Computer-Aided Civil and Infrastructure Engineering最新文献

{"title":"Structural complexity as a new paradigm in construction cost optimization","authors":"Yuan Ma,Chaolin Song,Laura Sardone,Rucheng Xiao,Giuseppe Carlo Marano","doi":"10.1016/j.cacaie.2026.100186","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100186","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"36 1","pages":"100186"},"PeriodicalIF":11.775,"publicationDate":"2026-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149090939","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":"Unified segmentation of surface and subsurface façade defects via infrared–visible fusion and temperature perturbation cloning","authors":"Xiaolan Zhuo,Ziyu Wang,Jiangpeng Shu,Xiaowu Zhang,Ziyue Zeng,Jingnan Zhang,Weibo Li","doi":"10.1016/j.cacaie.2026.100226","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100226","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"19 1","pages":"100226"},"PeriodicalIF":11.775,"publicationDate":"2026-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149090941","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":"Transformer-based controllable diffusion model for high-fidelity railway track slab crack image generation","authors":"Chao Zou, Zhixiong Chen, Yu Qian, Wenbo Hu, Feng Guo, Lizhuang Cui","doi":"10.1016/j.cacaie.2026.100250","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100250","url":null,"abstract":"Artificial intelligence (AI)-based visual inspection has achieved superior performance in detecting railway track slab cracks, delivering high accuracy yet maintaining computational efficiency. However, its widespread deployment continues to be challenged by the severe shortage of high-quality annotated images for effective model training, particularly for narrow, low-contrast track slab cracks with complex morphologies and confounding background textures. Although synthetic data have been used to support railway track slab crack segmentation, controllable generation at specified locations on crack-free backgrounds, with attention to both fine crack details and long-range continuity, remains insufficiently explored. We therefore propose a controllable Transformer-based diffusion model conditioned on a background image and a crack mask to generate high-fidelity track slab crack images. Differing from conventional diffusion models that employ U-Net-based architectures, we take advantage of a Transformer backbone to capture long-range spatial dependencies while preserving the global connectivity and topological consistency of irregular crack structures. A refined convolutional feature extraction module is developed to improve the representation of narrow crack boundaries, subtle texture variations, and local geometric details. By jointly modeling global crack topology and local fine-grained features, the proposed model enables the controllable generation of diverse crack patterns under varied conditions. Experimental results show that the generated railway track slab crack images possess realistic textures, clear boundaries, good structural continuity, and accurate preservation of details. In quantitative analysis, the proposed method achieves a PSNR of 27.5122 dB, a LPIPS of 0.0752, and a FID of 57.8792. The Edge Intersection over Union (IoU<ce:inf loc=\"post\">Edge</ce:inf>) and skeleton similarity reach 0.5645 and 0.5209, respectively, significantly outperforming similar methods. Furthermore, segmentation experiments demonstrate that the high-quality track slab crack images generated by this model can continuously improve the segmentation performance of downstream models. Compared with suboptimal methods (CDM), the proposed method improves the overall segmentation IoU and F1 scores by 5.90 and 4.80 percentage points, respectively. For mesh-like fine cracks, its IoU is 2.35 percentage points higher than that of CDM. The proposed method therefore provides a potential solution for reducing “effective training data anxiety” and manual annotation costs, supporting the development of more robust intelligent inspection systems for railway track slabs.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"26 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2026-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149221704","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":"Overbreak intelligent control method of drilling and blasting tunnel based on hyperparameter-tuned prediction and metaheuristic parameter optimization","authors":"Fu Huang, Jian Xiao, Changgeng Xie, Yongtao Wang, Min Zhang","doi":"10.1016/j.cacaie.2026.100242","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100242","url":null,"abstract":"Overbreak is inevitable for a tunnel excavated by drill-and-blast method, which has negative effects on both construction quality and economic efficiency of the tunnel. Since many previous studies mainly focus on overbreak intelligent prediction, the overbreak control method in the framework of the intelligent algorithm is not well investigated. In this study, an integrated “prediction-optimization-validation” method for tunnel overbreak control was proposed. Based on 356 sets of laser scanning data from four tunnels, a tunnel overbreak area prediction model was constructed based on an Optuna hyper parameter optimization algorithm. Using the minimum overbreak area as the optimization objective, an Grey Wolf Optimizer (GWO) was applied to optimize the blasting parameters and an optimal blasting scheme of the tunnel was determined. Moreover, a FEM-based fluid-structure interaction (FSI) numerical model was constructed to determine the overbreak area of the tunnel based on the optimal blasting scheme. By comparing the overbreak areas of the tunnel obtained from numerical simulation and those provided by the prediction model, the validity of the method proposed here is proved. The comparison of the overbreak area between the original blasting scheme and the optimal blasting scheme shows that the optimal blasting scheme can reduce the tunnel overbreak area by 44.0%. The method proposed in this paper provides a feasible solution for the overbreak control of tunnel excavated by drill-and-blast method.","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"1 1","pages":""},"PeriodicalIF":11.775,"publicationDate":"2026-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149087081","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":"Bidirectional Monte Carlo Tree Search for Coordinated Railway Alignment and Dominant-Structure Location Optimization in Fragmented Design Spaces","authors":"Qingxin Zeng,Hao Pu,Paul Schonfeld,Guanghui Wang,Taoran Song,Hong Zhang,Wei Li,Jianping Hu,Jiangtao Liu","doi":"10.1016/j.cacaie.2026.100241","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100241","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"17 1","pages":"100241"},"PeriodicalIF":11.775,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149090952","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":"POD-HDWNO: A hierarchical decoupled wavelet neural operator framework with POD reconstruction for efficient full-field response prediction of high-speed railway bridges","authors":"Libing Chen,Li Song,Zhiwu Yu,Xin Nie,Lilong Fan,Jian Hou","doi":"10.1016/j.cacaie.2026.100245","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100245","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"24 1","pages":"100245"},"PeriodicalIF":11.775,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149150484","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":"UAV-based pavement distress segmentation via controllable synthetic data augmentation and a lightweight selective fusion network under limited annotations","authors":"Jinfeng Yan,Cheng Fang,Yifan Wang,Yuhui Zhang,Fujiao Tang,Dawei Wang,Zepeng Fan","doi":"10.1016/j.cacaie.2026.100204","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100204","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"8 1","pages":"100204"},"PeriodicalIF":11.775,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148894869","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":"Real-time multi-step prediction of tunnel boring machine parameters using a decomposition-enhanced dual-stream deep learning model","authors":"Houle Zhang,Yongxin Wu,Mao Ye,Hanzhi Yang,Yue Hou","doi":"10.1016/j.cacaie.2026.100106","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100106","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"8 1","pages":"100106"},"PeriodicalIF":11.775,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149021545","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":"Pretrained geometry-physics-language surrogate for text-conditioned bridge response screening","authors":"Kewei Ren,Pang-jo Chun","doi":"10.1016/j.cacaie.2026.100162","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100162","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"36 1","pages":"100162"},"PeriodicalIF":11.775,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149021547","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":"Digital twinning of a steel composite bridge using semantic data structures","authors":"Rolando Chacón,Carlos Ramonell,Héctor Posada,Pablo Sierra,Christian Martínez de la Rosa,Rahul Tomar","doi":"10.1016/j.cacaie.2026.100142","DOIUrl":"https://doi.org/10.1016/j.cacaie.2026.100142","url":null,"abstract":"","PeriodicalId":156,"journal":{"name":"Computer-Aided Civil and Infrastructure Engineering","volume":"19 1","pages":"100142"},"PeriodicalIF":11.775,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149021551","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}