Tunnelling and Underground Space Technology最新文献_第6页

Advance prediction of rock mass classification in tunneling using improved D-S fusion and hybrid machine learning 基于改进D-S融合和混合机器学习的隧道围岩分类预测

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-19 DOI: 10.1016/j.tust.2026.107453

Shikuo Chen , Yifan Hou , Rui Wang , Xiaoyan Zhao

Accurate prediction of rock mass classification is imperative for optimizing safety and cost-efficiency in underground tunnel engineering. Despite its critical importance, conventional single-classification models often exhibit limitations in robustness and accuracy, hindering reliable risk assessment and design optimization. To overcome these persistent challenges, this study proposes a multi-model fusion framework grounded in D-S evidence theory, significantly enhancing classification reliability. Furthermore, an LSTM-based model is developed for ahead-of-face rock class prediction, leveraging geological data from excavated sections. Utilizing 325 field case datasets, unsupervised learning and SMOTE preprocessing were applied, with t-SNE visualization confirming markedly enhanced feature separability. Based on seven key geological indicators, five predictive models spanning classical rock mass rating systems and data-driven machine learning methods were established. These outputs were fused via a D-S evidence theory framework, significantly enhancing classification robustness. Furthermore, hyperparameters of the BP and RF models were optimized via global search algorithms to enhance base classifiers performance. Building upon their test-set metrics, we propose a refinement of the Basic Probability Assignment (BPA) function by integrating precision and accuracy. This modified BPA is adopted as the fusion index with an improved D-S evidence theory framework, establishing a robust rock mass classification model. Validated across three tunnels, the improved D-S model achieved 89.13% accuracy—outperforming all base classifiers. The integrated LSTM predictor further demonstrated robustness to temporal parameter variations. This integrated approach effectively mitigates single-model instability, significantly boosting classification accuracy and robustness. Crucially, its short-range ahead-of-face predictive capability enables proactive support design, enhancing tunnel construction safety.

准确的岩体分类预测对于优化地下隧道工程的安全性和成本效益至关重要。尽管其至关重要，但传统的单一分类模型往往在鲁棒性和准确性方面存在局限性，阻碍了可靠的风险评估和设计优化。为了克服这些持续存在的挑战，本研究提出了一个基于D-S证据理论的多模型融合框架，显著提高了分类的可靠性。此外，利用开挖断面的地质数据，开发了基于lstm的工作面前岩石分类预测模型。利用325个现场案例数据集，应用无监督学习和SMOTE预处理，t-SNE可视化证实显著增强了特征可分离性。基于7个关键地质指标，建立了跨越经典岩体评级系统和数据驱动机器学习方法的5个预测模型。这些输出通过D-S证据理论框架融合，显著增强了分类稳健性。此外，通过全局搜索算法优化BP和RF模型的超参数，以提高基分类器的性能。在他们的测试集指标的基础上，我们提出了一种基本概率分配（BPA）函数的改进，通过整合精度和准确性。采用改进的双酚a作为融合指标，结合改进的D-S证据理论框架，建立了稳健的岩体分类模型。经过三个隧道的验证，改进的D-S模型达到了89.13%的准确率，优于所有基础分类器。综合LSTM预测器进一步证明了对时间参数变化的鲁棒性。这种综合方法有效地减轻了单一模型的不稳定性，显著提高了分类精度和鲁棒性。最重要的是，它的短距离前方预测能力可以实现主动支持设计，提高隧道施工安全性。

{"title":"Advance prediction of rock mass classification in tunneling using improved D-S fusion and hybrid machine learning","authors":"Shikuo Chen , Yifan Hou , Rui Wang , Xiaoyan Zhao","doi":"10.1016/j.tust.2026.107453","DOIUrl":"10.1016/j.tust.2026.107453","url":null,"abstract":"<div><div>Accurate prediction of rock mass classification is imperative for optimizing safety and cost-efficiency in underground tunnel engineering. Despite its critical importance, conventional single-classification models often exhibit limitations in robustness and accuracy, hindering reliable risk assessment and design optimization. To overcome these persistent challenges, this study proposes a multi-model fusion framework grounded in D-S evidence theory, significantly enhancing classification reliability. Furthermore, an LSTM-based model is developed for ahead-of-face rock class prediction, leveraging geological data from excavated sections. Utilizing 325 field case datasets, unsupervised learning and SMOTE preprocessing were applied, with t-SNE visualization confirming markedly enhanced feature separability. Based on seven key geological indicators, five predictive models spanning classical rock mass rating systems and data-driven machine learning methods were established. These outputs were fused via a D-S evidence theory framework, significantly enhancing classification robustness. Furthermore, hyperparameters of the BP and RF models were optimized via global search algorithms to enhance base classifiers performance. Building upon their test-set metrics, we propose a refinement of the Basic Probability Assignment (BPA) function by integrating precision and accuracy. This modified BPA is adopted as the fusion index with an improved D-S evidence theory framework, establishing a robust rock mass classification model. Validated across three tunnels, the improved D-S model achieved 89.13% accuracy—outperforming all base classifiers. The integrated LSTM predictor further demonstrated robustness to temporal parameter variations. This integrated approach effectively mitigates single-model instability, significantly boosting classification accuracy and robustness. Crucially, its short-range ahead-of-face predictive capability enables proactive support design, enhancing tunnel construction safety.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107453"},"PeriodicalIF":7.4,"publicationDate":"2026-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146001005","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A two-stage robust optimization model for underground container logistics system investment under carbon tax and subsidies policies 碳税和补贴政策下地下集装箱物流系统投资的两阶段稳健优化模型

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-18 DOI: 10.1016/j.tust.2026.107444

Miaomiao Sun , Chengji Liang , Yu Wang , Nikolai Bobylev

Rapid urbanization is causing severe traffic congestion and carbon emissions. Leveraging underground space for Underground Container Logistics Systems (UCLS) is a promising solution, but its adoption is hindered by substantial capital investment barriers. To overcome this hurdle, government interventions, such as carbon taxes and subsidies, are considered critical economic levers. However, the quantitative impact of these combined policies on an operator’s investment decision, especially under demand uncertainty, remains unclear. This study develops a Two-Stage Robust Optimization (2S-RO) model from the port operator’s perspective to address this gap. The model determines the optimal strategic investment in UCLS routes (Stage 1) and the corresponding tactical container flow allocation (Stage 2), minimizing total costs under the worst-case demand scenario characterized by a budget-of-uncertainty set. The model is solved using a Column-and-Constraint Generation (C&CG) algorithm. A case study based on the Shanghai port region, consisting of 2 logistics parks and 3 container ports with annual demand of 20,500 Twenty-foot Equivalent Unit (TEU), analyzes 15 policy scenarios. Results reveal a policy combination “tipping point” effect: neither carbon tax nor subsidy alone triggers UCLS investment, but their combination at a threshold intensity (15 yuan/kg carbon tax + 15 % subsidy) makes UCLS economically viable, achieving 16.13 % cost savings, 59.08 % carbon emission reduction (from 2,847.6 to 1,165.2 tons/year), and a 2.9-year investment payback period. Flow allocation analysis shows that at this tipping point, 48.8 % of container flows shift from road transport to UCLS (37.6 % to shallow systems and 11.2 % to deep systems). Sensitivity analysis demonstrates that demand uncertainty, investment cost variations, and carbon emission caps significantly influence investment decisions: higher uncertainty requires stronger policy support, ±30 % cost variations substantially alter project viability, and emission caps must be set below 2000 tons/year (70 % of baseline) to effectively drive investment. This research provides a quantitative framework for operators to evaluate UCLS projects under uncertainty and offers evidence-based policy design guidance for policymakers, contributing to sustainable urban underground space utilization and port logistics decarbonization.

快速的城市化造成了严重的交通拥堵和碳排放。利用地下空间进行地下集装箱物流系统（UCLS）是一个很有前途的解决方案，但其采用受到大量资本投资壁垒的阻碍。为了克服这一障碍，政府干预，如碳税和补贴，被认为是关键的经济杠杆。然而，这些综合政策对运营商投资决策的定量影响，特别是在需求不确定的情况下，仍然不清楚。本研究从港口运营商的角度开发了一个两阶段稳健优化（2S-RO）模型来解决这一差距。该模型确定了UCLS路线的最优战略投资（阶段1）和相应的战术集装箱流量分配（阶段2），在以不确定预算集为特征的最坏需求情景下，使总成本最小化。该模型采用列约束生成（C&；CG）算法求解。以上海港区为例，分析了15种政策情景。上海港区由2个物流园区和3个集装箱港口组成，年需求量为20,500标准箱。结果表明，政策组合存在“引爆点”效应：碳税和补贴都不能单独触发UCLS投资，但在阈值强度（15元/kg碳税+ 15%补贴）下，两者组合使UCLS具有经济可行性，成本节约16.13%，碳排放量减少59.08%（从2847.6吨/年减少到1165.2吨/年），投资回收期为2.9年。流量分配分析表明，在这个临界点，48.8%的集装箱流量从公路运输转向UCLS（37.6%转向浅层系统，11.2%转向深层系统）。敏感性分析表明，需求不确定性、投资成本变化和碳排放上限显著影响投资决策：较高的不确定性需要更强的政策支持，±30%的成本变化将大大改变项目的可行性，排放上限必须设定在2000吨/年以下（基线的70%）才能有效推动投资。本研究为运营商在不确定性条件下评估UCLS项目提供了定量框架，为决策者提供了基于证据的政策设计指导，有助于城市地下空间可持续利用和港口物流脱碳。

{"title":"A two-stage robust optimization model for underground container logistics system investment under carbon tax and subsidies policies","authors":"Miaomiao Sun , Chengji Liang , Yu Wang , Nikolai Bobylev","doi":"10.1016/j.tust.2026.107444","DOIUrl":"10.1016/j.tust.2026.107444","url":null,"abstract":"<div><div>Rapid urbanization is causing severe traffic congestion and carbon emissions. Leveraging underground space for Underground Container Logistics Systems (UCLS) is a promising solution, but its adoption is hindered by substantial capital investment barriers. To overcome this hurdle, government interventions, such as carbon taxes and subsidies, are considered critical economic levers. However, the quantitative impact of these combined policies on an operator’s investment decision, especially under demand uncertainty, remains unclear. This study develops a Two-Stage Robust Optimization (2S-RO) model from the port operator’s perspective to address this gap. The model determines the optimal strategic investment in UCLS routes (Stage 1) and the corresponding tactical container flow allocation (Stage 2), minimizing total costs under the worst-case demand scenario characterized by a budget-of-uncertainty set. The model is solved using a Column-and-Constraint Generation (C&CG) algorithm. A case study based on the Shanghai port region, consisting of 2 logistics parks and 3 container ports with annual demand of 20,500 Twenty-foot Equivalent Unit (TEU), analyzes 15 policy scenarios. Results reveal a policy combination “tipping point” effect: neither carbon tax nor subsidy alone triggers UCLS investment, but their combination at a threshold intensity (15 yuan/kg carbon tax + 15 % subsidy) makes UCLS economically viable, achieving 16.13 % cost savings, 59.08 % carbon emission reduction (from 2,847.6 to 1,165.2 tons/year), and a 2.9-year investment payback period. Flow allocation analysis shows that at this tipping point, 48.8 % of container flows shift from road transport to UCLS (37.6 % to shallow systems and 11.2 % to deep systems). Sensitivity analysis demonstrates that demand uncertainty, investment cost variations, and carbon emission caps significantly influence investment decisions: higher uncertainty requires stronger policy support, ±30 % cost variations substantially alter project viability, and emission caps must be set below 2000 tons/year (70 % of baseline) to effectively drive investment. This research provides a quantitative framework for operators to evaluate UCLS projects under uncertainty and offers evidence-based policy design guidance for policymakers, contributing to sustainable urban underground space utilization and port logistics decarbonization.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107444"},"PeriodicalIF":7.4,"publicationDate":"2026-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145995547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Automated borehole layout adjustment method in drill-and-blast tunneling: A hierarchical constraint propagation framework 钻爆隧道自动井眼布置调整方法：一个层次约束传播框架

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-18 DOI: 10.1016/j.tust.2026.107478

Zinan Wang , Xiaomeng Shi , Zhaofei Chu , Lang Shi

In drill-and-blast tunneling, real-time borehole layout adjustments are essential when encountering geological hazards such as fault zones or weak interlayers, yet current practices rely on isolated single-point adjustments that create undetected spacing violations and compromise blast effectiveness. This study develops a Hierarchical Constraint Propagation Algorithm (HCPA) to address real-time coordination requirements in automated tunnel construction under dynamic geological conditions. The algorithm employs hierarchical constraint processing: Stage 1 enforces geological boundaries and displacement limits through adaptive positioning strategies, while Stage 2 balances inter-hole spacing via constraint propagation to maintain blast energy distribution. Validation using actual tunnel construction drawings and 1000 Monte Carlo simulations across diverse geological scenarios demonstrates 100% convergence success in practical conditions and 81.1% in adversarial geological scenarios while guaranteeing rigorous engineering safety. Comparative experiments demonstrate order-of-magnitude computational superiority over baseline methods while achieving minimum spacing violations. The sub-second computational efficiency satisfies real-time requirements of automated drilling operations in tunnel construction, providing a deterministic solution for constraint-aware parameter coordination in intelligent drill-and-blast systems.

在钻爆隧道施工中，当遇到断层带或薄弱夹层等地质灾害时，实时调整井眼布局至关重要，但目前的做法依赖于孤立的单点调整，这会造成未被发现的间距偏差，并影响爆破效果。针对动态地质条件下隧道自动化施工的实时协调要求，提出了一种层次约束传播算法（HCPA）。该算法采用分层约束处理，第一阶段通过自适应定位策略强化地质边界和位移限制，第二阶段通过约束传播平衡孔间间距，维持爆破能量分布。通过实际隧道施工图和1000个不同地质场景的蒙特卡罗模拟验证，在保证严格的工程安全的同时，在实际条件下的收敛成功率为100%，在敌对地质场景下的收敛成功率为81.1%。对比实验证明数量级的计算优势优于基线方法，同时实现最小的间距违规。亚秒级的计算效率满足了隧道施工中自动化钻井作业的实时性要求，为智能钻爆系统中约束感知参数协调提供了确定性解决方案。

{"title":"Automated borehole layout adjustment method in drill-and-blast tunneling: A hierarchical constraint propagation framework","authors":"Zinan Wang , Xiaomeng Shi , Zhaofei Chu , Lang Shi","doi":"10.1016/j.tust.2026.107478","DOIUrl":"10.1016/j.tust.2026.107478","url":null,"abstract":"<div><div>In drill-and-blast tunneling, real-time borehole layout adjustments are essential when encountering geological hazards such as fault zones or weak interlayers, yet current practices rely on isolated single-point adjustments that create undetected spacing violations and compromise blast effectiveness. This study develops a Hierarchical Constraint Propagation Algorithm (HCPA) to address real-time coordination requirements in automated tunnel construction under dynamic geological conditions. The algorithm employs hierarchical constraint processing: Stage 1 enforces geological boundaries and displacement limits through adaptive positioning strategies, while Stage 2 balances inter-hole spacing via constraint propagation to maintain blast energy distribution. Validation using actual tunnel construction drawings and 1000 Monte Carlo simulations across diverse geological scenarios demonstrates 100% convergence success in practical conditions and 81.1% in adversarial geological scenarios while guaranteeing rigorous engineering safety. Comparative experiments demonstrate order-of-magnitude computational superiority over baseline methods while achieving minimum spacing violations. The sub-second computational efficiency satisfies real-time requirements of automated drilling operations in tunnel construction, providing a deterministic solution for constraint-aware parameter coordination in intelligent drill-and-blast systems.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107478"},"PeriodicalIF":7.4,"publicationDate":"2026-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145995545","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Calculation and prediction of CO2 concentrations inside a ventilation gallery of Madrid Calle 30 urban tunnels 马德里Calle 30城市隧道通风廊内CO2浓度的计算与预测

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-17 DOI: 10.1016/j.tust.2025.107423

Lucía López-de Abajo, Marcos G. Alberti, Jaime C. Gálvez

Urban tunnels are exposed to high CO₂ concentrations, which can lead to concrete carbonation and considerably reduce the life span of the infrastructure. Carbonation prediction models are powerful tools to forecast the evolution of this phenomenon, with CO₂ concentration as a key input parameter. In this work, CO₂ concentrations inside a ventilation gallery of the Madrid Calle 30 urban tunnels were calculated since their construction in 2007 based on the evolution of the measured traffic intensity inside the tunnels and the evolution of the circulating fleet. A forecast of CO₂ concentrations until 2057 was also performed, considering the evolution of traffic in the area. Regarding the circulating fleet, two different scenarios were studied: one based on the evolution analysed from 2007 to 2022, and another accounting for measures to achieve climate neutrality by 2050. It was concluded that CO₂ concentrations inside an urban tunnel can be calculated and forecasted from traffic intensity data and the composition of its circulating fleet. Also, the methodology presented can be used and adapted to calculate CO₂ concentrations in other urban tunnels for concrete carbonation analysis.

城市隧道暴露在高浓度的二氧化碳中，这可能导致混凝土碳化，并大大减少基础设施的寿命。以CO2浓度为关键输入参数的碳酸化预测模型是预测这一现象演变的有力工具。在这项工作中，根据隧道内测量的交通强度的演变和循环车队的演变，计算了马德里Calle 30城市隧道通风廊内的二氧化碳浓度。考虑到该地区交通的演变，还对2057年前的二氧化碳浓度进行了预测。关于循环船队，研究了两种不同的情景：一种是基于2007年至2022年的演变分析，另一种是考虑到2050年实现气候中和的措施。结果表明，城市隧道内CO2浓度可以通过交通强度数据和其循环车队组成进行计算和预测。此外，所提出的方法也可用于计算其他城市隧道的二氧化碳浓度，用于混凝土碳化分析。

{"title":"Calculation and prediction of CO2 concentrations inside a ventilation gallery of Madrid Calle 30 urban tunnels","authors":"Lucía López-de Abajo, Marcos G. Alberti, Jaime C. Gálvez","doi":"10.1016/j.tust.2025.107423","DOIUrl":"10.1016/j.tust.2025.107423","url":null,"abstract":"<div><div>Urban tunnels are exposed to high CO<sub>2</sub> concentrations, which can lead to concrete carbonation and considerably reduce the life span of the infrastructure. Carbonation prediction models are powerful tools to forecast the evolution of this phenomenon, with CO<sub>2</sub> concentration as a key input parameter. In this work, CO<sub>2</sub> concentrations inside a ventilation gallery of the Madrid Calle 30 urban tunnels were calculated since their construction in 2007 based on the evolution of the measured traffic intensity inside the tunnels and the evolution of the circulating fleet. A forecast of CO<sub>2</sub> concentrations until 2057 was also performed, considering the evolution of traffic in the area. Regarding the circulating fleet, two different scenarios were studied: one based on the evolution analysed from 2007 to 2022, and another accounting for measures to achieve climate neutrality by 2050. It was concluded that CO<sub>2</sub> concentrations inside an urban tunnel can be calculated and forecasted from traffic intensity data and the composition of its circulating fleet. Also, the methodology presented can be used and adapted to calculate CO<sub>2</sub> concentrations in other urban tunnels for concrete carbonation analysis.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107423"},"PeriodicalIF":7.4,"publicationDate":"2026-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979444","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Data-driven causal factor analysis of metro construction incidents using complex network theory 基于复杂网络理论的地铁施工事故数据驱动原因分析

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-16 DOI: 10.1016/j.tust.2026.107464

Pan Zhang , Michael C.P. Sing , Albert P.C. Chan , Shengyu Guo

Near misses and accidents remain two of the most prevalent undesired incidents in metro construction, posing significant threats to safety. While these incidents often share common causal factors, the large number of factors involved in these incidents and the intricate relationships among them make it difficult to identify potential hazards and formulate targeted prevention measures. To address this challenge, this study applied complex network theory to systematically examine the interrelationships among causal factors of metro construction incidents. A case study approach was adopted, drawing on more than 4,000 near-miss and accident reports collected from multiple sources, such as government websites and construction sites. Accident chains and near-miss causation-attribute chains were extracted based on a comprehensive list of causal factors and work attributes (i.e., construction phase, construction area, and worker type). They were then used to construct a two-layer Metro Construction Incident Network (MCIN), capturing the multifaceted interactions between factors. Robustness assessment indicated that strength-based attack was one of the most effective strategies for incident prevention. Also, network topology analysis identified critical causal factors of accidents and near misses and revealed their occurrence patterns across different work attributes. Integrating work attributes into analysis provides greater flexibility for developing targeted prevention strategies for safety risks that are prone to incidents. The findings offer both theoretical insight for advancing accident causation analysis and practical guidance for improving safety risk management in metro construction.

未遂事故和意外事故仍然是地铁建设中最常见的两种不希望发生的事故，对安全构成重大威胁。虽然这些事件往往具有共同的因果因素，但这些事件涉及的因素众多，并且它们之间的关系错综复杂，因此很难识别潜在的危害并制定有针对性的预防措施。为了应对这一挑战，本研究运用复杂网络理论系统地考察了地铁施工事故成因之间的相互关系。采用个案研究方法，从多个来源（如政府网站和建筑工地）收集了4,000多份未遂事故和事故报告。基于原因因素和工作属性（即施工阶段、施工区域和工人类型）的综合列表，提取事故链和未遂事故因果属性链。然后，他们被用来构建一个双层地铁建设事件网络（MCIN），捕捉因素之间的多方面相互作用。鲁棒性评估表明，基于强度的攻击是最有效的事件预防策略之一。此外，网络拓扑分析确定了事故和未遂事故的关键原因因素，并揭示了它们在不同工作属性中的发生模式。将工作属性集成到分析中，为针对容易发生事故的安全风险制定有针对性的预防策略提供了更大的灵活性。研究结果既为推进事故成因分析提供了理论依据，也为加强地铁建设安全风险管理提供了实践指导。

{"title":"Data-driven causal factor analysis of metro construction incidents using complex network theory","authors":"Pan Zhang , Michael C.P. Sing , Albert P.C. Chan , Shengyu Guo","doi":"10.1016/j.tust.2026.107464","DOIUrl":"10.1016/j.tust.2026.107464","url":null,"abstract":"<div><div>Near misses and accidents remain two of the most prevalent undesired incidents in metro construction, posing significant threats to safety. While these incidents often share common causal factors, the large number of factors involved in these incidents and the intricate relationships among them make it difficult to identify potential hazards and formulate targeted prevention measures. To address this challenge, this study applied complex network theory to systematically examine the interrelationships among causal factors of metro construction incidents. A case study approach was adopted, drawing on more than 4,000 near-miss and accident reports collected from multiple sources, such as government websites and construction sites. Accident chains and near-miss causation-attribute chains were extracted based on a comprehensive list of causal factors and work attributes (i.e., construction phase, construction area, and worker type). They were then used to construct a two-layer Metro Construction Incident Network (MCIN), capturing the multifaceted interactions between factors. Robustness assessment indicated that strength-based attack was one of the most effective strategies for incident prevention. Also, network topology analysis identified critical causal factors of accidents and near misses and revealed their occurrence patterns across different work attributes. Integrating work attributes into analysis provides greater flexibility for developing targeted prevention strategies for safety risks that are prone to incidents. The findings offer both theoretical insight for advancing accident causation analysis and practical guidance for improving safety risk management in metro construction.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107464"},"PeriodicalIF":7.4,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979443","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Operational safety assessment of cracked tunnel linings reinforced with polypropylene fiber: investigation, field validation and numerical simulation 聚丙烯纤维加固隧道裂缝衬砌运行安全性评价：调查、现场验证和数值模拟

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-16 DOI: 10.1016/j.tust.2025.107426

Luoyin Li , Xinrong Liu , Ninghui Liang , Ganwen Hu , Kan Zhou , Xiaohan Zhou

Crack control is essential for ensuring the safety and durability of lining structures during tunnel operation. This study investigates the effectiveness of polypropylene fiber-reinforced concrete (PFRC) as a superior alternative to ordinary Portland cement concrete (OPC) for rehabilitating deteriorated linings. Through a combination of field investigations and numerical modeling, a comparative analysis of the mechanical properties and structural performance of OPC and PFRC linings was conducted. The results demonstrated that the secondary lining concrete strength in sections with severe tunnel deterioration fell below the required standards, classifying the lining safety level as Grade III, which necessitates urgent replacement. The incorporation of polypropylene fibers significantly enhances the load-bearing capacity and crack resistance of tunnel linings. Numerical simulations of cracked linings revealed that the stress intensity factor increases, and the stability safety factor decreases with increasing crack depth and width. Crack location and depth exerted a more pronounced influence on structural stability than crack width. Across all crack scenarios, PFRC consistently exhibited a higher safety factor than OPC, attributed to the fibre’s ability to mitigate stress concentration at crack tips. This research confirms the innovative application of PFRC in tunnel rehabilitation, providing a theoretical basis and practical strategy for improving crack resistance and long-term stability.

隧道施工过程中，裂缝控制是保证衬砌结构安全和耐久性的关键。本研究探讨聚丙烯纤维增强混凝土（PFRC）作为普通波特兰水泥混凝土（OPC）修复老化衬里的优越替代品的有效性。通过现场调查和数值模拟相结合的方法，对OPC和PFRC衬砌的力学性能和结构性能进行了对比分析。结果表明，隧道严重劣化路段二次衬砌混凝土强度低于要求标准，衬砌安全等级为III级，需要紧急更换。聚丙烯纤维的掺入显著提高了隧道衬砌的承载能力和抗裂能力。裂纹衬砌的数值模拟结果表明，随着裂纹深度和宽度的增加，衬砌的应力强度系数增大，稳定安全系数减小。裂缝位置和深度对结构稳定性的影响比裂缝宽度更显著。在所有裂缝情况下，PFRC始终表现出比OPC更高的安全系数，这归因于纤维能够减轻裂缝尖端的应力集中。本研究证实了PFRC在隧道修复中的创新应用，为提高隧道抗裂性和长期稳定性提供了理论依据和实践策略。

{"title":"Operational safety assessment of cracked tunnel linings reinforced with polypropylene fiber: investigation, field validation and numerical simulation","authors":"Luoyin Li , Xinrong Liu , Ninghui Liang , Ganwen Hu , Kan Zhou , Xiaohan Zhou","doi":"10.1016/j.tust.2025.107426","DOIUrl":"10.1016/j.tust.2025.107426","url":null,"abstract":"<div><div>Crack control is essential for ensuring the safety and durability of lining structures during tunnel operation. This study investigates the effectiveness of polypropylene fiber-reinforced concrete (PFRC) as a superior alternative to ordinary Portland cement concrete (OPC) for rehabilitating deteriorated linings. Through a combination of field investigations and numerical modeling, a comparative analysis of the mechanical properties and structural performance of OPC and PFRC linings was conducted. The results demonstrated that the secondary lining concrete strength in sections with severe tunnel deterioration fell below the required standards, classifying the lining safety level as Grade III, which necessitates urgent replacement. The incorporation of polypropylene fibers significantly enhances the load-bearing capacity and crack resistance of tunnel linings. Numerical simulations of cracked linings revealed that the stress intensity factor increases, and the stability safety factor decreases with increasing crack depth and width. Crack location and depth exerted a more pronounced influence on structural stability than crack width. Across all crack scenarios, PFRC consistently exhibited a higher safety factor than OPC, attributed to the fibre’s ability to mitigate stress concentration at crack tips. This research confirms the innovative application of PFRC in tunnel rehabilitation, providing a theoretical basis and practical strategy for improving crack resistance and long-term stability.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107426"},"PeriodicalIF":7.4,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Field data-based safety assessment and probabilistic deformation prediction of existing metro tunnels under adjacent excavation 基于现场数据的临近开挖地铁既有隧道安全评价与概率变形预测

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-15 DOI: 10.1016/j.tust.2026.107443

Wei-Bin Chen , Hai-Tong Liu , Yue Chen , Xiang-Sheng Chen , Tao Xu , Jing-Song Bai , Lin-Shuang Zhao

The stratum disturbance caused by excavation will threaten the structural integrity and operational safety of the existing metro tunnels. The data-driven approach proposed in this study mainly focuses on the safety assessment and probabilistic deformation prediction of existing metro tunnels under adjacent excavation operations. In deformation prediction, comparison of the Elman neural network, extreme gradient boosting, support vector machine, and random forest model shows the extreme gradient boosting achieves excellent accuracy and captures convergence variation patterns robustly. For safety assessment, principal component analysis fuses three key deformation indices to generate a comprehensive parameter

Q

. After normality tests confirm

Q

approximates a normal distribution, the “68-95 rule” classifies tunnel safety into 4 levels. For the left tunnel line, the 180-day forecast shows that the deployment of monitoring points under slightly enhanced Level 3 frequency can be moderately expanded. For the right tunnel line, the proportion of high/enhanced-frequency monitoring points can be proportionally reduced. In probabilistic deformation prediction, K-means clustering identifies two optimal clusters for both tunnel lines. Larger Bootstrap sampling enhances the statistical stability of the expendance percentage distribution. Left-line Cluster 2 shows persistently high expendance percentages while right-line Cluster 1 carries higher risk, likely owing to greater burial depth and in-situ stress. Level 1 high-frequency monitoring supplemented by multi-source data is recommended for both high-risk clusters. The proposed risk assessment framework is expected to promote the transformation from empirical thresholds to statistical thresholds and from static risk mapping to dynamic risk mapping.

开挖引起的地层扰动对既有地铁隧道的结构完整性和运行安全构成威胁。本文提出的数据驱动方法主要针对相邻开挖工况下既有地铁隧道的安全性评估和概率变形预测。在变形预测中，通过对Elman神经网络、极端梯度增强、支持向量机和随机森林模型的比较，表明极端梯度增强具有较好的预测精度和较强的收敛性。在安全性评价中，主成分分析将三个关键变形指标融合得到一个综合参数Q。正态性试验证实Q近似于正态分布后，“68-95规则”将隧道安全性分为4个等级。对于左侧隧道线，180天预报显示，3级频率略有增强的监测点部署可以适度扩大。对于正确的隧道线路，可以按比例减少高/增强频率监测点的比例。在概率变形预测中，K-means聚类为两条隧道线路识别两个最优聚类。较大的Bootstrap抽样增强了支出百分比分布的统计稳定性。左线簇2显示出持续的高消耗百分比，而右线簇1风险较高，可能是由于埋深和地应力较大。对于这两个高危集群，建议采用一级高频监测，并辅以多源数据。提出的风险评估框架有望促进从经验阈值到统计阈值的转变，从静态风险映射到动态风险映射。

{"title":"Field data-based safety assessment and probabilistic deformation prediction of existing metro tunnels under adjacent excavation","authors":"Wei-Bin Chen , Hai-Tong Liu , Yue Chen , Xiang-Sheng Chen , Tao Xu , Jing-Song Bai , Lin-Shuang Zhao","doi":"10.1016/j.tust.2026.107443","DOIUrl":"10.1016/j.tust.2026.107443","url":null,"abstract":"<div><div>The stratum disturbance caused by excavation will threaten the structural integrity and operational safety of the existing metro tunnels. The data-driven approach proposed in this study mainly focuses on the safety assessment and probabilistic deformation prediction of existing metro tunnels under adjacent excavation operations. In deformation prediction, comparison of the Elman neural network, extreme gradient boosting, support vector machine, and random forest model shows the extreme gradient boosting achieves excellent accuracy and captures convergence variation patterns robustly. For safety assessment, principal component analysis fuses three key deformation indices to generate a comprehensive parameter <span><math><mrow><mi>Q</mi></mrow></math></span>. After normality tests confirm <span><math><mrow><mi>Q</mi></mrow></math></span> approximates a normal distribution, the “68-95 rule” classifies tunnel safety into 4 levels. For the left tunnel line, the 180-day forecast shows that the deployment of monitoring points under slightly enhanced Level 3 frequency can be moderately expanded. For the right tunnel line, the proportion of high/enhanced-frequency monitoring points can be proportionally reduced. In probabilistic deformation prediction, K-means clustering identifies two optimal clusters for both tunnel lines. Larger Bootstrap sampling enhances the statistical stability of the expendance percentage distribution. Left-line Cluster 2 shows persistently high expendance percentages while right-line Cluster 1 carries higher risk, likely owing to greater burial depth and in-situ stress. Level 1 high-frequency monitoring supplemented by multi-source data is recommended for both high-risk clusters. The proposed risk assessment framework is expected to promote the transformation from empirical thresholds to statistical thresholds and from static risk mapping to dynamic risk mapping.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107443"},"PeriodicalIF":7.4,"publicationDate":"2026-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979440","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Impact of aftershocks on the response of a post-mainshock damaged metro station structure in seismic subsidence site 余震对地震沉降区主震后受损地铁车站结构响应的影响

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-14 DOI: 10.1016/j.tust.2026.107456

Zhong-Liang Zhang , Zhen-Dong Cui , Pengpeng He , Ronald Y.S. Pak

This study investigates the impact of aftershocks on the seismic response of a post-mainshock damaged metro station structure, with a particular focus on the complex dynamic characteristics of seismic subsidence sites. A three-dimensional finite element model was developed to replicate the collapse evolution of a post-mainshock damaged metro station under aftershocks. The results show that under strong mainshocks, the aftershock-induced displacement increment ratio can reach 1.37. Even following a weak mainshock, aftershocks can trigger approximately 40% additional site subsidence. The structural uplift increment ratio decreases with increasing aftershock intensity ratio, with an average value of about 4.4%. The EPWP increment ratio can reach up to 2.4 during aftershocks. Notably, the damage evolution of metro stations exhibits a mainshock threshold effect, i.e., stronger mainshocks lead to earlier damage initiation, with damage ratios exceeding 30%. Critically, aftershocks can exacerbate the damage, forming pervasive damage zones. Importantly, the inter-story drift shows a positive correlation with the damage ratio, surrounding soil displacement increment ratio, and EPWP increment ratio. A modified damage index is proposed to accurately evaluate structural damage under mainshock-aftershock sequences. The findings provide a valuable reference for the seismic design and post-earthquake rescue of metro stations in urban soft soil areas.

本文研究了余震对主震后受损地铁车站结构地震反应的影响，重点研究了地震沉降点的复杂动力特征。建立了三维有限元模型，模拟了主震后受损地铁车站在余震作用下的倒塌演化过程。结果表明：在强主震作用下，余震诱发位移增量比可达1.37；即使在一次微弱的主震之后，余震也会导致大约40%的地面沉降。构造隆升增量比随着余震烈度比的增大而减小，平均约为4.4%。余震时EPWP增量比最高可达2.4。值得注意的是，地铁车站的损伤演化表现出主震阈值效应，即主震越强，损伤发生越早，损伤率超过30%。关键的是，余震会加剧破坏，形成无处不在的破坏区。层间位移与损伤比、周围土体位移增量比、EPWP增量比呈正相关。提出了一种改进的损伤指标，以准确评价主余震作用下的结构损伤。研究结果为城市软土地区地铁车站的抗震设计和震后救援提供了有价值的参考。

{"title":"Impact of aftershocks on the response of a post-mainshock damaged metro station structure in seismic subsidence site","authors":"Zhong-Liang Zhang , Zhen-Dong Cui , Pengpeng He , Ronald Y.S. Pak","doi":"10.1016/j.tust.2026.107456","DOIUrl":"10.1016/j.tust.2026.107456","url":null,"abstract":"<div><div>This study investigates the impact of aftershocks on the seismic response of a post-mainshock damaged metro station structure, with a particular focus on the complex dynamic characteristics of seismic subsidence sites. A three-dimensional finite element model was developed to replicate the collapse evolution of a post-mainshock damaged metro station under aftershocks. The results show that under strong mainshocks, the aftershock-induced displacement increment ratio can reach 1.37. Even following a weak mainshock, aftershocks can trigger approximately 40% additional site subsidence. The structural uplift increment ratio decreases with increasing aftershock intensity ratio, with an average value of about 4.4%. The EPWP increment ratio can reach up to 2.4 during aftershocks. Notably, the damage evolution of metro stations exhibits a mainshock threshold effect, i.e., stronger mainshocks lead to earlier damage initiation, with damage ratios exceeding 30%. Critically, aftershocks can exacerbate the damage, forming pervasive damage zones. Importantly, the inter-story drift shows a positive correlation with the damage ratio, surrounding soil displacement increment ratio, and EPWP increment ratio. A modified damage index is proposed to accurately evaluate structural damage under mainshock-aftershock sequences. The findings provide a valuable reference for the seismic design and post-earthquake rescue of metro stations in urban soft soil areas.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107456"},"PeriodicalIF":7.4,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145962609","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Digital twin-based framework for multi-objective optimization of shield tunneling parameters 基于数字孪生的盾构隧道参数多目标优化框架

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-13 DOI: 10.1016/j.tust.2026.107457

Congzhen Yang , Zhikun Ding , Tianrui Liu , Zezhou Wu , Ke Chen

Optimizing shield tunneling parameters is critical for ensuring both safety and efficiency in tunnel construction. However, existing optimization approaches often underutilize operational data and are prone to entrapment in local optima. This study proposes a digital twin (DT) -based framework for multi-objective optimization (MOO) of shield tunneling parameters. The framework integrates data acquisition, preprocessing, modeling, and optimization within a layered architecture. Key parameters are identified using the shapley additive explanations (SHAP) method, while a hybrid optimization algorithm—artificial bee colony (ABC)–non-dominated sorting genetic algorithm III (NSGA-III)—combines the strengths of both algorithms and is applied across three operational scenarios. Optimized parameters are iteratively fed back into the DT to guide parameter adjustment. The framework is validated using data from the Shanghai Airport Link Line project. Full-parameter optimization yields the best performance, achieving an overall optimization rate of 32.02%, with particularly notable improvements in controlling the vertical deviation of shield head. Comparative analyses show that the proposed framework surpasses benchmark methods in convergence speed and solution quality, reducing shield attitude deviation by 2.21%–17.13%. These results underscore the framework’s potential as an effective decision-support tool for shield tunneling operations.

盾构施工参数的优化是保证隧道施工安全和效率的关键。然而，现有的优化方法往往没有充分利用操作数据，并且容易陷入局部最优。提出了一种基于数字孪生（DT）的盾构隧道参数多目标优化框架。该框架在分层体系结构中集成了数据采集、预处理、建模和优化。使用shapley加性解释（SHAP）方法确定关键参数，而混合优化算法-人工蜂群(ABC) -非主导排序遗传算法III (NSGA-III) -结合了两种算法的优势，并应用于三种操作场景。优化后的参数迭代反馈到DT中，指导参数调整。利用上海机场连接线项目的数据对该框架进行了验证。全参数优化效果最好，总体优化率为32.02%，在控制盾头垂直偏差方面效果尤为显著。对比分析表明，该框架在收敛速度和求解质量上均优于基准方法，将盾构姿态偏差降低了2.21% ~ 17.13%。这些结果强调了该框架作为盾构施工有效决策支持工具的潜力。

{"title":"Digital twin-based framework for multi-objective optimization of shield tunneling parameters","authors":"Congzhen Yang , Zhikun Ding , Tianrui Liu , Zezhou Wu , Ke Chen","doi":"10.1016/j.tust.2026.107457","DOIUrl":"10.1016/j.tust.2026.107457","url":null,"abstract":"<div><div>Optimizing shield tunneling parameters is critical for ensuring both safety and efficiency in tunnel construction. However, existing optimization approaches often underutilize operational data and are prone to entrapment in local optima. This study proposes a digital twin (DT) -based framework for multi-objective optimization (MOO) of shield tunneling parameters. The framework integrates data acquisition, preprocessing, modeling, and optimization within a layered architecture. Key parameters are identified using the shapley additive explanations (SHAP) method, while a hybrid optimization algorithm—artificial bee colony (ABC)–non-dominated sorting genetic algorithm III (NSGA-III)—combines the strengths of both algorithms and is applied across three operational scenarios. Optimized parameters are iteratively fed back into the DT to guide parameter adjustment. The framework is validated using data from the Shanghai Airport Link Line project. Full-parameter optimization yields the best performance, achieving an overall optimization rate of 32.02%, with particularly notable improvements in controlling the vertical deviation of shield head. Comparative analyses show that the proposed framework surpasses benchmark methods in convergence speed and solution quality, reducing shield attitude deviation by 2.21%–17.13%. These results underscore the framework’s potential as an effective decision-support tool for shield tunneling operations.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107457"},"PeriodicalIF":7.4,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145957103","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards low-carbon construction of metro station foundation pit: A probabilistic digital twin framework with self-supervised learning capability 面向地铁车站基坑低碳施工：具有自监督学习能力的概率数字孪生框架

IF 7.4 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Tunnelling and Underground Space Technology

Pub Date : 2026-01-13 DOI: 10.1016/j.tust.2026.107450

Weizong Lai , Yue Pan , Limao Zhang , Jin-Jian Chen , Jianjun Qin

Metro station foundation pit construction in large cities like Shanghai with high traffic demands and complicated geological conditions towards greater and deeper dimensions and contributes significant carbon emissions from intensive material and energy use. Efficient tool is urgently needed to quantify the carbon emissions and support engineering decisions, particularly as emissions data increasingly inform both regulatory compliance and carbon-related financial mechanisms. However, we recognize that both emission data and engineering inputs are highly uncertain, yet prevailing methods ignore this uncertainty and lack component-level modeling. To address this, a novel probabilistic digital twin (prob-DT) framework with self-supervised learning capability is proposed to uncover uncertainties in carbon emission quantification and identify the optimal engineering solution from carbon emission perspective. Integrating semantic and geometric information at the component level, prob-DT constructs a carbon knowledge base by self-supervised matching of consumption quotas to emission factors and models carbon emissions across construction stages. It propagates uncertainty and characterizes system-level risk via probabilistic analysis and Monte Carlo simulation. By comparing alternatives probabilistically, prob-DT identifies optimal low-carbon engineering schemes. Finally, the proposed prob-DT is instantiated for Digital Twin for Carbon Quantification for Metro Station Foundation Pit (DTCQ-MetroPit) system and applied to the Huangpi South Road Station project in Shanghai. Results indicate that the station exhibits relatively high carbon emissions due to its elongated geometry, which necessitates longer diaphragm walls and increased material consumption. Under an optimized strategy recommended by DTCQ-MetroPit, the Conditional Value at Risk (CVaR₉₅) of carbon emissions is reduced from 93,700 to 85,700 tons, demonstrating the framework’s effectiveness in guiding low-carbon engineering practices under uncertainties.

在上海这样交通需求大、地质条件复杂的大城市，地铁车站基坑建设向更大、更深的维度发展，并且由于材料和能源的密集使用，造成了巨大的碳排放。迫切需要有效的工具来量化碳排放并支持工程决策，特别是在排放数据越来越多地为监管合规和碳相关金融机制提供信息的情况下。然而，我们认识到排放数据和工程输入都是高度不确定的，但主流方法忽略了这种不确定性，缺乏组件级建模。为了解决这一问题，提出了一种具有自监督学习能力的概率数字孪生（probt - dt）框架，以揭示碳排放量化中的不确定性，并从碳排放的角度确定最优工程解决方案。probi - dt结合构件层面的语义信息和几何信息，通过消费配额与排放因子的自监督匹配构建碳知识库，并对各施工阶段的碳排放进行建模。它传播不确定性，并通过概率分析和蒙特卡罗模拟表征系统级风险。通过对备选方案进行概率比较，probd - dt确定了最优的低碳工程方案。最后，将提出的probo - dt方法应用于地铁车站基坑碳量化数字孪生系统（DTCQ-MetroPit），并应用于上海黄陂南路站工程。结果表明，该站由于其细长的几何形状而表现出相对较高的碳排放，这需要更长的连续墙和增加的材料消耗。在dtcq - metritit推荐的优化策略下，碳排放的条件风险值（CVaR95）从9.37万吨减少到8.57万吨，显示了该框架在不确定条件下指导低碳工程实践的有效性。

{"title":"Towards low-carbon construction of metro station foundation pit: A probabilistic digital twin framework with self-supervised learning capability","authors":"Weizong Lai , Yue Pan , Limao Zhang , Jin-Jian Chen , Jianjun Qin","doi":"10.1016/j.tust.2026.107450","DOIUrl":"10.1016/j.tust.2026.107450","url":null,"abstract":"<div><div>Metro station foundation pit construction in large cities like Shanghai with high traffic demands and complicated geological conditions towards greater and deeper dimensions and contributes significant carbon emissions from intensive material and energy use. Efficient tool is urgently needed to quantify the carbon emissions and support engineering decisions, particularly as emissions data increasingly inform both regulatory compliance and carbon-related financial mechanisms. However, we recognize that both emission data and engineering inputs are highly uncertain, yet prevailing methods ignore this uncertainty and lack component-level modeling. To address this, a novel probabilistic digital twin (prob-DT) framework with self-supervised learning capability is proposed to uncover uncertainties in carbon emission quantification and identify the optimal engineering solution from carbon emission perspective. Integrating semantic and geometric information at the component level, prob-DT constructs a carbon knowledge base by self-supervised matching of consumption quotas to emission factors and models carbon emissions across construction stages. It propagates uncertainty and characterizes system-level risk via probabilistic analysis and Monte Carlo simulation. By comparing alternatives probabilistically, prob-DT identifies optimal low-carbon engineering schemes. Finally, the proposed prob-DT is instantiated for Digital Twin for Carbon Quantification for Metro Station Foundation Pit (DTCQ-MetroPit) system and applied to the Huangpi South Road Station project in Shanghai. Results indicate that the station exhibits relatively high carbon emissions due to its elongated geometry, which necessitates longer diaphragm walls and increased material consumption. Under an optimized strategy recommended by DTCQ-MetroPit, the Conditional Value at Risk (CVaR<sub>95</sub>) of carbon emissions is reduced from 93,700 to 85,700 tons, demonstrating the framework’s effectiveness in guiding low-carbon engineering practices under uncertainties.</div></div>","PeriodicalId":49414,"journal":{"name":"Tunnelling and Underground Space Technology","volume":"171 ","pages":"Article 107450"},"PeriodicalIF":7.4,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145962002","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0