Pub Date : 2025-01-21DOI: 10.1016/j.autcon.2025.105974
Van Hien Nguyen, Tri Cuong Do, Kyoung Kwan Ahn
This paper presents a powertrain integrated with an energy regeneration system designed to decrease energy consumption and emissions in hybrid hydraulic excavators. The feature of this powertrain is its integration of a hydrostatic transmission (HST), which optimizes torque and speed between the pump and power sources. Additionally, the energy regeneration system includes two distinct methods: employing an independent hydraulic motor or utilizing the hydraulic motor mode of the main hydraulic pump/motor to recapture potential energy from the boom cylinder. Furthermore, an energy management approach, leveraging an equivalent consumption minimization strategy, is proposed to distribute power requirements, aiming to maximize engine efficiency. Compared to existing hybrid and conventional systems, both experimental and simulation results show that the proposed system can achieve energy savings of up to 61.64%. Additionally, the energy regeneration efficiency of the proposed system can reach up to 73% during the cylinder lowering process.
{"title":"Hybrid powertrain with dual energy regeneration for boom cylinder movement in a hydraulic excavator","authors":"Van Hien Nguyen, Tri Cuong Do, Kyoung Kwan Ahn","doi":"10.1016/j.autcon.2025.105974","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105974","url":null,"abstract":"This paper presents a powertrain integrated with an energy regeneration system designed to decrease energy consumption and emissions in hybrid hydraulic excavators. The feature of this powertrain is its integration of a hydrostatic transmission (HST), which optimizes torque and speed between the pump and power sources. Additionally, the energy regeneration system includes two distinct methods: employing an independent hydraulic motor or utilizing the hydraulic motor mode of the main hydraulic pump/motor to recapture potential energy from the boom cylinder. Furthermore, an energy management approach, leveraging an equivalent consumption minimization strategy, is proposed to distribute power requirements, aiming to maximize engine efficiency. Compared to existing hybrid and conventional systems, both experimental and simulation results show that the proposed system can achieve energy savings of up to 61.64%. Additionally, the energy regeneration efficiency of the proposed system can reach up to 73% during the cylinder lowering process.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"34 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027404","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}
Pub Date : 2025-01-21DOI: 10.1016/j.autcon.2025.105982
Chenxiang Ma, Chengcheng Xu
Highway interchanges have traditionally been designed through a time-consuming manual process. To enhance efficiency and effectiveness, this paper develops an objective-directed automatic and intelligent interchange design method using graph conditional variational autoencoder. Based on interchange graph representation and augmentation techniques, data are collected from diverse interchanges types and converted into graphs that store design parameters. Aiming at graph reconstruction and fitting data distribution, proposed model learns to generate optimized interchanges by embedding design objectives including throughput and total ramp length. For evaluation, predictors are used to directly output interchange properties, enabling the quick screening of structures. Results demonstrate significant improvements with generated designs showing up to 7.67 % increased throughput and 27.63 % reduced total ramp length compared to traditional methods. The generated set contains a high proportion of valid, novel and unique interchanges. These advancements highlight the potential for generative model in creating more efficient and valid interchanges.
{"title":"Objective-directed deep graph generative model for automatic and intelligent highway interchange design","authors":"Chenxiang Ma, Chengcheng Xu","doi":"10.1016/j.autcon.2025.105982","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105982","url":null,"abstract":"Highway interchanges have traditionally been designed through a time-consuming manual process. To enhance efficiency and effectiveness, this paper develops an objective-directed automatic and intelligent interchange design method using graph conditional variational autoencoder. Based on interchange graph representation and augmentation techniques, data are collected from diverse interchanges types and converted into graphs that store design parameters. Aiming at graph reconstruction and fitting data distribution, proposed model learns to generate optimized interchanges by embedding design objectives including throughput and total ramp length. For evaluation, predictors are used to directly output interchange properties, enabling the quick screening of structures. Results demonstrate significant improvements with generated designs showing up to 7.67 % increased throughput and 27.63 % reduced total ramp length compared to traditional methods. The generated set contains a high proportion of valid, novel and unique interchanges. These advancements highlight the potential for generative model in creating more efficient and valid interchanges.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"109 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027402","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}
Pub Date : 2025-01-18DOI: 10.1016/j.autcon.2025.105981
Bao-Luo Li, Jian-Sheng Fan, Jian-Hua Li, Yu-Fei Liu
Deformation detection of grid structures is vital. In complex environments, efficiently identifying locally crooked members among tens of thousands remains a significant challenge. Point cloud-based methods provide dependable solutions for instance segmentation and deformation recognition. However, existing approaches struggle with irrelevant and deficient data, diverse component forms, and low efficiency. This paper introduces non-rigid registration to grid structure scenarios and proposes a semirigid optimal step iterative point cloud registration and segmentation algorithm (SOSIT), specifically designed for grid structures. By leveraging geometric and physical priors, including the as-designed model topology, plane section and finite rotation assumptions, along with differential stiffness and stepwise softening constraints, SOSIT addresses critical challenges in spatial topology organization, transformation matrix representation, and spatially dependent stiffness variation. The algorithm achieves state-of-the-art (SOTA) performance, with a 129-fold increase in efficiency, a 10.1 % improvement in accuracy, and an 81.5 % enhancement in robustness, enabling automated and intelligent deformation inspection and monitoring.
{"title":"Semirigid optimal step iterative algorithm for point cloud registration and segmentation in grid structure deformation detection","authors":"Bao-Luo Li, Jian-Sheng Fan, Jian-Hua Li, Yu-Fei Liu","doi":"10.1016/j.autcon.2025.105981","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105981","url":null,"abstract":"Deformation detection of grid structures is vital. In complex environments, efficiently identifying locally crooked members among tens of thousands remains a significant challenge. Point cloud-based methods provide dependable solutions for instance segmentation and deformation recognition. However, existing approaches struggle with irrelevant and deficient data, diverse component forms, and low efficiency. This paper introduces non-rigid registration to grid structure scenarios and proposes a semirigid optimal step iterative point cloud registration and segmentation algorithm (SOSIT), specifically designed for grid structures. By leveraging geometric and physical priors, including the as-designed model topology, plane section and finite rotation assumptions, along with differential stiffness and stepwise softening constraints, SOSIT addresses critical challenges in spatial topology organization, transformation matrix representation, and spatially dependent stiffness variation. The algorithm achieves state-of-the-art (SOTA) performance, with a 129-fold increase in efficiency, a 10.1 % improvement in accuracy, and an 81.5 % enhancement in robustness, enabling automated and intelligent deformation inspection and monitoring.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"26 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027280","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}
Pub Date : 2025-01-18DOI: 10.1016/j.autcon.2025.105966
Penglu Chen, Wen Yi, Dong Su, Yi Tan, Jinwei Zhou, Xiangsheng Chen
The emergence of digital twins and construction simulation in underground space engineering has driven the demand for efficient Boolean operations on geological models to quickly simulate real-world excavation processes. Therefore, this paper proposes an efficient dynamic Boolean operation framework for large-scale geological models. Firstly, geological models are divided into finite subspace models using spatial bucketing algorithm and efficiently manages spatial triangle data with the R-tree algorithm. Intersecting subspace triangles are then converted into point clouds, and Ball-tree and K-means algorithms are employed to search and remove points, completing the Boolean operation between excavation equipment and geological models. Experiments show that the proposed method achieves a 13-fold speed improvement at 1 cm precision. Furthermore, Boolean operation speeds for point clouds of 10-different scales were analyzed, revealing the relationship between precision and time to meet diverse scenario requirements. The framework exhibits robustness and versatility, making it suitable for large-scale excavation and drilling simulations, including underground spaces and other construction projects.
{"title":"Simulating excavation processes for large-scale underground geological models using dynamic Boolean operations with spatial hash indexing and multiscale point clouds","authors":"Penglu Chen, Wen Yi, Dong Su, Yi Tan, Jinwei Zhou, Xiangsheng Chen","doi":"10.1016/j.autcon.2025.105966","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105966","url":null,"abstract":"The emergence of digital twins and construction simulation in underground space engineering has driven the demand for efficient Boolean operations on geological models to quickly simulate real-world excavation processes. Therefore, this paper proposes an efficient dynamic Boolean operation framework for large-scale geological models. Firstly, geological models are divided into finite subspace models using spatial bucketing algorithm and efficiently manages spatial triangle data with the R-tree algorithm. Intersecting subspace triangles are then converted into point clouds, and Ball-tree and K-means algorithms are employed to search and remove points, completing the Boolean operation between excavation equipment and geological models. Experiments show that the proposed method achieves a 13-fold speed improvement at 1 cm precision. Furthermore, Boolean operation speeds for point clouds of 10-different scales were analyzed, revealing the relationship between precision and time to meet diverse scenario requirements. The framework exhibits robustness and versatility, making it suitable for large-scale excavation and drilling simulations, including underground spaces and other construction projects.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"109 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027282","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}
Pub Date : 2025-01-18DOI: 10.1016/j.autcon.2025.105963
Shaojie Zhou, Jia-Rui Lin, Peng Pan, Yuandong Pan, Ioannis Brilakis
Deep learning (DL)-based point cloud segmentation is essential for understanding built environment. Despite synthetic point clouds (SPC) having the potential to compensate for data shortage, how synthetic color and mixing proportion impact DL-based segmentation remains a long-standing question. Therefore, this paper addresses this question with extensive experiments by introducing: 1) method to generate SPC with real colors and uniform colors from BIM, and 2) enhanced benchmarks for better performance evaluation. Experiments on DL models including PointNet, PointNet++, and DGCNN show that model performance on SPC with real colors outperforms that on SPC with uniform colors by 8.2 % + on both OA and mIoU. Furthermore, a higher than 70 % mixing proportion of SPC usually leads to better performance. And SPC can replace real ones to train a DL model for detecting large and flat building elements. Overall, this paper unveils the performance-improving mechanism of SPC and brings new insights to boost SPC's value.
{"title":"Impact of color and mixing proportion of synthetic point clouds on semantic segmentation","authors":"Shaojie Zhou, Jia-Rui Lin, Peng Pan, Yuandong Pan, Ioannis Brilakis","doi":"10.1016/j.autcon.2025.105963","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105963","url":null,"abstract":"Deep learning (DL)-based point cloud segmentation is essential for understanding built environment. Despite synthetic point clouds (SPC) having the potential to compensate for data shortage, how synthetic color and mixing proportion impact DL-based segmentation remains a long-standing question. Therefore, this paper addresses this question with extensive experiments by introducing: 1) method to generate SPC with real colors and uniform colors from BIM, and 2) enhanced benchmarks for better performance evaluation. Experiments on DL models including PointNet, PointNet++, and DGCNN show that model performance on SPC with real colors outperforms that on SPC with uniform colors by 8.2 % + on both OA and mIoU. Furthermore, a higher than 70 % mixing proportion of SPC usually leads to better performance. And SPC can replace real ones to train a DL model for detecting large and flat building elements. Overall, this paper unveils the performance-improving mechanism of SPC and brings new insights to boost SPC's value.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"58 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027285","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}
Pub Date : 2025-01-18DOI: 10.1016/j.autcon.2025.105965
Sandra Pozzer, Gabriel Ramos, Parham Nooralishahi, Ehsan Rezazadeh Azar, Ahmed El Refai, Fernando López, Clemente Ibarra-Castanedo, Xavier Maldague
This paper presents a framework that integrates passive infrared thermography (IRT) results with building information modeling (BIM) to improve subsurface delamination inspection in concrete infrastructures. The paper combines solar analysis with BIM for better thermography inspection planning and documents thermographic data on delamination within BIM environment using a semi-automatic AI procedure for delamination identification. The framework was tested on two case studies: one laboratory sample with inserted delaminations and one real-scale concrete structure. As a result, besides the increase in communication related to data visualization and centralization, the volume of the delivered data and the analysis time was reduced through the application of the proposed method. The findings promote better collaboration between the civil engineering and inspection industries, utilizing advanced nondestructive techniques for detecting concrete delamination. This approach improves planning, analysis, visualization, reporting, and data storage for passive IRT inspections, benefiting infrastructure stakeholders and streamlining maintenance management.
{"title":"Integration of thermographic inspection data with BIM for enhanced concrete infrastructure assessment","authors":"Sandra Pozzer, Gabriel Ramos, Parham Nooralishahi, Ehsan Rezazadeh Azar, Ahmed El Refai, Fernando López, Clemente Ibarra-Castanedo, Xavier Maldague","doi":"10.1016/j.autcon.2025.105965","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105965","url":null,"abstract":"This paper presents a framework that integrates passive infrared thermography (IRT) results with building information modeling (BIM) to improve subsurface delamination inspection in concrete infrastructures. The paper combines solar analysis with BIM for better thermography inspection planning and documents thermographic data on delamination within BIM environment using a semi-automatic AI procedure for delamination identification. The framework was tested on two case studies: one laboratory sample with inserted delaminations and one real-scale concrete structure. As a result, besides the increase in communication related to data visualization and centralization, the volume of the delivered data and the analysis time was reduced through the application of the proposed method. The findings promote better collaboration between the civil engineering and inspection industries, utilizing advanced nondestructive techniques for detecting concrete delamination. This approach improves planning, analysis, visualization, reporting, and data storage for passive IRT inspections, benefiting infrastructure stakeholders and streamlining maintenance management.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"38 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027283","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}
Artificial intelligence algorithms and multi-sensor technologies are advancing tunnel crack detection. However, image-based detection methods fail to account for tunnel section curvature, limiting their ability to represent the spatial geometry of cracks. To address these problems, this paper presents a tunnel crack assessment method combining simultaneous localization and mapping (SLAM) with deep learning-based segmentation. The SLAM algorithm reconstructs the tunnel point cloud map, and a two-dimensional (2D) convex hull point cloud unfolding with a cloth simulation filter (CSF) algorithm is applied for denoising. A deep learning segmentation model is used to segment the tunnel cracks. The cracks are projected into a three-dimensional (3D) point cloud map, and the crack length and spatial location are calculated. Field tests demonstrate that the method reduces tunnel reconstruction time to 27 s (a 99 % time saving), with a maximum radius error of 0.03 m and accurate 3D crack projections.
{"title":"Tunnel crack assessment using simultaneous localization and mapping (SLAM) and deep learning segmentation","authors":"Huitong Xu, Meng Wang, Cheng Liu, Yongchao Guo, Zihan Gao, Changqing Xie","doi":"10.1016/j.autcon.2025.105977","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105977","url":null,"abstract":"Artificial intelligence algorithms and multi-sensor technologies are advancing tunnel crack detection. However, image-based detection methods fail to account for tunnel section curvature, limiting their ability to represent the spatial geometry of cracks. To address these problems, this paper presents a tunnel crack assessment method combining simultaneous localization and mapping (SLAM) with deep learning-based segmentation. The SLAM algorithm reconstructs the tunnel point cloud map, and a two-dimensional (2D) convex hull point cloud unfolding with a cloth simulation filter (CSF) algorithm is applied for denoising. A deep learning segmentation model is used to segment the tunnel cracks. The cracks are projected into a three-dimensional (3D) point cloud map, and the crack length and spatial location are calculated. Field tests demonstrate that the method reduces tunnel reconstruction time to 27 s (a 99 % time saving), with a maximum radius error of 0.03 m and accurate 3D crack projections.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"2 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027281","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}
Pub Date : 2025-01-18DOI: 10.1016/j.autcon.2025.105983
Jiale Li, Song Zhang, Xuefei Wang
The prediction of pavement deterioration is critical for road maintenance and construction. A thorough understanding of road deterioration mechanisms can enhance the effectiveness of maintenance efforts and prevent further degradation. In this paper, a physics-informed neural network (PINN) was developed to incorporate insights from both big data and the macroscopic deterioration behavior of pavements. A fuzzy partial differential equation (FPDE) was employed as the representative constraint equation based on pavement fatigue cracking theory. Ten years of deterioration data were collected from a selected highway in China to validate the theoretical and practical aspects of the proposed method. The results indicate that the PINN model achieves superior physical consistency, with the prediction accuracy improving by 20.9 % and 11.4 % compared to the BPNN and XGBoost models, respectively. This study introduces a method that aligns data consistency with physical laws and enhances the interpretability of pavement deterioration.
{"title":"Physics-informed neural network with fuzzy partial differential equation for pavement performance prediction","authors":"Jiale Li, Song Zhang, Xuefei Wang","doi":"10.1016/j.autcon.2025.105983","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105983","url":null,"abstract":"The prediction of pavement deterioration is critical for road maintenance and construction. A thorough understanding of road deterioration mechanisms can enhance the effectiveness of maintenance efforts and prevent further degradation. In this paper, a physics-informed neural network (PINN) was developed to incorporate insights from both big data and the macroscopic deterioration behavior of pavements. A fuzzy partial differential equation (FPDE) was employed as the representative constraint equation based on pavement fatigue cracking theory. Ten years of deterioration data were collected from a selected highway in China to validate the theoretical and practical aspects of the proposed method. The results indicate that the PINN model achieves superior physical consistency, with the prediction accuracy improving by 20.9 % and 11.4 % compared to the BPNN and XGBoost models, respectively. This study introduces a method that aligns data consistency with physical laws and enhances the interpretability of pavement deterioration.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"15 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027279","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}
Ground-penetrating radar (GPR) is widely used in airport runway subsurface defect detection. However, variability in subsurface environments and operational frequencies of GPR systems across different airports can cause significant discrepancies in radar data, which influence defect assessments. To address this problem, this study proposes a deep learning algorithm named AD-DetNet, which is designed to maintain robust generalization performance across various airports under different radar frequency conditions. The AD-DetNet model integrates domain-specific knowledge pertinent to detecting subsurface defects in airport runways, which is suitable for various airport environments. In addition, the AD-DetNet model focuses on identifying and emphasizing common characteristics across diverse airports. Moreover, the proposed model incorporates unlabeled target-domain data during training and employs domain adaptation techniques to align features from different data domains. The results of extensive experiments demonstrate that the proposed AD-DetNet model can achieve superior generalization performance across numerous real-world airport datasets and can outperform current state-of-the-art object detection algorithms.
{"title":"Adaptive domain-aware network for airport runway subsurface defect detection","authors":"Haifeng Li, Wenqiang Liu, Nansha Li, Zhongcheng Gui","doi":"10.1016/j.autcon.2025.105969","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105969","url":null,"abstract":"Ground-penetrating radar (GPR) is widely used in airport runway subsurface defect detection. However, variability in subsurface environments and operational frequencies of GPR systems across different airports can cause significant discrepancies in radar data, which influence defect assessments. To address this problem, this study proposes a deep learning algorithm named AD-DetNet, which is designed to maintain robust generalization performance across various airports under different radar frequency conditions. The AD-DetNet model integrates domain-specific knowledge pertinent to detecting subsurface defects in airport runways, which is suitable for various airport environments. In addition, the AD-DetNet model focuses on identifying and emphasizing common characteristics across diverse airports. Moreover, the proposed model incorporates unlabeled target-domain data during training and employs domain adaptation techniques to align features from different data domains. The results of extensive experiments demonstrate that the proposed AD-DetNet model can achieve superior generalization performance across numerous real-world airport datasets and can outperform current state-of-the-art object detection algorithms.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"38 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027297","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}
To mitigate the impact of shield attitude prediction errors on operational decision-making, a framework centered on decomposition and deep learning is proposed to predict multiple shield attitudes. The shield time series data is decomposed into trends and fluctuations by integrating detrended fluctuation analysis and variational mode decomposition. A deep learning model augmented by the multi-head attention mechanism is proposed to independently predict trends and fluctuations. The multi-head attention mechanism enhances the prediction accuracy in simultaneously predicting six attitude parameters. Most prediction errors are allocated to fluctuations through decomposition. The precise prediction of trends provides significant insights into shield attitudes and reduces the risk of misleading outcomes. Compared with existing methods, the proposed method achieves greater precision while requiring fewer inference resources to predict all six attitude parameters. The contribution of multi-head attention and the reason behind prediction error allocation are analyzed via experiments and parameter sensitivity analysis.
{"title":"Prediction of shield machine attitude parameters based on decomposition and multi-head attention mechanism","authors":"Qiushi Wang, Wenqi Ding, Kourosh Khoshelham, Yafei Qiao","doi":"10.1016/j.autcon.2025.105973","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.105973","url":null,"abstract":"To mitigate the impact of shield attitude prediction errors on operational decision-making, a framework centered on decomposition and deep learning is proposed to predict multiple shield attitudes. The shield time series data is decomposed into trends and fluctuations by integrating detrended fluctuation analysis and variational mode decomposition. A deep learning model augmented by the multi-head attention mechanism is proposed to independently predict trends and fluctuations. The multi-head attention mechanism enhances the prediction accuracy in simultaneously predicting six attitude parameters. Most prediction errors are allocated to fluctuations through decomposition. The precise prediction of trends provides significant insights into shield attitudes and reduces the risk of misleading outcomes. Compared with existing methods, the proposed method achieves greater precision while requiring fewer inference resources to predict all six attitude parameters. The contribution of multi-head attention and the reason behind prediction error allocation are analyzed via experiments and parameter sensitivity analysis.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"13 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143027294","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：