959, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=HmY5kP12qH2UlNu4/bpD0w==, figureFileBig=UQd/5t5wVPiZkuhg244e6A==, tableContent=null), ArticleFig(id=1274369035470955128, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.1, caption=BN model diagram of surrounding rock deformation dynamic prediction, figureFileSmall=HmY5kP12qH2UlNu4/bpD0w==, figureFileBig=UQd/5t5wVPiZkuhg244e6A==, tableContent=null), ArticleFig(id=1274369035827470969, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=E0mByuXhLA9g/Kc5TE/iTA==, figureFileBig=tlMqgoIHvP5A4ZdpKctuKg==, tableContent=null), ArticleFig(id=1274369036142043770, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.2, caption=Typical displacement time-series curves in the case database, figureFileSmall=E0mByuXhLA9g/Kc5TE/iTA==, figureFileBig=tlMqgoIHvP5A4ZdpKctuKg==, tableContent=null), ArticleFig(id=1274369036217541243, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=Pey84FdOX5WXQWyvVOuBdA==, figureFileBig=AILBWh0tGsa9jJmuqPRI7A==, tableContent=null), ArticleFig(id=1274369036561474172, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.3, caption=BN structure construction flow chart by K2-score algorithm(mountain climbing method), figureFileSmall=Pey84FdOX5WXQWyvVOuBdA==, figureFileBig=AILBWh0tGsa9jJmuqPRI7A==, tableContent=null), ArticleFig(id=1274369036624388733, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=8h8v3xFmWPqJCOaN+vzbsg==, figureFileBig=86TiRV/V+Z/cb70CkMQy6Q==, tableContent=null), ArticleFig(id=1274369036687303294, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.4, caption=SBN model of surrounding rock deformation prediction, figureFileSmall=8h8v3xFmWPqJCOaN+vzbsg==, figureFileBig=86TiRV/V+Z/cb70CkMQy6Q==, tableContent=null), ArticleFig(id=1274369036771189375, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=ojh4ehB9z+h60wMzzicQLg==, figureFileBig=XID8ZhZEpSiXUADYqDLGGA==, tableContent=null), ArticleFig(id=1274369036834103936, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.5, caption=Schematic diagram of the construction methods of DBN model, figureFileSmall=ojh4ehB9z+h60wMzzicQLg==, figureFileBig=XID8ZhZEpSiXUADYqDLGGA==, tableContent=null), ArticleFig(id=1274369037027041921, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=waub1fkLn84k1GilW5Of+g==, figureFileBig=hZXxDnwMgMkwv4TJSrT1/Q==, tableContent=null), ArticleFig(id=1274369037400334978, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.6, caption=Vault settlement M-DBN model constructed based on time-series curve reconstructed by TNP 3 (Model 10), figureFileSmall=waub1fkLn84k1GilW5Of+g==, figureFileBig=hZXxDnwMgMkwv4TJSrT1/Q==, tableContent=null), ArticleFig(id=1274369037500998275, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=Dqaal80IxYnrr6QMN7aCiQ==, figureFileBig=7uBWwRO+aIZlhfdXr6VX2w==, tableContent=null), ArticleFig(id=1274369037844931204, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.7, caption=Vault settlement OM-DBN model constructed based on time-series curve reconstructed by TNP 4 (Model 16), figureFileSmall=Dqaal80IxYnrr6QMN7aCiQ==, figureFileBig=7uBWwRO+aIZlhfdXr6VX2w==, tableContent=null), ArticleFig(id=1274369037924622981, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=XSpYCE364y1ujBGWoYzvxA==, figureFileBig=RZ4C4shlq+/ZN1Sx8qRjSA==, tableContent=null), ArticleFig(id=1274369038029480582, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.8, caption=Forward transfer vault subsidence OM-DBN model based on time-series curve reconstructed by TNP 4(Model 15), figureFileSmall=XSpYCE364y1ujBGWoYzvxA==, figureFileBig=RZ4C4shlq+/ZN1Sx8qRjSA==, tableContent=null), ArticleFig(id=1274369038247584391, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=14GSrTj9S7e0/c4c9Ucpwg==, figureFileBig=7yybifitHHc35Dq6deSumQ==, tableContent=null), ArticleFig(id=1274369038432133768, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.9, caption=The change law of P and R values with observation time in the DBN model, figureFileSmall=14GSrTj9S7e0/c4c9Ucpwg==, figureFileBig=7yybifitHHc35Dq6deSumQ==, tableContent=null), ArticleFig(id=1274369038708957833, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=8U3U3Rbn7TlKsM+k+/SycQ==, figureFileBig=qVcYBEq/yb83CvEiUyzkhg==, tableContent=null), ArticleFig(id=1274369039166136970, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.10, caption=The change law of F1 values and ACC values with observation time in the DBN model, figureFileSmall=8U3U3Rbn7TlKsM+k+/SycQ==, figureFileBig=qVcYBEq/yb83CvEiUyzkhg==, tableContent=null), ArticleFig(id=1274369039585567371, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=9I3NPSITr8LEzm1EdWYxRA==, figureFileBig=SwgzkDOkCxHtnP591Nl8Xg==, tableContent=null), ArticleFig(id=1274369039707202188, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.11, caption=Major structure design of the tunnel in case 1, figureFileSmall=9I3NPSITr8LEzm1EdWYxRA==, figureFileBig=SwgzkDOkCxHtnP591Nl8Xg==, tableContent=null), ArticleFig(id=1274369039984026253, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=ueistN171b3+UpPQzFeoQQ==, figureFileBig=bj11MbJmHqMg+U51A5/3Tg==, tableContent=null), ArticleFig(id=1274369040072106638, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.12, caption=Comparison between in-site monitored displacement values and predicted displacement values, figureFileSmall=ueistN171b3+UpPQzFeoQQ==, figureFileBig=bj11MbJmHqMg+U51A5/3Tg==, tableContent=null), ArticleFig(id=1274369040340542095, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=eevSX+YZphb6Ssb+Lu0/Vg==, figureFileBig=i0O0qawKRV4q9O2jRvZTJQ==, tableContent=null), ArticleFig(id=1274369040437011088, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig.13, caption=Major structure design of the tunnel in case 2, figureFileSmall=eevSX+YZphb6Ssb+Lu0/Vg==, figureFileBig=i0O0qawKRV4q9O2jRvZTJQ==, tableContent=null), ArticleFig(id=1274369040575423121, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=7F35kmHvHbPMkw+jW9PbDg==, figureFileBig=OtP3cEyJjvlOwl8Vt6rO6w==, tableContent=null), ArticleFig(id=1274369040822887058, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Fig 14, caption=Prediction and updating of vault subsidence based on DBN model, figureFileSmall=7F35kmHvHbPMkw+jW9PbDg==, figureFileBig=OtP3cEyJjvlOwl8Vt6rO6w==, tableContent=null), ArticleFig(id=1274369040957104787, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Tunnel nameGeological environment condition and tunnel structure characteristicConstruction condition
Surrounding rock classificationSurrounding rock lithologySurrounding rock structureInitial in-situ stress stateDip angles of dominant discontinuities/(°)Tunnel burial depth/mTunnel excavation spanGroundwater conditionGroundwater controlAdvanced support activation
Xianghe tunnel D1K153+340[38]IVGranite, Sandstone, SlateBrokenHigh in-situ stress*Maximum 70512~14Water-richReasonableAdvanced anchor bolt
Xianghe tunnel D1K150+455[38]IVLime, DolomiteBrokenHigh in-situ stress40Maximum 70512~14Water-richReasonableAdvanced anchor bolt
Baozhen tunnel DK73+300[38]IVSilty shaleCrackedExtremely high in-situ stress400+>14Water-richReasonableAdvanced anchor bolt
Huama tunnel DK302+177[39]IVLime, SlateBrokenGeneral in-situ stress80Maximum 1 300+14.600Relatively water-richReasonableAdvanced anchor bolt
Baitanwu tunnel ZK7+815[40]IVWeathered porphyritic rhyolite, Argillaceous siltstoneBrokenGeneral in-situ stress52~12112~14Abundant groundwater activityReasonableAdvanced small pipe grouting
Tunnel nameConstruction conditionsDeformation monitoring objectMonitoring duration/dDeformation/mm
Excavation methodSurrounding rock disturbance intensitySupport timeSupport strengthObserved 1 dObserved 2 dObserved 3 dObserved 4 d
Xianghe tunnel D1K153+340[38]Short step excavation methodModerate disturbanceReasonableModerate intensityVault settlement2616.24023.83032.76040.430
Xianghe tunnel D1K150+455[38]Short step excavation methodModerate disturbanceReasonableModerate intensityVault settlement2618.76029.30037.32044.580
Baozhen tunnel DK73+300[38]Three-step excavation methodModerate disturbanceReasonableModerate intensityHorizontal convergence1819.02035.89050.57063.300
Huama tunnel DK302+177[39]Three-step excavation methodModerate disturbanceUnreasonableModerate intensity **Horizontal convergence3221.00046.00058.00065.000
Baitanwu tunnel ZK7+815[40]Ultra-short step methodModerate disturbanceUnreasonableLow intensityVault settlement5630.70045.10053.30060.600
), ArticleFig(id=1274369041212957332, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 1, caption=

Case data for prediction model construction of tunnel surrounding rock deformation (partial cases)

, figureFileSmall=null, figureFileBig=null, tableContent=
Tunnel nameGeological environment condition and tunnel structure characteristicConstruction condition
Surrounding rock classificationSurrounding rock lithologySurrounding rock structureInitial in-situ stress stateDip angles of dominant discontinuities/(°)Tunnel burial depth/mTunnel excavation spanGroundwater conditionGroundwater controlAdvanced support activation
Xianghe tunnel D1K153+340[38]IVGranite, Sandstone, SlateBrokenHigh in-situ stress*Maximum 70512~14Water-richReasonableAdvanced anchor bolt
Xianghe tunnel D1K150+455[38]IVLime, DolomiteBrokenHigh in-situ stress40Maximum 70512~14Water-richReasonableAdvanced anchor bolt
Baozhen tunnel DK73+300[38]IVSilty shaleCrackedExtremely high in-situ stress400+>14Water-richReasonableAdvanced anchor bolt
Huama tunnel DK302+177[39]IVLime, SlateBrokenGeneral in-situ stress80Maximum 1 300+14.600Relatively water-richReasonableAdvanced anchor bolt
Baitanwu tunnel ZK7+815[40]IVWeathered porphyritic rhyolite, Argillaceous siltstoneBrokenGeneral in-situ stress52~12112~14Abundant groundwater activityReasonableAdvanced small pipe grouting
Tunnel nameConstruction conditionsDeformation monitoring objectMonitoring duration/dDeformation/mm
Excavation methodSurrounding rock disturbance intensitySupport timeSupport strengthObserved 1 dObserved 2 dObserved 3 dObserved 4 d
Xianghe tunnel D1K153+340[38]Short step excavation methodModerate disturbanceReasonableModerate intensityVault settlement2616.24023.83032.76040.430
Xianghe tunnel D1K150+455[38]Short step excavation methodModerate disturbanceReasonableModerate intensityVault settlement2618.76029.30037.32044.580
Baozhen tunnel DK73+300[38]Three-step excavation methodModerate disturbanceReasonableModerate intensityHorizontal convergence1819.02035.89050.57063.300
Huama tunnel DK302+177[39]Three-step excavation methodModerate disturbanceUnreasonableModerate intensity **Horizontal convergence3221.00046.00058.00065.000
Baitanwu tunnel ZK7+815[40]Ultra-short step methodModerate disturbanceUnreasonableLow intensityVault settlement5630.70045.10053.30060.600
), ArticleFig(id=1274369041347175061, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Variable stateFactor node
Surrounding rock classification X1 Surrounding rock lithology X2Surrounding rock structure X3Initial in-situ stress state X4Dip angles of dominant discontinuities X5/(°)Tunnel burial depth X6/mTunnel excavation span X7/m
L1Ⅰ, ⅡLimestone, Dolomite, etc.Intact, Relatively IntactGeneral in-situ stress>75<50Small (5~8.5)
L2Glutenite, Basalt, etc.CrackedHigh in-situ stress30~7550~300Medium (8.5~12)
L3Clay rock, Shale, etc.Broken, ShatteredExtremely high in-situ stress<30>300Large (12~14)
L4Sandstone, Mudstone, etc.Extra large (>14)
Variable stateFactor node
Groundwater condition X8Groundwater control X9Advanced support activation X10Excavation method X11Surrounding rock disturbance intensity X12Support time X13Support strength X14
L1UndevelopedGoodGrouting to reinforce water blockageCRDMinor disturbanceReasonableHigh
L2Weakly developedMediumAdvanced small pipe grouting, pipe shedSidewall pit guide methodModerate disturbanceLateMedium
L3DevelopedBadCombined supportCore soil retention methodLarge disturbanceToo lateLow
L4UndevelopedBadNo advanced supportBenching tunnelling method (L4), Full-section excavation (L5)Large disturbanceToo lateLow
), ArticleFig(id=1274369041615610518, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 2, caption=

Influencing factor variables of surrounding rock deformation and their discretization

, figureFileSmall=null, figureFileBig=null, tableContent=
Variable stateFactor node
Surrounding rock classification X1 Surrounding rock lithology X2Surrounding rock structure X3Initial in-situ stress state X4Dip angles of dominant discontinuities X5/(°)Tunnel burial depth X6/mTunnel excavation span X7/m
L1Ⅰ, ⅡLimestone, Dolomite, etc.Intact, Relatively IntactGeneral in-situ stress>75<50Small (5~8.5)
L2Glutenite, Basalt, etc.CrackedHigh in-situ stress30~7550~300Medium (8.5~12)
L3Clay rock, Shale, etc.Broken, ShatteredExtremely high in-situ stress<30>300Large (12~14)
L4Sandstone, Mudstone, etc.Extra large (>14)
Variable stateFactor node
Groundwater condition X8Groundwater control X9Advanced support activation X10Excavation method X11Surrounding rock disturbance intensity X12Support time X13Support strength X14
L1UndevelopedGoodGrouting to reinforce water blockageCRDMinor disturbanceReasonableHigh
L2Weakly developedMediumAdvanced small pipe grouting, pipe shedSidewall pit guide methodModerate disturbanceLateMedium
L3DevelopedBadCombined supportCore soil retention methodLarge disturbanceToo lateLow
L4UndevelopedBadNo advanced supportBenching tunnelling method (L4), Full-section excavation (L5)Large disturbanceToo lateLow
), ArticleFig(id=1274369041682719383, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Variable stateDisplacement/mmSample size of crown settlement N1Sample size of horizontal convergence N2
L1≤504128
L250~1002057
L3100~1501926
L4150~200265
L5200~250104
L6250~30075
L7300~35063
L8350~40012
), ArticleFig(id=1274369041795965592, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 3, caption=

Discretization of displacement values of surrounding rock

, figureFileSmall=null, figureFileBig=null, tableContent=
Variable stateDisplacement/mmSample size of crown settlement N1Sample size of horizontal convergence N2
L1≤504128
L250~1002057
L3100~1501926
L4150~200265
L5200~250104
L6250~30075
L7300~35063
L8350~40012
), ArticleFig(id=1274369042131509913, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Monitored sectionTime nodeActual monitoring time/d
y0*y1y2y3y4y5y6y7y8y9y10y11y12y13y14**
Xianghe tunnel D1K153+340[38]L1L1L2L2L2L2L2L2L2L2L2L2L2L2L226
Xianghe tunnel D1K150+455[38]L1L2L2L3L3L3L3L3L3L3L3L3L3L3L326
Baozhen tunnel YDK79+117[38]L1L3L3L3L3L3L3L3L3L3L3L3L3L3L340
Huama tunnel DK302+177[39]L1L2L4L4L5L5L5L5L5L5L5L5L5L5L532
Baitanwu tunnel ZK7+815[40]L1L2L3L3L3L3L3L3L3L3L3L3L3L3L356
), ArticleFig(id=1274369042227978906, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 4, caption=

Partial time-series displacement sample data of TNP 4 vault settlement

, figureFileSmall=null, figureFileBig=null, tableContent=
Monitored sectionTime nodeActual monitoring time/d
y0*y1y2y3y4y5y6y7y8y9y10y11y12y13y14**
Xianghe tunnel D1K153+340[38]L1L1L2L2L2L2L2L2L2L2L2L2L2L2L226
Xianghe tunnel D1K150+455[38]L1L2L2L3L3L3L3L3L3L3L3L3L3L3L326
Baozhen tunnel YDK79+117[38]L1L3L3L3L3L3L3L3L3L3L3L3L3L3L340
Huama tunnel DK302+177[39]L1L2L4L4L5L5L5L5L5L5L5L5L5L5L532
Baitanwu tunnel ZK7+815[40]L1L2L3L3L3L3L3L3L3L3L3L3L3L3L356
), ArticleFig(id=1274369042307670683, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Level at time t1)Transition probability at time t
L1L2L3L4L5L6L7L8
L11.000000000
L20.3330.667000000
L300.3260.67400000
L4000.5710.4290000
L50000.2480.752000
L60000.0130.4860.50100
L7000000.3760.6240
L80000000.4570.543
), ArticleFig(id=1274369042471248540, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 5, caption=

Transition probability table of M-DBN

, figureFileSmall=null, figureFileBig=null, tableContent=
Level at time t1)Transition probability at time t
L1L2L3L4L5L6L7L8
L11.000000000
L20.3330.667000000
L300.3260.67400000
L4000.5710.4290000
L50000.2480.752000
L60000.0130.4860.50100
L7000000.3760.6240
L80000000.4570.543
), ArticleFig(id=1274369042571911837, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Model No.Reconstruction method of time-series curveConstruction method of transition networkTransition direction of the network
1TNP 1M-DBNForward transferred
2TNP 1M-DBNReverse transferred
3TNP 1OM-DBNForward transferred
4TNP 1OM-DBNReverse transferred
5TNP 2M-DBNForward transferred
6TNP 2M-DBNReverse transferred
7TNP 2OM-DBNForward transferred
8TNP 2OM-DBNReverse transferred
9TNP 3M-DBNForward transferred
10TNP 3M-DBNReverse transferred
11TNP 3OM-DBNForward transferred
12TNP 3OM-DBNReverse transferred
13TNP 4M-DBNForward transferred
14TNP 4M-DBNReverse transferred
15TNP 4OM-DBNForward transferred
16TNP 4OM-DBNReverse transferred
), ArticleFig(id=1274369042647409310, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 6, caption=

Test program for valuation of DBN model construction methods

, figureFileSmall=null, figureFileBig=null, tableContent=
Model No.Reconstruction method of time-series curveConstruction method of transition networkTransition direction of the network
1TNP 1M-DBNForward transferred
2TNP 1M-DBNReverse transferred
3TNP 1OM-DBNForward transferred
4TNP 1OM-DBNReverse transferred
5TNP 2M-DBNForward transferred
6TNP 2M-DBNReverse transferred
7TNP 2OM-DBNForward transferred
8TNP 2OM-DBNReverse transferred
9TNP 3M-DBNForward transferred
10TNP 3M-DBNReverse transferred
11TNP 3OM-DBNForward transferred
12TNP 3OM-DBNReverse transferred
13TNP 4M-DBNForward transferred
14TNP 4M-DBNReverse transferred
15TNP 4OM-DBNForward transferred
16TNP 4OM-DBNReverse transferred
), ArticleFig(id=1274369042710323871, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Reconstruction method of time-series curveVault settlementHorizontal convergence
M-DBNOM-DBNM-DBNOM-DBN
Forward transitionBackward transitionForward transitionBackward transitionForward transitionBackward transitionForward transitionBackward transition
TNP 161.2669.1360.3767.9671.1571.0873.1673.52
TNP 257.6364.3759.4163.3464.7365.1165.4365.27
TNP 354.0858.2658.0762.7465.0665.2761.1361.58
TNP 467.6572.5868.3174.1286.8286.5985.9385.19
), ArticleFig(id=1274369042928427680, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 7, caption=

Average values of the ACC at time nodes on each displacement time-series curve in the DBN model%

, figureFileSmall=null, figureFileBig=null, tableContent=
Reconstruction method of time-series curveVault settlementHorizontal convergence
M-DBNOM-DBNM-DBNOM-DBN
Forward transitionBackward transitionForward transitionBackward transitionForward transitionBackward transitionForward transitionBackward transition
TNP 161.2669.1360.3767.9671.1571.0873.1673.52
TNP 257.6364.3759.4163.3464.7365.1165.4365.27
TNP 354.0858.2658.0762.7465.0665.2761.1361.58
TNP 467.6572.5868.3174.1286.8286.5985.9385.19
), ArticleFig(id=1274369043003925153, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Case No.Factor node value
X1X2X3X4X5X6X7X8X9X10X11X12X13X14
1L3L1L3*L3L1L4L3L2L4L4L2L2
2L3L2L3L1L2L2L3L1L2L2L4L2L1L2
), ArticleFig(id=1274369043075228322, tenantId=1146029695717560320, journalId=1272208980697911299, articleId=1274300228095205417, language=CN, label=Table 8, caption=

Values of influencing factors of surrounding rock deformation

, figureFileSmall=null, figureFileBig=null, tableContent=
Case No.Factor node value
X1X2X3X4X5X6X7X8X9X10X11X12X13X14
1L3L1L3*L3L1L4L3L2L4L4L2L2
2L3L2L3L1L2L2L3L1L2L2L4L2L1L2
)], attaches=null, journal=Journal(id=1272208039877775360, delFlag=0, nameCn=岩石力学与工程学报, nameEn=Chinese Journal of Rock Mechanics and Engineering, nameHistory1=null, nameHistory2=null, issn=1000-6915, eissn=null, cn=42-1397/O3, 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=jE4tRO1Yjzxdww4NHlcaDw==, journalPrice=null, startedYear=null, abbrevIsoEn=Chinese Journal of Rock Mechanics and Engineering, journalRemark=null, publicationField=null, createdTime=1781247633193, updatedTime=1781247992638, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=C, firstLetterEn=C, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=jE4tRO1Yjzxdww4NHlcaDw==, picEn=yE9q2yMZJ9fDOBKxZi3AEg==, jcr=null, cjcr=null, exts=[JournalExt(id=1272209547574849641, language=CN, name=岩石力学与工程学报, 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=1781247992652, updatedTime=1781247992652, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://rockmech.whrsm.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://rockmech.whrsm.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://rockmech.whrsm.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1272209547620986986, language=EN, name=Chinese Journal of Rock Mechanics and Engineering, 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=1781247992663, updatedTime=1781247992663, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://rockmech.whrsm.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://rockmech.whrsm.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://rockmech.whrsm.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1272208980697911299, websiteList=[Website(id=1272211671482315193, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1272208980697911299, 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/yslxygcxb/CN, language=CN, createTime=1781248499031, createBy=18614031015, updateTime=1781248606948, updateBy=18614031015, name=岩石力学与工程学报-中文, tplId=1146099689490845704, title=岩石力学与工程学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1272212263650931197, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=articleTextType, value=kx, createTime=1781248640215, updateTime=1781248640215, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263629959674, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=banner, value=null, createTime=1781248640210, updateTime=1781248640210, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263667708416, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=grayFlag, value=0, createTime=1781248640219, updateTime=1781248640219, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263625765369, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=logo, value=https://castjournals.cast.org.cn/joweb/yslxygcxb/CN/file/pic?fileId=ZHPZ7MYfNE48n4/JFzLa0w==, createTime=1781248640209, updateTime=1781248640209, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263688679938, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=minRunFlag, value=0, createTime=1781248640224, updateTime=1781248640224, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263642542588, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/yslxygcxb/CN/file/pic, createTime=1781248640213, updateTime=1781248640213, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263676097025, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=silenceFlag, value=0, createTime=1781248640221, updateTime=1781248640221, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263638348283, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1781248640212, updateTime=1781248640212, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263659319806, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=themeColor, value=null, createTime=1781248640217, updateTime=1781248640217, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212263663514111, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671482315193, code=themeStyle, value=null, createTime=1781248640218, updateTime=1781248640218, creator=18614031015, updator=18614031015)]), Website(id=1272211671708807615, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1272208980697911299, 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/yslxygcxb/EN, language=EN, createTime=1781248499085, createBy=18614031015, updateTime=1781248602258, updateBy=18614031015, name=岩石力学与工程学报-英文, tplId=1146101810881728533, title=Chinese Journal of Rock Mechanics and Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1272212222580306419, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=articleTextType, value=kx, createTime=1781248630423, updateTime=1781248630423, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222563529200, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=banner, value=null, createTime=1781248630419, updateTime=1781248630419, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222597083638, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=grayFlag, value=0, createTime=1781248630427, updateTime=1781248630427, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222550946287, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=logo, value=https://castjournals.cast.org.cn/joweb/yslxygcxb/EN/file/pic?fileId=ZHPZ7MYfNE48n4/JFzLa0w==, createTime=1781248630416, updateTime=1781248630416, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222605472248, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=minRunFlag, value=0, createTime=1781248630429, updateTime=1781248630429, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222576112114, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/yslxygcxb/EN/file/pic, createTime=1781248630422, updateTime=1781248630422, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222601277943, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=silenceFlag, value=0, createTime=1781248630428, updateTime=1781248630428, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222571917809, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1781248630421, updateTime=1781248630421, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222588695028, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=themeColor, value=null, createTime=1781248630425, updateTime=1781248630425, creator=18614031015, updator=18614031015), WebsiteProps(id=1272212222592889333, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1272211671708807615, code=themeStyle, value=null, createTime=1781248630426, updateTime=1781248630426, creator=18614031015, updator=18614031015)])], journalTitle=岩石力学与工程学报, weixinUrl=null, journalUrl=https://rockmech.whrsm.ac.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Chinese Journal of Rock Mechanics and Engineering, journalPhotoCn=jE4tRO1Yjzxdww4NHlcaDw==, journalPhotoEn=yE9q2yMZJ9fDOBKxZi3AEg==, journalFirstLetter=C, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/yslxygcxb/CN/10.3724/1000-6915.jrme.2025.0477, detailUrlEn=https://castjournals.cast.org.cn/joweb/yslxygcxb/EN/10.3724/1000-6915.jrme.2025.0477, pdfUrlCn=https://castjournals.cast.org.cn/joweb/yslxygcxb/CN/PDF/10.3724/1000-6915.jrme.2025.0477, pdfUrlEn=https://castjournals.cast.org.cn/joweb/yslxygcxb/EN/PDF/10.3724/1000-6915.jrme.2025.0477, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于贝叶斯网络的隧道围岩变形动态预测模型
收藏切换
PDF下载
汪洪星 1 , 李珂瑶 1 , 张超 2, * , 阮俊浩 1 , 王丽萍 1 , 刘伟 3 , 巫尚蔚 1
岩石力学与工程学报 | 数值模拟与人工智能 2026,45(2): 553-577
收起
收藏切换
岩石力学与工程学报 | 数值模拟与人工智能 2026, 45(2): 553-577
基于贝叶斯网络的隧道围岩变形动态预测模型
全屏
汪洪星1 , 李珂瑶1, 张超2, * , 阮俊浩1, 王丽萍1, 刘伟3, 巫尚蔚1
作者信息
  • 1.重庆科技大学 安全科学与工程学院,重庆 400041
  • 2.中国科学院武汉岩土力学研究所 岩土力学与工程安全全国重点实验室,湖北 武汉 430071
  • 3.重庆大学 资源与安全学院,重庆 400044
  • WANG Hongxing (1983–), associate professor, is engaged in engineering disaster prediction and prevention based on machine learning algorithms. E-mail:

    汪洪星(1983–),现任副教授,主要从事基于机器学习算法的工程灾害预测和防控方面的研究工作。E-mail:

通讯作者:

* 张超(1978–),现任研究员,主要从事岩土工程灾害防控方面的研究工作。E-mail:
Dynamic prediction model of tunnel surrounding rock deformation based on Bayesian network
Hongxing WANG1 , Keyao LI1, Chao ZHANG2, * , Junhao RUAN1, Liping WANG1, Wei LIU3, Shangwei WU1
Affiliations
  • 1.School of Safety Science and Engineering, Chongqing University of Science and Technology, Chongqing 400041, China
  • 2.State Key Laboratory of Geomechanics and Geotechnical Engineering Safety, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan, Hubei 430071, China
  • 3.School of Resource and Safety Engineering, Chongqing University, Chongqing 400044, China
出版时间: 2026-02-01 doi: 10.3724/1000-6915.jrme.2025.0477
文章导航
收藏切换

基于现场监测位移数据驱动的隧道围岩变形动态预测方法,存在着较大的局限性和滞后性。综合利用隧道工程建设资料蕴含的物理信息和位移时序曲线蕴含的数学信息,通过物理信息机器学习的思想推导动态贝叶斯网络(dynamic Bayesian network,DBN)模型的构建方法,实现围岩变形的动态预测。通过离散化处理、位移时序曲线重构等方式,构建包含物理信息数据和极限位移数据的静态样本库,及包含物理信息数据和位移时序曲线数据的动态样本库。基于静态样本特征,改进K2-score算法建立极限位移预测的静态BN模型。基于静态BN模型和动态样本特征,融合平稳随机过程约束和Markov过程约束等先验信息,推导Markov-DBN的物理数据双驱动建模方法。融合约束增强优化的先验信息,构建优化Markov-DBN模型。5折交叉验证试验表明:Markov-DBN模型的预测能力随时间增加而快速降低,网络转移方向对其影响很大;优化Markov-DBN模型建立了所有时刻目标节点与变形影响因素节点的增强约束联系,其预测能力不随时间发生明显弱化,不受网络转移方向的影响,远高于Markov-DBN模型。案例分析表明:隧道工程施工前和施工初期,优化Markov-DBN模型即可实现位移时序曲线的有效预测,克服了传统方法的局限性和滞后性;隧道施工过程中,实时输入现场监测位移数据,可实现优化Markov-DBN模型的自我更新和围岩变形的动态预测。

隧道工程  /  围岩变形动态预测  /  多源信息融合  /  动态贝叶斯网络  /  物理数据混合建模  /  模型自我更新

Significant limitations and hysteresis are presented in dynamic prediction methods driven by on-site monitored displacement data for tunnel surrounding rock deformation. By comprehensively utilizing the physical information contained in tunnel construction project documents and the mathematical information from displacement time-series curves, a modelling method based on the dynamic Bayesian network (DBN) was developed using the concept of physical information machine learning (PIML) to achieve dynamic predictions of surrounding rock deformation. Through discretization processing and reconstruction of displacement time-series curves, a static sample database was established by combining physical information data with ultimate displacement data, while a dynamic sample database was created by integrating physical information data with displacement time-series curve data. Based on the characteristics of the static samples, the K2-score algorithm was improved to construct a static Bayesian network (BN) model for ultimate displacement prediction. Utilizing the static BN model and the characteristics of the dynamic samples, physical-data dual-drive modelling methods for the Markov DBN were derived by incorporating prior information, including the constraints of steady-state random processes and Markov process constraints. By integrating prior information for constraint-enhanced optimization, the optimized Markov DBN model was established. Five-fold cross-validation tests revealed that the prediction capability of the Markov DBN model decreased rapidly over time and that the network transition direction significantly affected this capability. In contrast, the prediction ability of the optimized Markov DBN model remained robust over time, was unaffected by the network transition direction, and significantly exceeded that of the Markov DBN model, as the optimized model enhanced constraint connections between target nodes and influencing factor nodes throughout the entire timeframe. Through engineering case analysis, it was concluded that before and during the early stages of tunnel construction, the optimized Markov DBN model could effectively predict displacement time-series curves, overcoming the limitations and hysteresis inherent in traditional methods. Furthermore, during construction, self-updating of the optimized Markov DBN model and dynamic predictions of surrounding rock deformation could be achieved by inputting the on-site monitored displacement data.

tunnel engineering  /  dynamic prediction of surrounding rock deformation  /  multi-source information fusion  /  dynamic Bayesian network  /  physical-data hybrid modelling  /  model self-update
汪洪星, 李珂瑶, 张超, 阮俊浩, 王丽萍, 刘伟, 巫尚蔚. 基于贝叶斯网络的隧道围岩变形动态预测模型. 岩石力学与工程学报, 2026 , 45 (2) : 553 -577 . DOI: 10.3724/1000-6915.jrme.2025.0477
Hongxing WANG, Keyao LI, Chao ZHANG, Junhao RUAN, Liping WANG, Wei LIU, Shangwei WU. Dynamic prediction model of tunnel surrounding rock deformation based on Bayesian network[J]. Chinese Journal of Rock Mechanics and Engineering, 2026 , 45 (2) : 553 -577 . DOI: 10.3724/1000-6915.jrme.2025.0477
隧道工程施工时常发生主体结构开裂、围岩垮塌等工程灾害,造成人员伤亡和工期延误。隧道工程灾害具有影响因素多、不确定性强等特点,其发生都伴随着围岩的持续变形[1]。根据国际岩石力学学会的定义[2]:“围岩变形是地质环境因素和施工因素共同作用下的动态非线性演化过程,是一种与时间相关的变形行为,……,具有复杂性、不确定性、多样性和时变性等特征”。进行围岩变形的精准预测,以变形速率、累计变形量、变形发展模式等动态特征指标对隧道施工安全进行预判,是防止隧道工程灾害发生的有效途径[3]。目前,围岩变形预测方法大致可以分为2类:以围岩的物理力学指标为基础的力学分析方法,和以围岩工程资料或现场位移监测为基础的数据驱动预测方法[4-6]
力学分析方法通过经验理论公式或弹黏塑性本构方程进行解析计算或数值计算,分析围岩变形的演化特征,实现围岩变形的预测,如:O. Aydan等[7]提出运用切向相对应变来预测围岩挤压变形潜在趋势;K. Bian等[8]开展了一系列的地应力试验和水–力耦合模拟试验,获取了卸荷条件下的围岩变形水–力耦合本构模型,提出了隧道工程开挖过程中岩体膨胀变形的数值分析法。由于计算公式的简化和围岩物理力学参数的不确定性,当越来越多的隧道工程面临超长、深埋、高地应力等复杂地质环境时,力学分析方法已不能准确指导隧道工程施工[9]。基于围岩工程资料的数据驱动预测方法,运用机器学习算法探求围岩变形的内在规律从而实现围岩变形的预测,在工程中得到越来越广泛的应用[10-11]
基于工程资料的数据驱动预测方法,以地质勘查报告和工程设计资料为基础,根据围岩级别、隧道埋深、施工方法等样本数据,探索围岩变形趋势与影响因素之间的内在关系,构建变形预测模型,如:X. D. Feng和R. Jimenez[12]使用贝叶斯网络(Bayesian network,BN)模型预测隧道挤压大变形的发生概率;X. G. Wu等[13]建立了基于BN的地铁隧道施工风险模型,能够在施工前实现隧道坍塌变形风险的预测。上述方法计算隧道围岩达到极限变形状态(大变形、坍塌变形等)的概率,为隧道工程的选址、开挖支护工法的选择等提供指导[14]。但是,上述方法无法给出围岩变形的动态变化过程,无法提取变形速率、变形发展模式等动态特征指标,无法实现围岩稳定性的动态预测。
根据新奥法的施工理念,采用监控量测的方法对隧道施工过程中的围岩变形进行监测、预测和分析,是保障隧道施工安全的最有效措施[15]。基于监测位移的数据驱动预测方法,应用机器学习算法探索变形与时间之间的数学关系,构建围岩变形预测模型,可实现围岩稳定性的动态预测。该类方法的重点在数据的获取和算法模型的构建,如:T. G. Feng等[16]提出了一种人工蜂群算法搜索策略改进随机森林(random forest,RF)模型,通过现场监测位移数据对该模型进行训练,实现了隧道围岩变形的高精度预测;S. Chen等[17]通过围岩变形实测数据构建混沌时间序列,建立混沌时间序列–模糊神经网络模型进行隧道围岩变形预测,预测的精度较高且稳定性较强。但是,该类预测方法具有一定的滞后性和局限性[2]:一方面,需要较长监测时间后,才可采集大量数据进行模型训练和预测计算,无法对隧道施作初期变形做出及时的预测,更无法在隧道施作之前的设计阶段进行围岩变形的预测[18];另一方面,仅凭监测数据的数学信息进行围岩变形的预测,没有合理利用隧道的地质环境条件、隧道结构、施工方法等物理信息,导致预测效果具有一定的片面性,无法根据现场物理信息的变化进行预测结果的及时更新,且无法应对现场位移监测经常出现的数据缺失问题[19]
基于工程资料的数据驱动预测方法,有效地利用了围岩变形影响因素的物理信息;基于现场监测位移的数据驱动预测方法,有效地利用了监测位移时序数据的数学信息;但是,二者没有实现互补。如果融合上述2种方法,综合利用隧道工程资料的物理信息和监测位移时序曲线的数学信息,构建围岩变形的动态预测模型,将有效克服上述2种预测方法的不足,实现围岩变形的全过程动态预测。融合多源信息的机器学习(machine learning,ML)动态模型已有相关的研究,为变形动态预测模型的构建提供了探索的方向,如:张 振等[20]通过融合雷诺平均(Reynolds-averaged Navier-Stokes equations,RANS)流场物理控制方程与大涡模拟(large eddy simulation, LES)温度场数据,构建湍流普朗特数的物理信息神经网络(physic information neural network,PINN),可动态调整湍流热扩散强度以实现温度场的精准预测;卢鑫月等[21]提出了基于动态贝叶斯网络(dynamic Bayesian network,DBN)和模糊综合评价的地铁隧道施工动态风险评估方法,该方法综合应用风险因素指标和现场监测位移指标来计算施工风险,可根据实时位移监测数据实现风险概率的动态更新;但是,该方法无法实现位移时序曲线的预测。
DBN模型可结合物理先验信息设置约束条件,融合样本数据统计相关性进行混合建模,以克服稀疏/受限样本数据环境带来的模型训练挑战,是一种高效的物理信息机器学习(physic information machine learning,PIML)模型,可提升模型预测准确率、计算效率和物理一致性[13]。相较于PINN[20]等融合物理控制方程和样本数据构建ML模型的方法,应用DBN方法构建隧道围岩变形动态预测模型更具优势:(1) 现有物理控制方程(本构模型)无法准确地描述隧道围岩变形的发生规律;而根据隧道围岩变形的物理信息数据和历史监测位移的数学信息数据,即可构建变形动态预测的DBN模型,且将显著提高计算效率;(2) DBN模型没有严格的输入项和输出项[11]——在DBN模型训练和测试时,历史监测位移数据是输出项;在DBN模型应用时,早期监测位移数据可以是输出项或者输入项,而未来监测位移数据则是输出项,从而实现变形的动态预测和更新。由此可见,DBN非常契合融合隧道工程资料、历史监测位移数据、现场监测位移数据等多源信息数据的变形动态预测模型的构建和应用。
因此,本研究致力于探寻适用于隧道围岩变形动态预测的DBN模型构建和应用方法。采集隧道工程案例,融合工程资料物理信息数据和历史监测位移时序曲线的数学信息数据构造样本库;基于PIML的思想,融合样本数据和物理先验信息约束,推导DBN建模方法;设置对比试验,评价和筛选最优的DBN模型;通过案例分析和讨论,验证DBN模型的工程有效性,展示其动态预测和实时更新的方法。期望构建的围岩变形动态预测DBN模型,能够有效克服现有预测方法的滞后性和局限性等不足,实现隧道围岩变形的全过程高精度动态预测。
贝叶斯网络也称为贝叶斯信度网络(Bayesian confidence network,BCN),由有向无环图(directed acyclic graph,DAG)和条件概率表(conditional probability table,CPT)组成,可以用于隧道围岩变形动态预测模型的构建,如图1所示。图1(a)中,变量X1X2,…,Xn为自变量(因素节点),表示围岩变形的影响因素;Y为因变量(目标节点),表示围岩变形量。有向边由父节点指向子节点;自变量指向因变量的有向边,表示影响因素对围岩变形的直接影响;自变量之间的有向边,表示影响因素之间的指向性作用关系。图1(a)中的DAG,称为BN结构,表示节点间的指向性逻辑关系;图1(b)中的CPT,称为BN参数,表示不同父节点组合状态下,子节点任一取值状态的取值概率。BN模型表征所有节点之间的量化逻辑关系,不存在绝对的输入项(自变量)或输出项(因变量)——本文的研究将证明,BN模型可以输入因素节点计算目标节点,也可以输入目标节点计算因素节点。图1(a),(b)所示的BN模型没有考虑目标系统随时间的动态变化过程,称为静态贝叶斯网络模型(static Bayesian network model,SBN),只能用来构建表征隧道围岩极限变形量与影响因素之间预测关系的静态预测模型。
动态贝叶斯网络是SBN在时间上的扩展,由初始网络B0和转移网络Bt组成,如图1(c)所示。初始网络即SBN,表示节点之间的初始逻辑关系;转移网络是后续时间节点上的BN模型。图1(c)中,t时刻的转移网络Bt包含了(x1t  xntyt)等节点构成的SBN,还包含了t时刻之前的初始网络B0和转移网络B1B2  Bt1t时刻网络节点的影响。与初始网络B0类似,转移网络Bt也包含了转移网络结构(DAG)和转移网络参数(CPT)。可见,可以基于隧道围岩极限变形预测的SBN模型,构建位移时序曲线预测的DBN模型,实现围岩变形的动态预测。由于受到t时刻之前系统状态的影响,转移网络Bt的结构和参数都是非常复杂且庞大的,需要加以一定的约束使其简化,以增加其适用性。
对于图1(a),(b)所示的SBN模型,应用贝叶斯公式进行围岩变形量的计算,贝叶斯公式[13]
P(Y|X=e)=P(Y)P(X=e|Y)P(X=e)=P(X=e|Y)P(Y)P(X=e)=P(X=eY)P(X=e)
式中:X=(x1  xn)为证据项(自变量),表示围岩变形的影响因素;Y为目标项(因变量),表示围岩变形量;P(Y|X=e)为后验概率,即已知X的某个新证据e情况下,Y为某个取值状态的概率;P(Y)为先验概率,是考虑X的新证据e之前变量Y为某个取值状态的概率,可以根据历史数据学习获得;P(X=e|Y)X的似然度,也是一个条件概率,也可以基于历史数据计算得到;P(X=e)X的新证据发生的概率;P(X=e|Y)(x1 xny)的联合概率分布。
不同于朴素贝叶斯和半朴素贝叶斯直接根据式(1)的似然度P(X=e|Y)进行计算,BN模型的网络结构极其复杂,需要结合Markov链规则和条件独立性规则对式(1)最右端的联合概率分布P(X=eY)进行分解,通过CPT中的条件概率来进行联合概率的计算,计算式[22-23]
P(x1  xny)=P(x1)P(x2|x1)P(xn|x1x2xn1)P(y|x1x2  xn)=i=1nP[xi|π(xi)]P[y|π(y)]
式中:π(xi)为变量xi的父节点集合;π(y)y的父节点集合;P[xi|π(xi)]P[y|π(y)]xiy的条件概率,可以从BN模型的CPT中获取。
同理,应用Markov链规则和条件独立性规则对式(1)最右端的概率分布P(X=e)进行分解,将分解后的结果与式(2)一起代入式(1),得到用条件概率表示的计算式[13]
P(Y|X=e)=P(X=eY)P(X=e)=i=1nP[xi|π(xi)]P[y|π(y)]yΠi=1nP[xi|π(xi)]P[y|π(y)]
根据式(3),即可在获取证据项X = e的情况下,从BN模型的CPT中提取条件概率计算目标项Y的取值概率。
对于如图1(c)所示的DBN模型,将DBN模型等价为多个SBN模型组合而成的复合BN模型,针对所有时刻节点M=[XTYT]=(x11x21  xn1y1x12x22  xn2y2  x1Tx2T  xnTyT)构建DBN的DAG和CPT后,应用与式(3)同样的方式,构建下式[21]实现不同时刻围岩变形量Yt的计算:
P(YT|XT=e)=P(XT=eYT)P(XT=e)=t=0TΠi=1nP[Xti|π(Xti)]Πt=0TP[yt|π(yt)]yΠt=0TΠi=1nP[Xti|π(xti)]Πt=0TP[yt|π(yt)]
式中:YT为作为目标项在t=(0  1  T)时刻的围岩位移值,XT=e为作为证据项的影响因素在t=(0  1  T)时刻的取值,P(XT=eYT)YTXT的联合分布概率,P(XT=e)XT的边缘分布概率,P[xti|π(xti)]为任一t时刻第i个因素节点的条件概率,P[yt|π(tt)]为任一t时刻目标节点的条件概率。
式(4)和(3)的构建方法,是完全相同的,其区别仅在于:式(3)描述的是某一个时刻,隧道围岩位移值与其影响因素之间的关系;而式(4)描述了动态系统中,在给定所有时刻影响因素取值的情况下,目标变量在各时刻取值的联合概率分布。
当BN结构和参数不是很复杂的时候,可以从BN模型最顶层的根节点(没有父节点的节点)开始向下逐层计算实现式(3),(4)的精确求解,也可以使用变量消元法对其进行精确求解。当模型比较复杂的时候,可使用团树法(joint tree method)对式(3),(4)进行精确求解;当模型过于复杂的时候,需要使用吉布斯采样法、Markov链蒙特卡罗法(Markov chain Monte Carlo method,MCMC)、变分法等近似求解法[24]。本研究的模型复杂度适中,将采用团树法对其进行求解,求解速度较快。
一般的情况下,围岩变形影响因素XT=(x0x1x2  xi  xT),可在隧道施工前的设计阶段通过地质勘查资料和工程设计资料确定下来的,并且大部分情况下不同时刻的xit是保持一致的。因此,基于式(4),在隧道工程的设计阶段,就可输入围岩影响因素取值,计算得到任一时刻t=(0  1  T)的变形量取值,获取位移YT的时序曲线,用做隧道工程稳定性判断、二次衬砌施作时机选择和监测预警值设置的依据。
在隧道工程的施工阶段,从获取第1个时刻的监测位移数据Yreal0开始,将每一个时刻tm的监测位移数据Yreal=(Yreal0Yreal1Yreal2  Yrealtm)作为新的证据项代入DBN模型中,可对位移时序曲线上后续的Ypred=(Ypredtm+1Ypredtm+2  Ypredtm+m)进行更新,如下所示:
P[Ypred|(XTYreal)=e]=P[(XTYreal)=eYpred]P[(XTYreal)=e]={{t=0Ti=1nP[xti|π(xti)]}{t=0tmP[yrealt|π(yrealt)]}{t=tmTP[ypredt|π(ypredt)]}}/{ypred{t=0Ti=1nP[xti|π(xti)]}{t=0tmP[yrealt|π(yrealt)]}{t=tmTP[ypredt|π(ypredt)]}
在施工阶段,如果现场补充勘探或者超前钻探得到的地质环境条件,与前期工程地质勘察中提供的地质环境条件不同;或者隧道施作现场的工程结构或开挖支护方法发生了改变,可将上述更新后的工程地质条件、工程结构特征或开挖支护方法等,与监测位移数据Yreal一起形成新的证据项(XnewYreal)=e代入式(5)中,对原始的证据项(X Yreal)=e进行替换,计算得出更新后的P[Ypred| (XnewYreal)=e],从而实现对现场新发现证据项的有效利用。
综上所述,通过采集隧道工程的工程资料数据和历史监测位移数据构造融合围岩动态变形的物理信息和数学信息的样本库,可以建立隧道围岩变形动态预测的DBN模型。基于构建的DBN模型,在隧道工程施工之前或者施作早期,仅输入工程资料数据或者辅以少量的早期监测数据,可实现施工期间位移时序曲线的预测;施工期间,当获取了更多的现场监测位移数据,或现场的工程地质数据或施工方法发生变化时,可以及时输入新增的监测位移值或变更的工程资料数据,实现位移时序曲线的动态更新。构建的隧道围岩变形动态DBN模型,将有效克服现有方法的动态预测能力不足、滞后性和局限性等问题。
通过知网数据库,设置包括影响因素和时序位移数据等关键词的检索策略,采集文献;结合工程资料搜集,获取近几十年来隧道工程案例数据共400组。每组案例数据中包含了围岩变形影响因素取值和历史监测位移时序曲线。基于重要性和易获取性原则,通过理论分析和文献计量分析等方法,计算影响因素的重要程度系数,筛选了包括地质环境条件、隧道结构特征、施工条件等多方面共14个影响因素,作为围岩变形预测BN模型的因素变量。为满足BN模型构建的需要,从400组案例数据中筛选出数据较为完备的130组案例数据,用于模型训练样本库的构建,部分案例数据如表1所示。
影响因素变量既有离散型变量也有连续型变量和描述型变量。但是,SBN和DBN更加青睐于对离散型变量进行计算。因此,参照张广泽等[25-36]的研究成果,对所有变量进行离散化,通过一个离散取值表征一个取值区间(见表2),如:采用规范[26]中对围岩级别的定义,对“围岩级别”这个变量进行离散化;基于规范[27]对隧道埋深、隧道开挖跨度的分类,分别对“隧道埋深”、“隧道开挖跨度”等变量进行离散化。离散化的区间划分越小,对变量的信息挖掘更准确;但是,过小的区间划分将导致变量取值状态的样本数量过少,影响机器学习模型训练的效果。因此,本研究在进行变量离散化时,尽量遵从现有规范标准或研究成果对上述变量的区间划分方式,既可保障BN模型训练的合理性,也更符合工程技术人员的技术习惯。变量离散化后可契合BN模型的计算,同时可克服影响因素取值的随机性和不确定性对计算结果的影响。
表2可知,14个影响因素均被划分3~4个取值状态(仅X11被划分为5个取值状态)。每个变量的离散化基本上覆盖了所有可能的取值状态。取值状态L1L2L3L4L5表征了影响因素导致隧道围岩发生变形的难易程度,L1表示最不易导致围岩变形,L5表示最容易导致围岩发生变形。每一个取值状态,都可从隧道工程的地勘报告、设计资料或评价资料等工程资料中获取。总体而言,影响因素的离散化符合工程应用习惯,易于进行针对性的取值。同时,SBN和DBN基于概率理论对表征取值区间的离散型变量进行概率计算,符合隧道工程全概率可靠性分析的要求[37]。根据表2所示的离散化处理方法,得到每个案例中每个影响因素的取值,据以构建影响因素样本库,代表了构建隧道围岩变形动态预测模型的物理信息数据。
根据隧道工程围岩变形的特征和变形监测的习惯,分别采集拱顶下沉和水平收敛2个变形量的时序曲线数据,如图2所示。对拱顶下沉和水平收敛的取值分布进行统计分析发现,拱顶下沉的取值分布状态与水平收敛区别较大,不具有同步性。同一个案例中,拱顶下沉和水平收敛不同步的情况也很多——同一隧道的拱顶下沉取值状态与水平收敛取值状态的差异很大,如图2所示。如果将拱顶下沉和水平收敛2个位移值放在同1个预测模型里面,各个影响因素的取值相同而2个位移值不同步,将导致数据学习的学习逻辑出现混乱,构建出的模型无法反映影响因素之间、影响因素与位移值之间的真实关系[41]。因此,本研究分别构建拱顶下沉的样本库和水平收敛的样本库,继而分别构建各自的BN预测模型。
图1(c)所示,DBN模型表达的是动态系统中SBN模型随时间的变化过程。理论上来说,时间是连续型的变量,DBN模型应该是无数个SBN模型的动态变化过程——其实现难度很大。如前所述,BN模型更加青睐于离散型变量的计算。因此,DBN中将连续性的时间也进行离散化,将DBN表达为有限个时间节点上SBN的集合:初始时间节点上的BN为初始网络B0,后续时间节点t上的BN为转移网络Bt;任何1个时间节点上的Bt包含了该时刻因素节点和目标节点之间的指向性连接,也包含了该时刻以前的BN与该时刻所有节点之间的指向性连接网络。
本文中,围岩位移时序曲线数据是以1 d为单位进行采集的(见图2),相当于是已经以1 d为单位对时间变量进行了离散。为了降低时间节点的数量,以降低DBN模型的复杂度,根据案例库中历史位移监测数据的特点,提出以下4种时间节点划分方法以实现位移时序曲线的重构:TNP(time node partition) 1,将所有的历史位移监测数据以3 d为一个时间节点进行划分;TNP 2,以5 d为一个时间节点进行划分;TNP 3,根据真实监测时间长度,将其等分为10个时间节点;TNP 4,将所有原始历史位移监测数据补充至最大天数70 d再以5 d为一个时间节点进行划分——对于稳定型位移时序曲线,可以直接将最终的稳定位移值(极限位移值)向后补充至70 d;对于其他类型的位移时序曲线,根据其线型特征分别选用线性插值函数、拉格朗日插值函数、样条插值函数(三次样条插值、自然插值、B样条插值等)等插值方法,对其进行外推插值,且通过均方根误差、平均绝对误差、相关系数等多个指标对插值误差和插值合理性进行综合评价分析,为每一个位移时序曲线确定最优的插值外推方法。
因为不同隧道有不同的工程要求,采集的时序曲线时长是不一致的——案例库中最短的监测时间为18 d,最长的监测时间为70 d。由于不同案例监测时长的不同,TNP 1和TNP 2构造的时序曲线存在着较多的缺省数据,将一定程度上影响模型的构建效果。TNP 3构造的时序曲线与真实时间不对应,导致时间节点的物理意义不明确。TNP 4有效地利用了历史监测位移数据的变化趋势,弥补了部分案例数据采集不充分的缺陷,但是其合理性仍然有待验证。
基于上述4种时间节点划分方式构造的位移时序曲线,取时序曲线最右端的累计值(极限变形量)作为隧道围岩变形量的静态表征,形成极限位移样本库进行SBN预测模型的构建;取时序曲线上不同时间节点的累计值作为变形量的动态表征,形成时序位移样本库进行DBN预测模型的构建。基于TNP1~3方法构造时序曲线形成的极限位移样本库是相同的,基于TNP4方法构造时序曲线的极限位移样本库是不同的。TNP1~4构造时序曲线的时序位移样本库都不同。
针对每一个时间节点处的位移值(包括拱顶下沉和水平收敛值),参照国内外相关规范和张广泽等[25]的研究成果,将位移值离散为8个取值区间,如表3所示。由表3可知,即使经过筛选,部分样本数据仍然存在空缺值。好在相对于其他的机器学习模型,BN模型对于存在空缺值的样本具有更好的处理能力;后续,还将融入隧道围岩变形的物理信息,结合数据信息进行混合建模,以保证模型的有效性。离散后的位移值,更加有利于BN模型的构建;据以构建的BN预测模型,可以根据现场工程资料和前期监测数据,正向推理得到离散型位移值,直接用于隧道变形的监测预警。表3所示的离散化操作符合工程应用习惯,且比大部分规范中的区间划分更为精细,可以提升变形预测模型在险情预警时的精度。但同时,也导致表3中的位移值样本分布不均衡——这是基于大量实测数据统计得到的客观结果,真实反映了隧道围岩变形的实际分布规律,不存在取样偏差问题。真实分布的样本数据,反映了隧道围岩位移值的真实分布规律,据以构建的BN模型更为符合真实情况。因此,本研究没有采用措施对样本数据进行调控,而是保持真实样本进行模型的构建和训练。
基于上述位移时序曲线重构和位移值离散的方法,可完成极限位移样本库和时序位移样本库的构建,构建的TNP 4拱顶下沉的部分时序位移样本如表4所示。时序位移样本库代表了构建隧道围岩变形动态预测模型的数学信息数据。融合上述影响因素样本库和极限位移样本库,得到据以构建极限位移预测SBN模型的静态样本库;融合影响因素样本库和时序位移样本库,得到据以构建时序位移预测DBN模型的动态样本库。
由此可见,构建的动态预测BN模型本质上是关于位移取值状态的分类预测模型,不是传统意义上的回归预测模型。工程上进行隧道变形监测预警时,基于变形预警值对监测位移数据进行分析——很多时候,变形预警值也是取值区间,即离散型变量。因此,基于构建的分类预测模型,根据现场工程资料和前期监测数据进行正向推理得到的离散型位移值,可以直接用于隧道变形监测预警。
为了克服小样本量给模型训练和测试带来的不确定性,引入K(5)折交叉验证试验[42-43]对BN模型进行训练和测试:把130组样本数据随机划分成5等份,轮流将每份数据都做一次测试集,其余的4份作为训练集,进行模型的训练和验证;交叉验证进行5次,取5次验证指标的平均值来评价模型的性能。需要注意的是,本研究所有DBN模型的网络结构都是根据130组样本构建的;5折交叉验证试验是针对各个模型的参数(条件概率表)进行的。
K2-score算法是常用的BN结构构建方法,其通过爬山法对可能存在的指向边进行搜索,通过下式[28]所示的评分函数对网络结构进行评价:
argmaxBS[P(BSD)]=P(BS)argmaxBSi=1n[j=1qi(ri1)!(Nij+ri1)!Nijk=1riNijk!]=P(BS)argmaxBS[i=1ng(xiπi)]
式中:P(BSD)为模型的评分值,评分值越高,模型性能越优,BS为BN结构,D为样本数据;P(BS)为模型的先验概率,可以通过历史数据获取,本研究中该先验概率取为定值;n为变量个数;qi为父节点集的状态等级个数;ri为变量的状态等级个数;Nijk为样本数据中变量xi的父节点集πi的状态为j时,该变量状态为k的数据量;Nij为样本数据中变量xi的父节点集πi的状态为j时,该变量所有状态的数据量。
K2-score算法构建SBN模型结构的主要流程为(见图3):首先,构建一个空网络或简易网络作为初始网络;接着,根据节点排序和节点的最大父节点数量,给每个节点添加(删除)父节点(默认为通过爬山法进行父节点的增减),观察网络结构的评分函数值变化——若添加或删除一个父节点后评分函数值增加,则保留该操作,否则复原并继续下一个父节点的添加或删除,直到遍历所有的父节点或者达到最大父节点数量,完成该节点的父节点增减操作;然后,继续下一个子节点的父节点添加(删除)操作。通过对每一个节点进行类似的操作直到遍历所有的节点,结束爬山法的搜索,完成SBN模型网络结构的构建。
常规的K2-score算法中,初始网络、节点排序和最大父节点数量通常是通过专家经验的方法进行确定,具有较大的主观性,存在计算量太大、计算效率低、随机性大等问题,最终导致计算结果不可靠。鉴于此,本研究通过样本数据进行影响因素的解释结构模型(interpretive structure model,ISM)分析,对初始网络、节点排序和最大父节点数量进行约束设置,实现K2-score算法的优化:(1)将影响因素的ISM作为K2-score算法的初始网络;(2)节点排序为[X2X3X1X4X6X11X7X10X8X12X13X5X9X14Y];(3)最大父节点数量为6。应用优化的K2-score算法,分别基于拱顶下沉静态样本库和水平收敛静态样本库——取TNP1~3方法构造时序曲线形成的极限位移,构建隧道围岩变形的SBN结构,如图4所示。
图4中,2个SBN结构都符合工程逻辑,如:岩性、围岩结构、初始地应力、主要结构面倾角和地层含水状况等因素,对围岩级别产生直接影响作用;开挖工法、开挖跨度及各工程地质因素造成围岩扰动,通过围岩扰动程度间接影响围岩变形;支护强度、支护时机、围岩扰动程度、围岩级别作为重要的关键因素,直接影响围岩变形等,都由SBN结构清楚地展示出来。但是,图4中的2个SBN模型结构也存在着部分不同的连接关系,如水平收敛SBN结构中,地层含水状况节点通过地下水控制节点间接影响围岩变形,而拱顶下沉SBN结构中两者均直接指向围岩变形节点;水平收敛SBN结构中,开挖工法与支护强度之间存在直接连接,而拱顶下沉SBN结构中则不存在此连接。上述差异是拱顶下沉和水平收敛样本数据深层次差异的直观体现,证明了应用优化K2-score算法深度挖掘历史数据进行SBN模型构建的必要性,同时证明了分别针对拱顶下沉和水平收敛构建预测模型的合理性。
尽管已经进行了初步筛选,样本库中仍然存在着较多的缺省值。此时,常规的精确计算方法如最大似然估计法(maximum likelihood estimation,MLE)和贝叶斯估计法等无法在样本存在缺省值时完成BN参数的计算,只能使用期望最大化法(expectation-maximization,EM)、MC法和Gaussian法等近似算法计算条件概率表。其中,EM法是最常用的方法,其算法思想是:(1)随机设置1个初始参数θ0;(2)根据θ0对样本中的缺省值进行填充形成完整的样本,通过MLE算法计算参数θ的期望对数似然函数,计算式[37]
lgL(θ|Dl)=l=1Nxlp(xl|Dlθt)lgp(Dlxl|θ)
式中:N为样本量;xl为样本Dl中缺省变量的集合;θt为当前迭代步骤得到的参数;p(xl|Dl θt)为BN参数为θt时,根据样本Dl计算得到的缺省值取值为xl的概率,当xl为空集时,p(xl|Dl θt)=1lgp(Dlxl|θ)为添加修补样本xl后形成完整样本(Dlxl)时参数θ的似然度,可以根据下式[19]进行计算:
lgp(Dlxl|θ)=i=1nj=1qik=1rilgp(Dlxl|θijk)=i=1nj=1qik=1riNijklgθijk
式中:θijk为根据添加修补样本xl后形成的完整样本(Dlxl)计算得到的,父节点集π(xi)为第j个取值组合时,节点xi取第k个状态的条件概率,k=1riθijk=1
(3)对数似然函数对θ求导,得到期望最大时的参数θ1;(4)根据θ1对原始样本中的缺省值进行更新,重复步骤(2)和(3)的操作,对θ进行迭代更新,得到θ2θ3θ4,…,θn;(5)直到对数似然函数(见式(7))的极值达到稳定,算法收敛完成。本质上,EM算法是MLE算法的迭代优化过程。
针对如图4所示的SBN的结构,基于静态样本库的130组样本,应用EM法进行数据学习得到的BN模型参数,如图4中每个节点下的概率值所示。图4中每个根节点的边缘概率即为条件概率;中间节点和叶节点等子节点的条件概率表非常大,无法直观地展示,因此通过下式[13]将条件概率转化为边缘概率进行展示:
P(xm)=xi=x1  xnyxixmP(x1  xny)=x1=x1  xnyxixm{j=1nP[xj|π(xj)]P[y|π(y)]}
对于完全动态的DBN,不同时间节点上的因素节点和目标节点的取值都会发生变化,不同时间节点上BN的结构和参数都不同。同时,对应图1中任一时间节点上的Bt,都可能受到前面时刻所有BN的影响,如图5(a)所示。对于完全动态的DBN模型,其构建难度和计算量是非常大的,不利于DBN的工程应用。
鉴于此,在应用DBN进行动态系统的分析和预测时,需要应用物理信息机器学习(PIML)的方法,对动态系统附加一定的约束条件,以降低DBN构建难度和计算量[44]。对于本研究中的PIML模型,有个几个概念需要明确:(1)融合物理信息和数学信息的变形动态预测模型中,“物理信息”和“数学信息”指的是样本中,工程资料数据蕴含的隧道围岩变形影响因素信息和位移时序曲线数据中蕴含的位移变化规律信息;(2)通过物理信息机器学习——物理数据双驱动的方法构建DBN模型时,“数据信息”指的是构建DBN模型时通过机器学习(machine learning,ML)算法挖掘的,样本数据中影响因素变量和位移时序曲线变量等所有变量的统计相关性信息;而“物理先验信息”指的是构建DBN模型时,为了弥补稀疏/受限的样本数据环境给ML带来的不足,引入对模型构建过程进行约束的专业领域内公认的物理先验信息,以提升模型的预测准确率、计算效率和物理一致性。本研究中,利用物理数据信息进行双驱动混合建模时,物理先验信息主要为服务于DBN结构、参数的训练和优化的相关约束信息。
对于隧道围岩变形这一动态系统,刘次华[45]的研究成果认为:影响因素XT=(x0x1x2  xi  xT)如地质环境条件、施工方法和隧道结构特征等,在t=(0  1  T)的有限时间范围内是不会发生明显变化的,即该系统的变化过程可以近似为1个平稳随机过程——平稳随机过程约束1:
在有限的时间段内,任一时刻t的系统状态是一致平稳的。即:不考虑t时刻以前系统作用的情况下,DBN模型中任一t时刻BN的网络结构与初始网络B0中完全相同。
因此,式(4)中各时刻影响因素的条件概率乘积项可以被简化,即:可将式(4)中除初始时刻外的i=1nP[xti|π(xti)]都将设置为1,从而得到表征不同时刻围岩变形量Yt计算的条件约束式:
P(YT|X=e)=P(X=eYT)P(X=e)=i=1nP[xi|π(xi)]t=0TP[yt|π(yt)]yi=1nP[xi|π(xi)]t=0TP[yt|π(tt)]
类似地,基于式(5),也可以得到表征根据现场监测位移值进行位移时序曲线更新的条件约束式:
P[Ypred|(XYreal)=e]=P[(XYreal)=eYT]P[(XYreal)=e]={i=1nP[xi|π(xi)]t=0tmP[yrealt|π(yrealt)]t=tmTP[ypredt|π(ypredt)]}/{ypredi=1nP[xi|π(xi)]t=0tmP[yrealt|π(yrealt)]t=tmTP[ypredt|π(ypredt)]}
此时,将DBN结构中时刻t(不包括初始时刻)的π(yt)π(yrealt)π(ypredt)中除π(y0)外的因素节点剔除,不会影响该时刻节点y的CPT计算。从而可得,XT=X=(x10x20  xn0),式(10),(11)中π(xi)即为初始网络B0中的π(xi),任一时刻t(不包括初始时刻)的π(yt)π(yrealt)π(ypredt)仅包含了前面所有时刻的节点y和初始网络B0中节点y的父节点,如图5(b)所示。
根据上述约束1确定的DBN,时刻t的目标节点y与前面所有时刻的目标节点y相关。当围岩变形监测的时间较长时,将导致DBN的结构异常复杂,目标节点的条件概率表过大。因此,进一步对DBN动态系统进行约束,假设其为Markov过程:事物经n次变动,其第n次结果仅与第(n1)次结果有关,与之前的结果无关;事物这种变化过程只与近期状态相关、与过去状态无关的性质被称为无后效性;n个具有无后效性的连续变动事物在变化过程中构成的集合即为Markov链,事物这种变化演绎的过程称为Markov过程。
因此,如果将目标节点y=(y1y2  yT)的变化过程当成Markov过程,则可以得到以下的Markov过程约束2:
假设概率的动态变化过程满足Markov性,即系统在t时刻的状态只与其在t-1时刻的状态相关,而与t-1时刻之前的状态无关,计算式为
P(yt|y1y2  yt1)=P(yt|yt1)
此时,式(10),(11)中,任一时刻t(不包括初始时刻)的π(yt)π(yrealt)π(ypredt)仅包含了最临近前一时刻的节点y,从而实现DBN结构的进一步简化:DBN结构只剩下初始网络结构B0和不同时刻围岩变形节点yt依次连接形成的转移网络结构(见图5(c)6)。图6为基于TNP 3重构时序曲线的130个动态样本,应用如图5(c)所示方法构建的拱顶下沉动态预测DBN模型;图中的条件概率表为5折交叉试验的第1组训练样本集通过EM算法进行计算得到的。将上述简化方法构建的DBN称为M-DBN。
基于上述方法构建的M-DBN结构,由B0和不同时刻目标节点间的连线——转移网络构成。B0的条件概率表是已知的,只剩下转移概率P(yt|yt1)需要确定。如前所述,围岩变形动态变化系统满足平稳随机过程和Markov过程约束,则可以得到以下的约束3:
假设任意2个相邻时刻的条件概率转移过程是平稳的,即转移概率P(yt|yt1)在整个DBN中保持不变。
Markov过程中,由于更新的信息不总是准确的,有可能会导致概率漂移。因此,转移概率的计算需要充分挖掘位移时序曲线的概率转移信息:将Markov链特征y的数列划分为N个状态,Ei(j)为数列的第i(j)种状态;数列中从t1时刻状态Ei转移到t时刻状态Ej的转移频数记为mijEi所处状态的频数为Mi(总转移频数),则转移概率P(yt|yt1)Transition probabilityTransition probability由所有样本中所有时间节点的位移状态转移数据计算得到,由下式[43]所示:
P(yt|yt1)=mijMi(i=1  2  Nj=1  2  N)
算例1:时序曲线重构方式为TNP 3,根据划分好的10个时间节点,由第1个时间节点至最后一个时间节点计算围岩变形节点的转移概率。应用拱顶下沉动态预测的5折交叉试验第1组训练样本集进行转移概率的计算,以L1等级为例:统计所有样本数据中L1等级在t1时刻的总转移频数,分别统计出在t时刻转移至L1L2L3L4L5L6L7L8等级的转移频数,各转移频数在总转移频数中的占比即为对应的转移概率,如表5所示。
B0加上不同时间节点间的转移网络和转移概率,即构成了M-DBN模型。但是,如果时间节点划分的方式发生变化,表5中的数据都会发生变化,从而导致M-DBN模型发生变化。
上述M-DBN的构建方法,极大地降低了复杂动态系统DBN模型的构建难度和计算量,具有很强的适用性。但是,因为任一时刻的yt只与前一时刻的yt1有关,而与围岩变形的影响因素(x1  x14)无关,容易陷入“数字游戏”的陷阱:(1) B0中的目标节点y0(x1  x14)产生了联系,对围岩变形的物理信息(影响因素取值)有一定程度的有效利用;当围岩变形监测时间较长时,后期变形节点对物理信息的利用有限,仅依靠前一个时刻的变形数据可能无法对后期变形进行有效预测——即发生如前所述的“概率漂移”现象;(2) M-DBN模型中,任意相邻时刻的转移概率相等,与工程实际情况不符合;实际工程中,位移时序曲线的类型很多,而M-DBN转移概率描述的只是其中的一种情况。
鉴于此,本研究提出改变上述M-DBN构建过程中的约束条件,对M-DBN模型进行约束增强优化:(1)保留上述约束1,即DBN模型结构中仅包含了初始网络结构和不同时刻的目标节点;(2)删除上述约束3,即认为任何相邻时刻转移网络的转移概率是不同的,该转移概率将基于围岩变形的动态样本库进行数据学习而获得,最大程度地保证转移概率与动态样本数据的契合度;(3)保留上述约束2的部分约束,即仍然认为t时刻的目标节点yt只与t1时刻的目标节点yt1相关,而与其他时刻的目标节点不相关;(4)增加优化增强约束4:保留任意时刻的目标节点yt与初始网络B0中目标节点y0的父节点的指向性连接,从而强化任意时刻的目标节点与围岩变形物理信息的联系。
基于约束4,式(10),(11)中,任一时刻t(不包括初始时刻)的π(yt)π(yrealt)π(ypredt)除包含最临近前一时刻节点yt1外,还包含初始网络B0中节点y0的父节点,如图5(d)所示。将上述约束增强优化后的M-DBN模型称为优化M-DBN(optimized M-DBN,OM-DBN)模型。相对于M-DBN,OM-DBN的模型结构和参数更加复杂一些,但是更加符合工程实际情况。
算例2:位移时序曲线重构方式为TNP 4,构造拱顶下沉的动态样本数据集,采用如图5(d)所示的OM-DBN方法构建转移网络结构。应用5折交叉试验的第2组训练样本集,采用式(7),(8)所示EM方法构建DBN的网络参数,得到拱顶下沉的OM-DBN模型如图7所示。
图7中,节点y1y2,……,y15对应不同时刻的目标节点y。从图7可以看出,OM-DBN模型由初始网络和不同时间节点的转移网络构成:(1) OM-DBN模型的初始SBN模型与图4(a)中所示的SBN是完全相同的;(2)从第1个时间节点以后,任一时刻的BN得到了极大的简化——除了目标节点及其父节点,其他的节点都已经删除——如上所述,这种处理是合理的。因此,任一时刻t的转移网络由该时刻目标节点yt、初始网络B0中节点y0的父节点和前一时刻的目标节点yt-1构成,转移概率由转移网络节点的样本数据通过EM方法计算获得,完全避免了约束2可能会导致的“概率漂移”问题。如果将转移网络中目标节点与网络中的指向性连线去掉(见图7中的点划线箭头),则OM-DBN模型结构退化为M-DBN模型结构——任一时刻t的转移网络退化为该时刻目标节点ytt1时刻目标节点yt1的连线,转移概率退化为P(yt|yt1)。但是,该转移概率是根据不同时刻目标节点的样本数据通过EM法计算得到的,与M-DBN中的转移概率不同。
图1所示,当时间节点确定以后,DBN表达的就是SBN在不同时刻的转移。按照实际工程中的逻辑,DBN中初始网络应该是隧道开挖支护完成初期、变形监测最初始的SBN,即DBN中转移网络的转移方向,与物理意义上时间进行的方向是一致的。但是,由表3可知,监测初期的围岩变形量普遍很小,无法体现不同影响因素作用下围岩变形的差异性,可能导致据以构建的BN模型不尽合理。因此,前面以变形监测最后时刻的极限位移值(包括外插值)构建的SBN(见图4)作为DBN的初始网络,相对物理时间进行逆向转移得到DBN模型,如图67所示。逆向转移方式构建的DBN模型与围岩变形过程的物理意义不一致,但是,从数学模型的角度来讲,DBN模型中的指向性连线并不代表真实的因果关系,其表征的是变量之间的一种概率计算路径;逆向转移方式同样表达的是动态系统在不同时刻的量化逻辑关系,能够表征动态系统的动态演化特征。
为验证逆向转移方式构建DBN模型的合理性,并探索正向转移方式构建DBN模型的可行性,基于“约束1”和图4所示的SBN,构建正向转移方式的DBN。“约束1”认为隧道围岩动态变形是一个随机平稳过程,即:DBN模型在初始时刻的网络结构,与其在最后时刻的网络结构应该保持一致。因此,可以利用图4的SBN作为正向转移DBN的初始网络,初始网络的结构和参数都与图4中所示的SBN相同;但是,其目标节点y0表示的是最小时间节点处的位移值,该节点的条件概率表与图4中的SBN不同,需要通过初始时刻样本数据进行学习而获得。随着时间的增加,SBN逐步转移形成DBN;不同时刻的目标节点取值也随之发生变化,形成转移网络。转移网络参数将通过正向转移样本进行学习获得,且与逆向转移DBN的参数不同。
算例3:假设时序曲线重构方式为TNP 4,采用正向转移方式构建拱顶下沉的动态样本库,基于数据集中所有的130组动态样本数据,采用图5(d)所示的OM-DBN方式构建转移网络结构;应用5折交叉试验的第2组训练样本集,采用式(7),(8)所示EM方法计算OM-DBN的网络参数,得到DBN模型如图8所示。
图7相比,图8中正向转移的OM-DBN模型与逆向转移OM-DBN模型的网络结构(包括初始网络结构和转移网络结构)是一样的,初始网络中因素节点(x1  x14)的条件概率是相同的。但由于样本中目标节点y0的取值不同,导致该节点的条件概率也不同。因为转移方向的不同,不同时间节点处目标节点y=(y1y2  yT)的样本取值不同,所以转移网络的条件概率也都不同。不同网络转移方向构建的DBN模型的合理性和有效性,将在下文中进行评价。
由上述分析可知,根据采集的样本数据构建DBN模型,时序曲线重构方式(TNP 1,TNP 2,TNP 3和TNP 4)、转移网络构建方法(M-DBN,OM-DBN)和网络转移方向等不同的模型构建方法,决定了不同的DBN模型结构和参数。针对拱顶下沉和水平收敛的动态预测,采用多因素完全方案设计对比试验,对每一种模型构建方法进行考察,如表6所示。由表6可知,针对拱顶下沉和水平收敛的动态预测各构建了16个DBN模型,共32个。其中,模型10对应于图6所示的DBN模型,模型16对应于图7所示的DBN模型,模型15对应于图8所示的DBN模型。应用5折交叉验证试验的方法,对该32个DBN模型的优劣进行评价,筛选最优的DBN模型。
分类预测模型的评价指标较多,除了最常用的总体精度(overall accuracy,ACC)以外,还有准确率(precision,P)、召回率(recall rate,R)和F1值等更为全面的评价指标。ACC表示测试样本中测试结果正确的样本数量所占的比例;准确率P也称为查准率,表示正确预测为某一类的样本数量,占所有预测为该类的样本数量的比例,代表了ML算法对某一类样本进行准确预测的能力;召回率R也称为查全率,表示正确预测为某一类的样本数量,占真实为该类的样本数量的比例,代表了ML算法将某类样本正确识别出来的能力。精确度P和召回率R相互制约,通常P较高时,R就较低;反之亦然。因此,引入F1值来综合反映模型的预测能力;F1值越高,表明模型的预测能力较好[1146-47]。本研究将综合应用ACCPRF1对DBN模型进行评价。
应用上述分类评价指标对DBN模型进行评价,实际上是针对位移时序曲线上每一个时间节点处的位移值预测效果进行评价。尽管未使用回归模型中常用的方根误差、平均绝对误差或相关系数等指标进行评价,上述多个分类评价指标在每一个时间节点处的应用,从多个维度反映了位移时序曲线中预测位移值与实测位移值的拟合程度,实现了时序曲线拟合质量的评估。
表6中DBN模型的ACC表7所示。此处ACC指每个位移时序曲线上时间节点处ACC的平均值。由表7可知,对于拱顶下沉和水平收敛,将变形数据补充至70 d再划分时间节点(TNP 4)构建模型的ACC均为最高,TNP3构建的模型预测正确率最低。因此,TPN4是最合理的时间节点划分方式,后续的分析都将基于TNP4的时序曲线重构方法展开。
对于拱顶下沉,OM-DBN的效果优于M-DBN;逆向转移的ACC更高。对于水平收敛,与拱顶下沉的结果有所不同:相对于OM-DBN模型,M-DBN模型的ACC更高;正向转移的ACC更高。可见,本研究针对拱顶下沉和水平收敛发生机制的不同而分别进行DBN构建,是合理的。但是,上述评价只是模型预测能力的整体表达,接下来将结合PRF1值等指标对DBN模型的预测能力做更深层次的分析。
选取TNP4构建的8个DBN模型,对其在15个时间节点的预测能力做进一步的深入分析,如图9所示。图中,图例中的F表示转移网络为正向转移(forward),B表示逆向转移(backward)。由图9所示,拱顶下沉DBN模型与水平收敛DBN模型的预测性能是非常相似的:OM-DBN模型的P(R)值始终大于M-DBN模型的;对于M-DBN模型,逆向转移网络的P(R)值始终大于正向转移网络的;对于OM-DBN模型,正向转移网络的P(R)值与逆向转移网络的相差不大,且随着观测时间的增长而此消彼长。由于任一样本初始观测时间点的位移值都是L1(见表4),所以8个模型在初始观测时间点的P(R)值都是1.0;随着观测时间的增长,同一时间节点处不同样本的取值差异逐渐变大,预测的难度增大,所以8个模型的P(R)值都是随着观测时间的增大而逐步减小的。
对于M-DBN模型,P值和R值的变化规律是相似的,都是随着观测时间增加持续减小。对于OM-DBN模型,P值和R值的变化规律有所不同:在第40 d以前,二者的变化规律相似;在第40 d以后,P值有一定的回升后保持稳定,R值则是持续下降后保持稳定。可见,由于OM-DBN模型在各个时间节点都可以有效利用隧道围岩变形的物理信息,其预测性能更好。
对于拱顶下沉和水平收敛的DBN模型,其P值和R值的变化规律总体是相似的,但是仍然存在着一定的区别,如对于拱顶下沉的OM-DBN模型,在观测时间的后半段,P值和R值的差异保持不变的;而对于水平收敛的OM-DBN模型,在观测时间的后半段,P值和R值的差异是逐渐减小的。因此,需要通过F1值来综合评价DBN模型的预测性能。
拱顶下沉和水平收敛DBN模型的F1值和ACC值随观测时间的变化规律,如图10所示。由图10可知,任何1个DBN模型的F1值和ACC值具有较好的一致性,在任何1个时间节点,F1值较大的时候,ACC值也较大。可见,应用F1值综合评价DBN模型的预测性能是可行的。拱顶下沉DBN模型与水平收敛DBN模型的F1值和ACC值存在着一定的差异,对于OM-DBN模型,在观测时间的最后阶段,拱顶下沉DBN模型的F1值和ACC值之间的差异保持稳定,而水平收敛DBN模型的F1值和ACC值之间的差异逐渐增大。但总体而言,拱顶下沉DBN模型与水平收敛DBN模型的F1值和ACC值变化规律是非常相似的。
M-DBN模型转移网络中的目标节点只与前(后)一个时间节点处的目标节点有连接,与地质环境条件、施工方法条件等影响因素节点之间没有直接的连接;随着观测时间的变化,转移网络中目标节点获取影响因素节点的有效信息越来越少。在物理信息的减少和目标值差异性增加的双重影响下,M-DBN模型的F1值和ACC值持续地降低,且降低的幅度很大。OM-DBN模型任一时间节点转移网络中的目标节点,与前(后)一个时间节点的目标节点和影响因素节点等之间都有直接的联系,所以观测时间的增长不会明显降低F1值和ACC值——降低到第25 d以后,便不再明显降低。因此,OM-DBN模型的F1值和ACC值都远高于M-DBN模型。
DBN模型中,第1 d处的目标节点取值都是L1,无法体现不同影响因素对目标节点的影响,即第1 d目标节点与影响因素节点的直接连接无法获取足够多的有效物理信息。正向转移M-DBN模型中,除了第1 d处的目标节点外,其他目标节点未与影响因素之间建立直接联系,因此无法获取足够多的有效物理信息,其F1值和ACC值远低于逆向转移M-DBN模型。
OM-DBN模型中,无论是通过正向转移还是逆向转移来构建转移网络,每一个观测时间节点的目标节点都与影响因素节点建立了直接的连接,能够获取充足的有效物理信息。因此,OM-DBN模型逆向转移的F1值和ACC值均与正向转移的相差不大。
M-DBN模型的转移概率是一个定值,同一种转移方向时,不同时间节点的转移概率是相等的;不同转移方向条件下转移概率的差异也是固定的。因此,正向转移模型与逆向转移模型的差异是相对稳定的,不会出现交叉的现象。
OM-DBN模型中任何时间节点处的转移概率是根据样本数据进行学习得到的,同一种转移方向条件下不同时间节点处的转移网络,会根据样本数据的特征而发生变化,转移概率会根据样本数据分布发生不可预知的变化。因此,正向转移模型与逆向转移模型的差异是不稳定的,交叉的现象较为普遍,且差异也不是很大。
综上所述,对于围岩拱顶下沉和水平收敛的动态预测,DBN模型采用TNP4的时间节点划分方式是最合理的;OM-DBN模型显著优于M-DBN模型;OM-DBN模型中,正向转移与逆向转移的方式差异不大。接下来,基于优选的OM-DBN模型(时序曲线重构方式为TNP4,转移网络的转移方向为正向转移——见表6中的模型15,图8)进行隧道工程案例的应用分析与讨论。
化马隧道为一双线大断面隧道,起止里程为DK301+282~DK313+862,全长1 258 m,最大埋深约1 300 m。DK302+177断面设计净高为13.65 m,最大净宽为15.48 m,如图11所示。以IV类围岩为主,岩性为灰岩夹板岩,岩体破碎,地下水发育,较富水。采用曲墙带仰拱复合式衬砌,初期支护采用锚喷支护。基于表1所示的影响因素离散化方法,根据工程勘察资料和工程设计资料,得到各个影响因素的取值如表8所示。从隧道施作开始,实施了持续70 d的现场变形监测,如图12所示。由图12可知,由于监测设备或环境扰动等原因,有20 d的监测数据丢失。
分别采用本研究所构建的DBN模型和支持向量机(support vector machine,SVM)2种方法对该隧道的位移时序曲线进行预测:(1)将表8中的14个影响因素取值输入图8的DBN模型,通过式(10)进行推理计算,得到15个时间节点的拱顶下沉值,以实现基于DBN的位移时序曲线预测,如图12所示;(2)采集前26 d的现场监测位移数据构造出21个样本进行SVM模型训练(SVM模型训练的超参数取常规默认值),应用得到的SVM模型对第26 d以后的位移值进行预测,如图12所示。
图12中,DBN模型预测结果为离散型数值的位移等级。现场监测位移和SVM模型预测结果本来是连续型的数值,为了便于与DBN模型的预测结果进行对比分析,也将其按照表4所示的规则转化成离散型的位移等级。如前所述,通过离散型变量对隧道围岩变形量进行表达,是具有工程意义的。
图12可知,SVM模型的预测结果与现场监测位移的吻合度非常好,每一个预测位移值都与实测值保持了一致,其ACCPRF1值4个评价指标的值都为1.00。但是,如图12所示,由于位移监测设备或环境扰动等原因,现场监测位移数据经常会出现数据丢失的情况;此时,SVM等“基于现场监测位移的数据驱动预测方法”将失效。
DBN模型的预测结果与现场监测位移值保持了很好的一致性,仅有第30 d的预测位移等级与实测位移等级不同(预测值高1个等级)。DBN模型的ACCPRF1值4个评价指标的值分别为0.977,0.900,1.000和0.950。此处评价指标的计算与上一节中的计算不同,是把每一个时间节点的位移值作为一个测试样本进行统计计算的。可见,DBN模型的预测效果也是非常好的。同时,如图12所示,现场监测位移数据的部分缺失,并没有对DBN模型的预测结果产生太大的影响,DBN模型仍然可以实现完整位移时序曲线的预测,直到第70 d。
由此可见:(1) SVM模型需要根据大量的早期监测位移数据才可以实现后续位移的预测;DBN模型有效利用隧道围岩变形影响因素数据等物理信息,即可在隧道施工前实现全过程位移时序曲线的预测,显著克服了“基于现场监测位移的数据驱动预测方法”的滞后性,且预测准确率非常高。(2)当现场监测位移存在数据缺失时,SVM模型将无法对后续位移数据进行预测;DBN模型在不输入现场监测位移数据的情况下,实现位移时序曲线的预测,具有非常好的工程适用性。(3)隧道开始施作获取监测数据后,将监测位移值输入DBN模型,DBN模型将能够根据现场监测位移数据提供的数学信息,对位移时序曲线进行及时的更新,这个工作将在下节案例二中进行展示。
节6.1的案例分析主要是依据式(10)进行的:在输入围岩变形影响因素取值的情况下评价输出的围岩变形计算值与样本值的差异或契合度。上述评价分析展示了DBN模型在完全不依赖现场监测数据的情况下,利用围岩变形物理信息对位移值进行动态预测的过程,这是本研究构建DBN模型相对于传统“基于现场监测位移的数据驱动预测方法”的最大优势之一。接下来,展示如何应用式(11),有效利用现场监测数据进行DBN模型的更新。文笔山隧道为分离式隧道,其左线的起止桩号为ZK149+885~ZK152+630,全长2 745 m,最大埋深约为254 m,如图13所示。ZK152+120断面处埋深为85 m,围岩等级为IV级,围岩岩性为玄武岩夹凝灰岩,地下水不发育,岩体富水性弱,岩体自稳能力较差,侧壁稳定性差。各个影响因素的取值如表8所示。将表8所示的影响因素取值输入图8所示的DBN模型,通过式(10)进行推理计算,得到15个时间节点的拱顶下沉值如图14所示。
图14可知,此隧道的围岩状态不是很好,围岩级别较高、围岩的岩性水稳性差、围岩结构破碎,隧道的埋深较大、跨度也较大;但是,好在初始地应力不高、主要结构面的倾角较小、地层含水状态较好。重要的是,隧道的施工条件因素(开挖工法、超前支护工法、支护强度、支护时机、地下水控制、围岩扰动程度等)是通过基于本研究构建的SBN和DBN进行反向推理计算获取的能够保证隧道围岩稳定的取值,保证了施工方法的合理性。因此,由图14可知,隧道施作后15个时间节点的拱顶下沉值都是处于L1等级(0~50 mm),表现出了较好的变形稳定性。同时,由图14可知,虽然每个时间节点的位移值都处于L1等级,表示每一时刻yt的8个取值状态中,L1等级的取值概率最大;但是,随着时间的推移,yt处于L1等级的取值概率是逐渐变小的。说明随着时间的推移,该隧道的拱顶下沉有可能会逐步增长到L2等级。
隧道开挖支护后,第一时间在围岩初期支护结构上布设监测点,获取围岩的拱顶下沉和水平收敛等位移值。将监测到的位移值按照相应的时间节点输入图8所示的DBN模型中,通过式(11)进行推理计算,对后续的位移值进行更新预测,如图14所示。第1 d和第5 d的拱顶下沉监测值与预测值是一致的,将第1 d和第5 dytL1等级取值概率设置为100%,完成对模型的更新;模型更新后,后续的变形预测值没有发生变化。第10 d的拱顶下沉监测值(L2等级)与预测值不同,将第10 d的y3取值改为L2等级(将其取值概率设置为100%),完成模型的更新。模型更新后发现,后续的变形预测值发生了较大的变化——随着时间的推移,变形预测值将逐步增加到L3等级。将后续各个时间节点的现场监测位移值持续代入DBN模型,通过式(11)进行推理计算,将能够完成隧道围岩拱顶下沉预测值的持续更新。DBN模型充分挖掘了隧道围岩变形影响因素的物理信息,不仅能在完全不依赖现场监测数据的情况下进行位移时序曲线的预测,还可提供充分的数据信息以保障隧道施工初期位移时序曲线动态更新的有效性。
由上述案例分析可知,围岩变形的DBN预测模型不仅可以在隧道工程的设计阶段和施工阶段前期依据式(10)完成围岩变形的预测,为隧道工程选址、二次衬砌施作时机选择和监测预警值设置等提供依据,也可以在隧道施工的过程中,根据现场位移监测值和式(11)对围岩变形预测进行持续的更新,为隧道围岩稳定状态的判断、二次衬砌施作时机的调整、预警状态的确定等提供依据。除此之外,基于构建完成的围岩变形DBN预测模型,还可以进行隧道施工参数的优化、施工过程中的工程病害诊断等工作,将在其他的研究成果中进行展示。
针对隧道围岩变形的动态预测问题,通过文献资料采集,获取了130组包含隧道围岩变形影响因素物理信息和位移时序曲线数学信息的案例数据;通过离散化处理和位移时序曲线重构,构建了融合变形影响因素数据、极限位移数据和位移时序曲线数据等多源信息的样本库。基于BN理论框架,综合利用样本数据的物理信息和数学信息,推导了融合物理先验信息和样本数据的PIML建模方法,建立了围岩变形动态预测的DBN模型,得到以下研究成果:
(1)提出了基于ISM优化K2-score算法的BN模型构建方法,建立了符合工程逻辑的极限变形预测SBN模型。
(2)提出了符合围岩变形动态过程物理特征的平稳随机过程约束和Markov过程约束等物理先验信息,融合多源样本数据驱动,构建了围岩变形动态预测M-DBN模型。
(3)融合约束增强优化先验信息,实现任一时刻变形量与物理信息的深度融合,建立了优化M-DBN(OM-DBN)模型。
(4)综合考虑位移时序曲线的重构方法、DBN模型的构建方法和网络转移方向,分别针对拱顶下沉和水平收敛构建了共32个DBN模型。通过5折交叉试验进行评价,得出:① TNP4是最优的时序曲线重构方法;② M-DBN模型的预测能力随时间增加而快速降低,网络转移方向对其影响很大;③ OM-DBN模型的预测能力不随时间变化而明显降低,不受网络转移方向的显著影响,且远高于M-DBN模型。
(5)采用TNP4时序曲线重构方法+OM-DBN建模技术+正向网络转移方向等关键技术构建的OM-DBN模型,是最优的围岩变形动态预测DBN模型。
案例分析结果表明,在隧道施工前或初期,输入隧道工程建设资料提供的因素取值,OM-DBN模型能够实现位移时序曲线的有效预测,弥补了传统“基于现场监测位移数据驱动”预测方法的局限性和滞后性;在施工过程中,增加输入监测位移数据,OM-DBN模型能够进行自我更新,实现位移时序曲线的动态更新。
  • 重庆市教育委员会科学技术研究项目(KJQN202301530)
  • 国家自然科学基金青年基金资助项目(52304125)
  • 重庆市自然科学基金面上项目(CSTB2023NSCQ-MSX0828)
参考文献 引证文献
排序方式:
[1]
郑艾辰,黄锋,林志,. 2008年至2016年我国隧道工程施工安全事故统计与分析[J]. 施工技术201746(增1):833–836.
(ZHENG AichenHUANG FengLIN Zhiet al. The statistics and analysis of safety accident during tunnel engineering construction in China from 2008 to 2016[J]. Construction Technology201746(Supp.1):833–836.(in Chinese))
[2]
吕擎峰,李钰,牛荣,. 基于深度学习的特殊岩土隧道围岩变形预测研究[J]. 应用基础与工程科学学报202331(6):1 590–1 600.
( QingfengLI YuNIU Ronget al. Research on deformation prediction of surrounding rock in special geotechnical tunnels based on deep learning[J]. Journal of Fundamental and Engineering Science202331(6):1 590–1 600.(in Chinese))
[3]
孙振宇,张顶立,侯艳娟,. 基于现场实测数据统计的隧道围岩全过程变形规律及稳定性判据确定[J]. 岩土工程学报202143(7):12–20.
(SUN ZhenyuZHANG DingliHOU Yanjuanet al. Whole-process deformation laws and determination of stability criterion of surrounding rock of tunnels based on statistics of field measured data[J]. Chinese Journal of Geotechnical Engineering202143(7):12–20.(in Chinese))
[4]
王述红,董福瑞. 基于变形预测和参数反演的山岭隧道围岩稳定性分析[J]. 岩土工程学报202345(5):1 024–1 035.
(WANG ShuhongDONG Furui. Stability analysis of surrounding rock of mountain tunnels based on deformation prediction and parameter inversion[J]. Chinese Journal of Geotechnical Engineering202345(5):1 024–1 035.(in Chinese))
[5]
WANG YLIU YWANG Z Fet al. Investigation on progressive failure process of tunnel lining induced by creep effect of surrounding rock:A case study[J]. Engineering Failure Analysis2023144:106946.
[6]
SHI SZHAO RLI Set al. Intelligent prediction of surrounding rock deformation of shallow buried highway tunnel and its engineering application[J]. Tunnelling and Underground Space Technology201990:1–11.
[7]
AYDAN OAKAGI TKAWAMOTO T. The squeezing potential of rocks around tunnels:Theory and prediction[J]. Rock Mechanics and Rock Engineering199326(2):137–163.
[8]
BIAN KLIU JLIU Zet al. Mechanisms of large deformation in soft rock tunnels a case study of Huangjiazhai Tunnel[J]. Bulletin of Engineering Geology and the Environment201978(1):431–444.
[9]
王明年,王志龙,张霄,. 深埋隧道围岩形变压力计算方法研究[J]. 岩土工程学报202042(1):81–90.
(WANG MingnianWANG ZhilongZHANG Xiaoet al. Method for calculating deformation pressure of surrounding rock of deep-buried tunnels[J]. Chinese Journal of Geotechnical Engineering202042(1):81–90.(in Chinese))
[10]
马鑫民,陈攀,陈晨,. 基于机器学习的煤巷围岩稳定性预测与应用[J]. 矿业科学学报20238(2):156–165.
(MA XinminCHEN PanCHEN Chenet al. Prediction of surrounding rock stability of coal roadway based on machine learning and its application[J]. Journal of Mining Science and Technology20238(2):156–165.(in Chinese))
[11]
WANG LPAN QWANG S. Data-driven predictions of shield attitudes using Bayesian machine learning[J]. Computers and Geotechnics2024166:106002.
[12]
FENG X DJIMENEZ R. Predicting tunnel squeezing with incomplete data using Bayesian networks[J]. Engineering Geology2015195:214–224.
[13]
WU X GJIANG ZZHANG L Met al. Dynamic risk analysis for adjacent buildings in tunneling environments:a Bayesian network based approach[J]. Stochastic Environmental Research and Risk Assessment201529(5):1 447–1 461.
[14]
ZHOU JZHU S LQIU Y Get al. Predicting tunnel squeezing using support vector machine optimized by whale optimization algorithm[J]. Acta Geotechnica202217(4):1 343–1 366.
[15]
梁文添,陈劲慧. 新奥法在三连拱特大断面隧道施工中的应用[J]. 岩石力学与工程学报201736(11):2 763–2 770.
(LIANG WentianCHEN Jinhui. Application of new Austrian tunnelling method to a large span trinocular cross section tunnel[J]. Chinese Journal of Rock Mechanics and Engineering201736(11):2 763–2 770.(in Chinese))
[16]
FENG T GWANG C RZHANG Jet al. An improved artificial bee colony-random forest(IABC-RF) model for predicting the tunnel deformation due to an adjacent foundation pit excavation[J]. Underground Space20227(4):514–527.
[17]
CHEN SZHU YHUANG L. Application of chaotic time series analysis to the prediction of tunnel surrounding rock displacement[J]. Modern Tunnelling Technology201552(3):75–81.
[18]
LI Z HMA E LLAI J Xet al. Tunnel deformation prediction during construction:An explainable hybrid model considering temporal and static factors[J]. Computers and Structures2024294:107276.
[19]
BAI C HXUE Y GQIU D Het al. Analysis of factors affecting the deformation of soft rock tunnels by data envelopment analysis and a risk assessment model[J]. Tunnelling and Underground Space Technology2021116:104111.
[20]
张振,苏欣荣,袁新. 基于物理信息神经网络的气膜冷却湍流模型反演学习[J]. 动力工程学报202444(9):1 459–1 465.
(ZHANG ZhenSU XinrongYUAN Xin. PINN-based inversion learning of turbulence model for film cooling[J]. Journal of Chinese Society of Power Engineering202444(9):1 459–1 465.(in Chinese))
[21]
卢鑫月,许成顺,侯本伟,. 基于动态贝叶斯网络的地铁隧道施工风险评估[J]. 岩土工程学报202244(3):492–501.
(LU XinyueXU ChengshunHOU Benweiet al. Risk assessment of metro construction based on dynamic Bayesian network[J]. Chinese Journal of Geotechnical Engineering202244(3):492–501.(in Chinese))
[22]
CHING W K,程伟琪,黄曦敏,. 马尔可夫链:模型、算法与应用[M]. 北京:清华大学出版社,2015:13–27.
(CHING W KCHENG WeiqiHUANG Ximinet al. Markov Chains:models,algorithms and applications second edition[M]. Beijing:Tsinghua University Press,2015:13–27.(in Chinese))
[23]
SMITH J QANDERSON P E. Conditional independence and chain event graphs[J]. Artificial Intelligence2008172(1):42–68.
[24]
杨悦,陈孝国,高霞,. 基于神经网络–蒙特卡罗法的隧道初衬可靠度研究[J]. 齐齐哈尔大学学报:自然科学版201733(1):5–9.
(YANG YueCHEN XiaoguoGAO Xiaet al. Study on reliability of tunneling initial lining based on radial basis function neural network-Monte Carlo method[J]. Journal of Qiqihar University:Natural Science201733(1):5–9.(in Chinese))
[25]
张广泽,邓建辉,王栋,. 隧道围岩构造软岩大变形发生机理及分级方法[J]. 工程科学与技术202153(1):1–12.
(ZHANG GuangzeDENG JianhuiWANG Donget al. Mechanism and classification of tectonic-induced large deformation of soft rock tunnels[J]. Engineering Science and Technology202153(1):1–12.(in Chinese))
[26]
中华人民共和国国家标准编写组. GB/T 10003—2016铁路隧道设计规范[S]. 北京:中国铁道出版社,2016.
(The National Standards Compilation Group of People's Republic of China. GB/T 10003—2016 Code for design of railway tunnel[S]. Beijing:China Railway Publishing House,2016.(in Chinese))
[27]
中华人民共和国行业标准编写组. JTG 3370.1—2018公路隧道设计规范[S]. 北京:人民交通出版社,2018.
(The Professional Standards Compilation Group of People's Republic of China. JTG 3370.1—2018 Specification for design of highway tunnel[S]. Beijing:China Communications Press,2018.(in Chinese))
[28]
COOPER GHERSKOVITS E. A Bayesian method for the induction of probabilistic networks from data[J]. Machine Learning19929(4):309–347.
[29]
朱永全,宋玉香. 隧道工程[M]. 北京:中国铁道出版社,2018:21–30.
(ZHU YongquanSONG Yuxiang. Tunnel engineering[M]. Beijing:China Railway Publishing House,2018:21–30.(in Chinese))
[30]
孙景来,刘保国,储昭飞,. 隧道坍塌事故类型划分及其主要特征[J]. 中国铁道科学201839(6):44–51.
(SUN JinglaiLIU BaoguoCHU Zhaofeiet al. Type classification and main characteristics of tunnel collapses[J]. China Railway Science201839(6):44–51.(in Chinese))
[31]
陈子全,何川,吴迪,. 高地应力层状软岩隧道大变形预测分级研究[J]. 西南交通大学学报201853(6):1 237–1 244.
(CHEN ZiquanHE ChuanWU Diet al. Study of large deformation classification criterion for layered soft rock tunnels under high geostress[J]. Journal of Southwest Jiaotong University201853(6):1 237–1 244.(in Chinese))
[32]
任德惠,张平. 不同倾角结构面对巷道稳定性的影响[J]. 煤炭学报198813(3):13–19.
(REN DehuiZHANG Ping. Influence of structural planes with different dip angles on roadway stability[J]. Journal of China Coal Society198813(3):13–19.(in Chinese))
[33]
吴秋军. 铁路隧道围岩变形机理与分级方法研究[博士学位论文][D]. 成都:西南交通大学,2017.
(WU Qiujun. Deformation mechanism and classification methods for surrounding rock of railway tunnels[Ph. D. Thesis][D]. Chengdu:Southwest Jiaotong University,2017.(in Chinese))
[34]
王冠勇. 分析影响隧道围岩稳定性因素[J]. 交通世界(建养・机械)2012,(13):234–235.
(WANG Guanyong. Analysis of factors affecting the stability of tunnel surrounding rock[J]. Traffic World (Construction and Maintenance・Machinery)2012,(13):234–235.(in Chinese))
[35]
吉小明. 隧道开挖的围岩损伤扰动带分析[J]. 岩石力学与工程学报200524(10):1 697–1 702.
(JI Xiaoming. Study on mechanical and hydraulic behavior of tunnel surrounding rock masses in excavation-disturbed zone[J]. Chinese Journal of Rock Mechanics and Engineering200524(10):1 697–1 702.(in Chinese))
[36]
来弘鹏,林永贵,谢永利,. 支护时机对软弱围岩公路隧道力学特征影响的试验研究[J]. 岩土工程学报200931(3):390–395.
(LAI HongpengLIN YongguiXIE Yongliet al. Influence of supporting opportunity on stress characteristics of soft-weak surrounding rocks in highway tunnels[J]. Chinese Journal of Geotechnical Engineering200931(3):390–395.(in Chinese))
[37]
ŠPAČKOVÁ OSTRAUB D. Dynamic Bayesian network for probabilistic modeling of tunnel excavation processes[J]. Computer-Aided Civil and Infrastructure Engineering201328(1):1–21.
[38]
李海斌,翟秋柱,张优,. PSO-BP神经网络在隧道围岩变形预测中的应用[J]. 路基工程2017,(5):164–169.
(LI HaibinZHAI QiuzhuZHANG Youet al. Application of PSO-BP neural network in prediction of tunnel rock deformation[J]. Subgrade Engineering2017,(5):164–169.(in Chinese))
[39]
王树栋,刘开云. 长大隧道软弱围岩施工大变形智能预测方法[J]. 中国铁道科学200829(2):82–87.
(WANG ShudongLIU Kaiyun. Intelligence method for large deformation prediction of weak surrounding rock construction in long and large tunnel[J]. China Railway Science200829(2):82–87.(in Chinese))
[40]
解志刚. 隧道围岩变形预测及其在工程中的应用[硕士学位论文][D]. 兰州:兰州交通大学,2012.
(XIE Zhigang. Tunnel surrounding rock deformation and prediction and its application in engineering[M. S. Thesis][D]. Lanzhou:Lanzhou Jiaotong University,2012.(in Chinese))
[41]
NIU Z LWANG Y QFAN S Y. Research on the deformation control of surrounding rock about large-section tunnel in strong-medium weathered slate[J]. Frontiers in Earth Science202310:1–15.
[42]
汪学清,刘爽,李秋燕,. 基于K折交叉验证的SVM隧道围岩分级判别[J]. 矿冶工程202141(6):126–128.
(WANG XueqingLIU ShuangLI Qiuyanet al. Classification and discrimination of surrounding rock of tunnel based on SVM of K-Fold cross validation[J]. Mining and Metallurgical Engineering202141(6):126–128.(in Chinese))
[43]
毛奕喆,龚国芳,周星海,. 基于马尔可夫过程和深度神经网络的TBM围岩识别[J]. 浙江大学学报:工学版202155(3):458–466.
(MAO YizheGONG GuofangZHOU Xinghaiet al. Identification of TBM surrounding rock based on Markov process and deep neural network[J]. Journal of Zhejiang University:Engineering Science202155(3):458–466.(in Chinese))
[44]
GHAHRAMANI Z. An introduction to hidden Markov models and Bayesian Networks[J]. International Journal of Pattern Recognition and Artificial Intelligence200115(1):9–42.
[45]
刘次华. 随机过程及其应用[M]. 北京:清华大学出版社,2004:1–2.
(LIU Cihua. Stochastic processes and their applications[M]. Beijing:Tsinghua University Press,2004:1–2.(in Chinese))
[46]
YE X WZHANG X LZHANG H Qet al. Prediction of lining upward movement during shield tunneling using machine learning algorithms and field monitoring data[J]. Transportation Geotechnics202341:101002.
[47]
杨背背,殷坤龙,杜娟. 基于时间序列与长短时记忆网络的滑坡位移动态预测模型[J]. 岩石力学与工程学报201837(10):2 010–2 018.
(YANG BeibeiYIN KunlongDU Juan. A model for predicting landslide displacement based on time series and long and short term memory neural network[J]. Chinese Journal of Rock Mechanics and Engineering201837(10):2 010–2 018.(in Chinese))
2026年第45卷第2期
PDF下载
2
1
引用本文
BibTeX
文章信息
doi: 10.3724/1000-6915.jrme.2025.0477
  • 接收时间:2025-07-03
  • 首发时间:2026-06-18
  • 出版时间:2026-02-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-07-03
  • 修回日期:2025-09-27
基金
Science and Technology Research Program of Chongqing Municipal Education Commission(KJQN202301530)
重庆市教育委员会科学技术研究项目(KJQN202301530)
National Natural Science Foundation for Young Scholars of China(52304125)
国家自然科学基金青年基金资助项目(52304125)
Chongqing Natural Science Foundation Project(CSTB2023NSCQ-MSX0828)
重庆市自然科学基金面上项目(CSTB2023NSCQ-MSX0828)
作者信息
    1.重庆科技大学 安全科学与工程学院,重庆 400041
    2.中国科学院武汉岩土力学研究所 岩土力学与工程安全全国重点实验室,湖北 武汉 430071
    3.重庆大学 资源与安全学院,重庆 400044

通讯作者:

* 张超(1978–),现任研究员,主要从事岩土工程灾害防控方面的研究工作。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/yslxygcxb/CN/10.3724/1000-6915.jrme.2025.0477
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏