Computer-Aided Civil and Infrastructure Engineering最新文献

{"title":"A multibranch and multiattention framework for floor plan segmentation","authors":"Axel De Nardin, Silvia Zottin, Andrea Toma, Claudio Piciarelli, Gian Luca Foresti","doi":"10.1111/mice.70030","DOIUrl":"https://doi.org/10.1111/mice.70030","url":null,"abstract":"Automated floor plan analysis is crucial in architecture, urban planning, and interior design. Floor plan segmentation is a foundational step for tasks such as surface area estimation and three‐dimensional building reconstruction. However, automatic semantic segmentation of floor plan images faces unique challenges, including high interclass similarity, ambiguous room boundaries, and varying floor plan styles. We introduce a novel multibranch and multiattention framework for deep floor plan segmentation, explicitly designed to handle the challenges of interclass similarity, ambiguous room boundaries, and diverse architectural styles. Our method leverages intrabranch channel attention and cross‐branch positional attention to refine both boundary recognition and room‐type segmentation, significantly enhancing robustness and accuracy across multiple datasets. Through extensive experiments on the raster‐to‐vector (R2V) and R3D datasets, we demonstrate how our approach sets a new state‐of‐the‐art for floor plan segmentation, outperforming general‐purpose and specialized models alike.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"27 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144747504","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}

{"title":"Cover Image, Volume 40, Issue 19","authors":"","doi":"10.1111/mice.70018","DOIUrl":"https://doi.org/10.1111/mice.70018","url":null,"abstract":"<p><b>The cover image</b> is based on the article <i>Infrared thermography and 3D pavement surface unevenness measurement algorithm for damage assessment of concrete bridge decks</i> by Mikiko Yamashita et al., https://doi.org/10.1111/mice.13406.\u0000\u0000 <figure>\u0000 <div><picture>\u0000 <source></source></picture><p></p>\u0000 </div>\u0000 </figure></p>","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"40 19","pages":""},"PeriodicalIF":8.5,"publicationDate":"2025-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/mice.70018","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144695856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cover Image, Volume 40, Issue 19","authors":"","doi":"10.1111/mice.70019","DOIUrl":"https://doi.org/10.1111/mice.70019","url":null,"abstract":"<p><b>The cover image</b> is based on the article <i>Parallel computing aided analyses of dynamic buckling for railway track infrastructure</i> by D. Agustin et al., https://doi.org/10.1111/mice.70004.\u0000\u0000 <figure>\u0000 <div><picture>\u0000 <source></source></picture><p></p>\u0000 </div>\u0000 </figure></p>","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"40 19","pages":""},"PeriodicalIF":8.5,"publicationDate":"2025-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/mice.70019","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144695857","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Learning loads on in‐service underground infrastructure with a trans‐dimensional Bayesian inversion method","authors":"Zhiyao Tian, Xianfei Yin, Shunhua Zhou","doi":"10.1111/mice.70024","DOIUrl":"https://doi.org/10.1111/mice.70024","url":null,"abstract":"Monitoring external loads on underground structures is crucial for structural health assessment. Inverting earth pressures from observable structural responses, such as deformation data, holds promise. However, existing methods often rely on presumptions about pressure complexity, which can be infeasible for many poorly performing in‐service infrastructures. This paper proposes a trans‐dimensional Bayesian method that simultaneously infers both the complexity and magnitude of earth pressures by parameterizing a set of a priori unknown variables, where the exact number of parameters remains undetermined. A Bayesian framework is employed to represent the posterior distribution of these parameters, with a trans‐dimensional Markov chain specifically designed for statistical inference of the distributed pressures. Case studies demonstrate that the proposed method outperforms traditional methods, which are limited by rigid presumptions. Furthermore, it is shown that the inferred pressures can reproduce comprehensive structural responses, such as internal forces, providing new tools and insights for structural health monitoring of underground infrastructures.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"106 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144684620","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}

{"title":"Coarse‐grained molecular dynamics reveals rejuvenation behavior in aged asphalt","authors":"Liang He, Haitao Huang, Wentao Zhang, Changjiang Dai, Guannan Li, Alessio Alexiadis, Mengzhe Tao, Wim Van den bergh, Karol J. Kowalski","doi":"10.1111/mice.70017","DOIUrl":"https://doi.org/10.1111/mice.70017","url":null,"abstract":"To address the issue of insufficient spatiotemporal scales in all‐atom molecular dynamics (MD) simulations, this study employs coarse‐grained MD (CGMD) based on the MARTINI 3.0 force field to simulate the rejuvenation behavior of asphalt. Based on all‐atom molecular structure coarse‐grained mapping, molecular models of both virgin and aged asphalt are constructed, with iterative optimization of parameters. Coarse‐grained models of asphalt were constructed and validated through density, glass transition temperature, and visualization analysis. Rejuvenation diffusion model is constructed, penetration tube tests are designed to analyze the rejuvenation behavior, and further verified through rheological performance and microstructural analysis. The results show that the rejuvenation efficiency depends on temperature, dosage, and structure of the rejuvenator molecule. Aromatic oil shows stronger interaction, while peanut oil demonstrates higher diffusivity. This study explored the application of MARTINI 3.0 force field in CGMD, providing guidance for the further application of coarse‐grained methods in asphalt behavior analysis.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"282 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144677417","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}

{"title":"A multimodal deep learning approach for predicting traffic accident severity using crash records, road geometry, and textual descriptions","authors":"Yue Liu, Zhixiang Gao, Hanzhang Ge, Ziyu Chen, Guohua Liang, Yonggang Wang, Yuting Zhang","doi":"10.1111/mice.70023","DOIUrl":"https://doi.org/10.1111/mice.70023","url":null,"abstract":"Accurately predicting traffic accident severity is critical for improving road safety management and targeted prevention strategies. This study proposes a novel multimodal deep learning framework integrating structured accident records, detailed road geometry, and unstructured textual descriptions. To our knowledge, this research offers the first large‐scale dataset that combines linear roadway geometry with accident reports for comprehensive severity analysis. The proposed model employs advanced multimodal fusion techniques, including attention‐based gating mechanisms, a mixture‐of‐experts module, and cross‐feature interactions, effectively capturing complex interdependencies among various data modalities. In addition, this paper implements a three‐tiered interpretability analysis at the modality, expert, and feature levels, using SHapley Additive exPlanations values to transparently explain the model predictions. Experimental results demonstrate that our model significantly outperforms traditional and baseline methods, particularly in identifying severe accidents. Interpretability analyses highlight critical insights into accident severity, emphasizing textual descriptions and detailed road characteristics.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"25 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144678164","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}

{"title":"Sensitivity of inflow turbulence uncertainty on wind pressure on high‐rise buildings using large eddy simulations","authors":"L. W. Chew, A. F. Melaku, M. F. Ciarlatani, C. Gorlé","doi":"10.1111/mice.70016","DOIUrl":"https://doi.org/10.1111/mice.70016","url":null,"abstract":"Large eddy simulations (LES) can aid the prediction of wind loading on buildings, provided that representative inflow turbulence properties are prescribed. This study conducts LES to assess the sensitivity of the mean, root mean square (rms) fluctuation, and peak pressure coefficients (<jats:italic>C<jats:sub>p</jats:sub></jats:italic>) on building surfaces to the uncertainties in the incoming flow turbulence. Compared to wind tunnel measurements, the simulated mean <jats:italic>C<jats:sub>p</jats:sub></jats:italic> is well predicted, and the variation in inflow turbulence has a negligible effect. The rms <jats:italic>C<jats:sub>p</jats:sub></jats:italic> increases with increasing turbulence intensities and increasing turbulence length scales. Increasing inlet values of turbulence intensities and turbulence length scales reduces the root mean square errors (RMSE) of rms <jats:italic>C<jats:sub>p</jats:sub></jats:italic> from 0.049 to 0.026 and from 0.047 to 0.024, respectively, on the side surfaces with flow separation. The minimum <jats:italic>C<jats:sub>p</jats:sub></jats:italic> responds similarly, where the RMSE is reduced from 0.382 to 0.280 and from 0.385 to 0.286. The maximum <jats:italic>C<jats:sub>p</jats:sub></jats:italic> on the windward surface achieves the lowest RMSE of 0.089 at nominal inlet values. The agreement between LES and experiment improves significantly after incorporating uncertainties in the input turbulence properties by repeating simulations with smaller and larger values from the estimated turbulence inputs. Wind tunnel experiments often do not measure the complete turbulence properties of the incoming flow, thereby obscuring the validation process of simulation results. The findings recommend wind tunnel experiments to measure and report the complete turbulence properties of the incoming flow for accurate prediction of wind loading.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"6 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144639814","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}

{"title":"Parallel computing aided analyses of dynamic buckling for railway track infrastructure","authors":"Dan Agustin,&nbsp;Qing Wu,&nbsp;Maksym Spiryagin,&nbsp;Colin Cole,&nbsp;Esteban Bernal","doi":"10.1111/mice.70004","DOIUrl":"10.1111/mice.70004","url":null,"abstract":"<p>This paper presents a scalable parallel computing framework for simulating track buckling under dynamic train loads, enabling large-scale railway track stability analysis. A three-dimensional (3D) track model is developed using finite element-based Euler–Bernoulli beam formulations for rails, dynamic force inputs, and nonlinear interactions at the sleeper–ballast interface to capture dynamic buckling behavior. To address computational challenges in simulating extended track sections, the framework employs message passing interface–based parallelization, optimizing load balancing, and minimizing interprocess communication overhead. Unlike approaches that simulate long tracks virtually by recycling a small domain, the proposed method maintains complete dynamic and structural detail across the entire track length. It dynamically adjusts lateral rail stiffness and incorporates thermal compression effects to enable simulation of buckling behavior, while efficiently scaling across high-performance computing clusters. Case studies demonstrate the framework's ability to simulate large-scale tracks under combined thermal gradients and dynamic train loads, achieving near-linear speedup and reducing runtime by up to 90% compared to serial approaches. Additionally, a machine learning–based buckling risk assessment is presented as a use case, where a model trained on long-track simulation results predicts buckling risk across extended sections. By integrating 3D track dynamics, parallel computing, and data-driven risk assessment, this work provides a powerful tool for evaluating railway infrastructure resilience under extreme operational conditions.</p>","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"40 19","pages":"2943-2968"},"PeriodicalIF":8.5,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/mice.70004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144629562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Computer vision wireless sensors for displacement influence line/surface measurement of footbridges using stationary pedestrian loading","authors":"Miaomin Wang, Huiqi Liang, Zuo Zhu, Huifang Wu, Fuyou Xu, Ki‐Young Koo, James Brownjohn","doi":"10.1111/mice.70008","DOIUrl":"https://doi.org/10.1111/mice.70008","url":null,"abstract":"Although using a heavy vehicle at a consistent speed is a common method to estimate displacement influence lines or displacement influence surfaces (DILs/DISs) of vehicular bridges, it requires algorithms to separate the dynamic and static components from measured displacements. This decomposition process can introduce uncertainties in the results. Additionally, employing vehicles is logistically impractical for most footbridges. To overcome these issues, this paper proposes a new, practical framework using computer vision to measure DILs/DISs on footbridges. It combines a stationary pedestrian loading strategy with a computer vision input–output wireless sensor network (CVIO‐WSN). This framework has two main features: (1) pedestrians follow the “step‐and‐stand” rule, and their weight acts as a static load when they stand still at discrete locations across the footbridge for DIL/DIS measurement; (2) CVIO‐WSN consists of input nodes for human load localization and output nodes for simultaneous structural response measurement, allowing load and response data to be collected in a contactless way that minimizes disruption to operational structures. Two laboratory experiments were conducted to validate this system. The first evaluated the timestamping accuracy between two identical sensor nodes tracking the same moving target, showing an average synchronization error of 2.39 ms. The second assessed the localization accuracy of the input nodes, with the average error of 14.0 mm on the X‐axis and 16.9 mm on the Y‐axis. The method was then applied to an experimental floor structure and an operational full‐scale footbridge. In the first application, the input nodes tracked a human through a sequence of 77 locations across the floor, while the output node measured structural displacement at the center, successfully obtaining the structural DIS. In the second application, the method localized four humans (pedestrians) moving to discrete locations across an operational arch footbridge and briefly remaining stationary while measuring displacement at two points of the structure. Although the measurement results were promising, using heavier pedestrians or increasing their number is recommended to improve the signal‐to‐noise ratio of the structural displacement measurements.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"9 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144611156","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}

{"title":"Large language model for post‐earthquake structural damage assessment of buildings","authors":"Yongqing Jiang, Jianze Wang, Xinyi Shen, Kaoshan Dai","doi":"10.1111/mice.70010","DOIUrl":"https://doi.org/10.1111/mice.70010","url":null,"abstract":"A rapid and accurate assessment of structural damage to buildings in the aftermath of earthquakes is critical to emergency responses and engineering retrofit decisions. However, current in situ building damage assessment is primarily conducted through visual inspections by engineering professionals and deep learning techniques using single‐modal information, which are time‐consuming and unable to effectively integrate visual and textual information. In recent years, multimodal learning methods and large language models (LLMs), which could process visual and linguistic information, have emerged as viable alternatives for damage assessment of building constructions. In this study, a vision question–answering model for structural damage assessment (SDA‐Chat) is developed that automatically generates professional textual interpretations of structural damage images via multi‐round visual question–answering (VQA) interactions. A three‐stage training strategy that includes instruction fine‐tuning is designed to improve the model's VQA accuracy. The cross‐modality projector based on dimension reshaping and parallel network architecture is developed to enhance the accuracy and speed of alignment of multimodal features. Comparative experiments are conducted on the self‐constructed dataset containing 8195 pairs of structural damage images and corresponding damage description texts, focusing on various advanced LLMs. The results highlight that the SDA‐Chat can simultaneously process seven different tasks, demonstrating the effectiveness of the proposed method. The highest question–answering accuracy and efficiency of the model reached 83.04% and 435.31 tokens/s, respectively. In addition, high‐precision and lightweight solutions are designed for different application scenarios.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"46 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2025-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144611162","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}