Advanced Engineering Informatics最新文献

{"title":"Cross-attention multi-scale state space model for remaining useful life prediction of aircraft engines","authors":"Da Zhang ,&nbsp;Bingyu Li ,&nbsp;Feiyu Wang ,&nbsp;Zhiyuan Zhao ,&nbsp;Junyu Gao ,&nbsp;Xuelong Li","doi":"10.1016/j.aei.2025.103817","DOIUrl":"10.1016/j.aei.2025.103817","url":null,"abstract":"<div><div>Health monitoring and remaining useful life (RUL) prediction of aircraft engines are critical for aviation safety and maintenance decision-making. However, existing methods struggle to fully exploit the nonlinear interactive features across multi-sensor signals, limiting their ability to represent global degradation trends. Additionally, the dynamic interplay mechanisms between long-term macroscopic deterioration and short-term local anomaly patterns remain insufficiently captured, compromising the granular expression of features. To address these challenges, we propose CM-Mamba, a cross-attention multi-scale state space model for RUL prediction. Specifically, we first devise a dual-channel multi-scale patching strategy to separately extract global long-range degradation features and local short-term anomaly patterns. Then, a bidirectional state space model (Mamba) with reverse scanning mechanism is employed to capture global degradation trends across sensors and enhance spatiotemporal correlations. Moreover, windowed self-attention is adopted to refine local sensor degradation details, complemented by a cross-attention mechanism to facilitate global–local feature interactions. After fusing multi-scale features, a fully connected network generates RUL predictions. Experiments based on the C-MAPSS dataset demonstrate that this method significantly improves prediction accuracy under complex conditions and multiple fault modes, validating its superiority in cross-variable correlation modeling and multiscale degradation dynamics analysis.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103817"},"PeriodicalIF":9.9,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049598","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":"Integrated machine learning framework for performance-based seismic assessment of post-tensioned steel-timber hybrid frames with energy-dissipating braces","authors":"Fei Chen ,&nbsp;Zheng Li ,&nbsp;Minghao Li","doi":"10.1016/j.aei.2025.103821","DOIUrl":"10.1016/j.aei.2025.103821","url":null,"abstract":"<div><div>Performance-based seismic assessment (PBSA) is essential for evaluating both structural safety and post-earthquake recovery. While machine learning (ML) has been used to predict seismic responses, its integration into full PBSA workflows—particularly for loss and downtime estimation—remains limited. This study formalizes a streamlined, building-specific ML-PBSA framework and demonstrates its application to post-tensioned steel–timber hybrid (PTSTH) frames with energy-dissipating braces. Seismic input uncertainty was addressed using Latin hypercube sampling and ground motion selection by return period. Feature engineering and ML algorithms were jointly optimized via tree-based Bayesian method. Surrogate models achieved high predictive accuracy (<em>R</em>² = 0.940–0.966) for key response parameters, with a 95th percentile model improving upper-bound prediction of residual inter-story drift. Fragility analysis confirmed the suitability of ML outputs for damage assessment. The ML models were further integrated into the full PBSA process, including seismic loss and downtime estimation. Results closely matched those from nonlinear time history analysis, with prediction errors under 5 % for seismic events with 975-year return periods or longer.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103821"},"PeriodicalIF":9.9,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049595","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":"Ontology-based prompting with large language models for inferring construction activities from construction images","authors":"Cheng Zeng ,&nbsp;Timo Hartmann ,&nbsp;Leyuan Ma","doi":"10.1016/j.aei.2025.103869","DOIUrl":"10.1016/j.aei.2025.103869","url":null,"abstract":"<div><div>Recognizing construction activities from images enhances decision-making by providing context-aware insights into project progress, resource allocation, and productivity. However, conventional approaches, such as supervised learning and knowledge-based approach, struggle to generalize to the dynamic nature of construction sites due to limited annotated data and rigid knowledge patterns. To address these limitations, we propose a novel method that integrates Large Language Models (LLMs) with structured domain knowledge via ontology-based prompting. In our approach, visual features such as entities, spatial arrangements, and actions are mapped to predefined concepts in a construction-specific ontology, resulting in symbolic scene representations. In-context learning is employed by constructing prompts that include multiple structured examples, each describing a scenario with its associated activities. By analyzing these ontology-grounded examples, the LLM learns patterns that connect symbolic representations to construction activity labels, enabling generalization to new, unseen scenes. We evaluated the method using GPT-based models on a dataset covering 29 construction activity types. The model achieved an activity recognition accuracy of 73.68 %, and 50.00 % when jointly identifying the activity and its associated entities. Ablation studies confirmed the positive effects of including Chain-of-Thought reasoning, diverse visual concepts, and richer context examples. These results demonstrate the potential of ontology-informed prompting to support scalable and adaptive visual understanding in construction domains.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103869"},"PeriodicalIF":9.9,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049596","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 twin-driven unsupervised waveform segmentation for bearing quantitative diagnosis","authors":"Xinyu Lu,&nbsp;Zongyang Liu,&nbsp;Hanyang Liu,&nbsp;Jing Lin","doi":"10.1016/j.aei.2025.103833","DOIUrl":"10.1016/j.aei.2025.103833","url":null,"abstract":"<div><div>The quantitative diagnosis of bearing is a prerequisite for informed maintenance decisions, ensuring the high-efficiency operation of modern production facilities. Existing studies utilize dual-impulse extraction-based signal processing techniques or neural network-based intelligent methods for defect size estimation. However, the former is subject to expert knowledge and complicated interferences, while the latter is limited by data resources and black-box attributes. Simulation-based digital twin (DT) technology provides intrinsic mechanism insights and cost-effective data generation. Inspired by this, a DT-driven unsupervised waveform segmentation (DTUWS) method is proposed in this paper to address the above problems. Specifically, a high-fidelity DT model of bearing is first established based on the modeling-update concept of DT technology. The hyper-real observation capability of the DT model is leveraged to generate vibration responses and pixel-level fault semantic labels. Then, the U-Net structure is combined with multi-task learning to construct an unsupervised waveform segmentation model for feature extraction and knowledge transfer. The predicted semantic labels of unlabeled raw field signals are post-processed to derive defect sizes. The diagnosis mechanism of DTUWS is intuitive and interpretable. Experiments on two distinct bench tests demonstrate that DTUWS can achieve accurate and robust quantitative diagnosis without field pre-testing and manual feature extraction.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103833"},"PeriodicalIF":9.9,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049594","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":"Semantic-driven spatial fusion for noise-resilient distance measurement in autonomous inspection of insulators","authors":"Zhikang Yuan ,&nbsp;Junqiu Tang ,&nbsp;Zixiang Wei ,&nbsp;Fei Xie ,&nbsp;Qi Gong ,&nbsp;Shuojie Gao ,&nbsp;Lijun Jin ,&nbsp;Yingyao Zhang","doi":"10.1016/j.aei.2025.103823","DOIUrl":"10.1016/j.aei.2025.103823","url":null,"abstract":"<div><div>Computer vision-based methods have shown great promise in obtaining object distances, significantly improving the efficiency of power distribution system component construction acceptance. However, the complex backgrounds of overhead power lines pose significant challenges to measurement accuracy. To address this, we propose a novel approach that fuses semantic segmentation and spatial reconstruction for noise-resilient distance measurement. The method begins with instance segmentation to generate semantic masks of insulators, followed by binocular vision-based spatial reconstruction. By leveraging depth and density information, DD-Clustereo model is designed to adaptively distinguish valid points from background noise, ensuring precise measurements of the shortest distances between insulators. Experimental results demonstrate that the fusion of semantic and spatial features effectively eliminates background interference, achieving an average error rate of just 2.16%. This work highlights the transformative potential of information fusion in empowering power component inspection through machine vision.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103823"},"PeriodicalIF":9.9,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049597","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":"Graph approach for Gibson’s ecological optics with dynamics of network motifs","authors":"Gi-bbeum Lee,&nbsp;Ji-Hyun Lee","doi":"10.1016/j.aei.2025.103865","DOIUrl":"10.1016/j.aei.2025.103865","url":null,"abstract":"<div><div>Dynamic visual perception in complex environments is central to understanding the interaction between organisms and their surroundings. Ecological optics depicts that the visual system gains optical information from ambient light, which is structured by relative movements between organism and environment. Recent advances have developed theoretical models of optical information, commonly formalized as optical flows, that account for the perception–action link. However, these frameworks have had limited capacity to inform environmental design, due to a gap between the micro-scale, formalized models of optical information and meso-scale, semantic analyses of observer experience. To address this gap, building on basic principles of ecological optics, we develop a framework that characterizes observers’ perception–action patterns by integrating graph-theoretic concepts and measures. Our framework discretizes spatial and temporal trajectories of ambient light experienced by an observer, in the form of a weighted directed graph. This graph approach directly reveals dynamics of perception–action patterns via network motifs—recurring subgraph patterns within a larger graph. Information entropy, as a temporal measure of information content, indicates the distinct modes of the dynamics. As a demonstration, a state analysis shows that several transient states in the motif-based dynamics exhibit good correlations with observers’ inclination toward places from survey data, validating its potential for spatial analysis. Overall, the proposed framework paves the way towards real-world applications in optimizing dynamic interactions between observer and environment.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103865"},"PeriodicalIF":9.9,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145027227","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":"Structural evaluation of cracked shield tunnels using computer-vision-based model updating techniques","authors":"Xiangyu Chang ,&nbsp;Youqi Zhang ,&nbsp;Chengjia Han ,&nbsp;Yuguang Fu ,&nbsp;Jianxiao Mao ,&nbsp;Hao Wang","doi":"10.1016/j.aei.2025.103818","DOIUrl":"10.1016/j.aei.2025.103818","url":null,"abstract":"<div><div>Accurate and efficient assessment of structural damage in shield tunnels is essential for ensuring the safety and reliability of transportation systems. Cracks in tunnel linings are common, necessitating regular structural integrity assessments to ensure safety. Traditional modeling of such damage is often complex and time-consuming. Therefore, the objective of this study is to automate the entire process from detecting tunnel damage in images to conducting numerical analyses for shield tunnels, thereby enabling rapid assessment of structural integrity. We propose a segment-based method that updates a finite element (FE) model of shield tunnels to reflect geometric changes due to cracks, utilizing computer vision (CV) techniques and geometric analyses. Firstly, the Segment Anything Model, along with CV techniques, is used to identify the shapes and sizes of tunnel components from full and partial tunnel segment images. Then, a Dual VMamba U-Net (DVMamba-UNet) is proposed to identify cracks and provide detailed crack information, i.e., crack masks. Finally, geometric analysis is employed to develop algorithms that automatically transform coordinates and select elements within FE models, facilitating the update of geometric changes. Residual capability assessments of updated FE models are used to evaluate the structural damage and the tunnel segment condition. Two case studies are conducted to verify the effectiveness of the proposed approach and algorithms. The results show that the proposed method allows for automatic updates to the FE tunnel model based on damage detected in images through CV techniques and geometric analyses. Additionally, updated FE tunnel models representing different damage levels are developed and analyzed using numerical simulations. This approach not only proves effective in evaluating structural damage in shield tunnels but also offers potential as a data processing and model updating modules within future Digital Twin frameworks for tunnel infrastructure.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103818"},"PeriodicalIF":9.9,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145027224","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":"An efficient machine learning-based model for duration prediction of construction tasks with large-scale datasets","authors":"Yaping Liu,&nbsp;Huan Luo","doi":"10.1016/j.aei.2025.103820","DOIUrl":"10.1016/j.aei.2025.103820","url":null,"abstract":"<div><div>Infrastructure project delays increasingly cause substantial economic losses and operation risks, yet conventional methods for construction schedule risk analysis remain reliant on subjective empirical judgments. While machine learning (ML) methods have mitigated some limitations, prevailing approaches focus on macro-level predictions using small-sample datasets, largely neglecting textual data in construction tasks. To address these issues, this study proposes KLWLS-SVMR, a novel ML model that integrates textual and numerical features to predict construction task durations. The proposed model quantifies unstructured task descriptions through topic modeling, constructs optimal feature sets via a PCA-random forest hybrid mechanism, and integrates k-means clustering with locally weighted support vector regression to enhance prediction accuracy and computational efficiency for large-scale datasets. The superior performance of proposed method is demonstrated by comparison with 11 popularly used ML methods based on a dataset covering 140,378 real-world tasks. Compared to the best-performing benchmarks, Extremely Randomized Trees (<span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>=</mo><mn>0</mn><mo>.</mo><mn>968</mn></mrow></math></span>) and Adaptive Boosting (<span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>=</mo><mn>0</mn><mo>.</mo><mn>960</mn></mrow></math></span>), the proposed model achieves a higher <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> value of 0.973 while reducing computational time by 25.0% and 79.7%, respectively. Compared to the original feature set, the KLWLS-SVMR trained with optimal feature sets formulated by the proposed hybrid mechanism shows an 11.3% increase in <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span>, 53.5% and 51.8% decrease in RMSE and MAE, respectively, while significantly improving computational efficiency by 40.0%. Rigorous hypothesis testing confirms that all ML models trained with the optimal feature sets exhibit statistical significance (<span><math><mrow><mi>p</mi><mo>=</mo><mn>0</mn><mo>.</mo><mn>006</mn><mo>≪</mo><mn>0</mn><mo>.</mo><mn>05</mn></mrow></math></span>) for prediction performance improvement. This work advances ML applications in construction engineering by providing a practical technical pathway for optimizing task-level resource scheduling and risk management.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103820"},"PeriodicalIF":9.9,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145027225","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":"Optimal reliability and efficiency design with experimental validation of a locking mechanism for morphing wings using data-driven surrogate models","authors":"Jiwei Yao ,&nbsp;Shiqiu Gong ,&nbsp;Fei Ji ,&nbsp;Yifan Wang ,&nbsp;Jing Zhao","doi":"10.1016/j.aei.2025.103831","DOIUrl":"10.1016/j.aei.2025.103831","url":null,"abstract":"<div><div>To satisfy the engineering demand for millisecond-level reliable locking and holding during the fold-to-deploy process of morphing wings in high-speed aircraft, this study proposes a novel spring-taper-pin wedge-locking mechanism (STPWLM). The STPWLM is optimized to enhance locking reliability and efficiency and validated through physical experiments. At supply pressures of 0.4 MPa, 0.6 MPa, and 0.8 MPa, the optimized prototype reduces link rebound oscillations by 100 %, 100 %, and 66.7 %, and shortens locking duration by 100 %, 100 %, and 56.98 %, respectively, compared with the unoptimized prototype. Specifically, a mechanical structure model is established according to the functional requirements and working principle. Through engineering expertise, 11 primary structural parameters are subsequently identified. An analytical model of the force transmission and motion properties of the STPWLM is established to identify four key structural parameters, thereby clarifying the design optimization guidelines. The locking performance metric is defined as a composite measure that combines the number of link rebound oscillations and the locking duration, jointly reflecting reliability and efficiency. To address the difficulty of explicitly calculating this metric, a transient dynamic finite element simulation model is constructed, and comparative simulations are used to verify the design guidelines. Optimal Latin hypercube design (OLHD) coupled with data-driven surrogate models is employed to predict the locking performance metric, significantly reducing the computational cost of simulations. The design of experiments (DOE) method is utilized as a decision support tool for sensitivity analysis of the main and interaction effects of the core structural parameters, taper pin positioning distance and taper angle, thus providing a solid basis for optimal design. A multi-island genetic algorithm (MIGA) is applied to determine the optimal combination of the core structural parameters. Morphing wing prototypes before and after optimization are fabricated, and a series of comparative physical experiments are conducted. Experimental results confirm the accuracy of the analytical model and demonstrate the superiority of the optimal design. These findings provide valuable guidance for the engineering design and application of morphing wings with an integrated locking mechanism.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103831"},"PeriodicalIF":9.9,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049593","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":"FMANet: Fused mamba attention model with multi-type preprocessing for simulated crack-contaminated complex environments","authors":"Junwen Zheng ,&nbsp;Houxin Lv ,&nbsp;Hangtian Song ,&nbsp;Jiang Li ,&nbsp;Rongrong Bai ,&nbsp;Lingkun Chen ,&nbsp;Qizhi Chen ,&nbsp;Lizhong Jiang","doi":"10.1016/j.aei.2025.103808","DOIUrl":"10.1016/j.aei.2025.103808","url":null,"abstract":"<div><div>This paper proposes FMANet (Fusion Mamba Attention Network), a crack segmentation network that integrates Mamba modules and hybrid attention mechanisms, and combines a set of data preprocessing methods that integrate multiple simulated real environment interferences. FMANet enhances segmentation accuracy and anti-interference capabilities with a visual state spatial model and parallel hybrid attention module. The data preparation creates a Chameleon Crack Dataset (CCD) of various cracks using color transformation, Berlin noise, and Gaussian noise/blurring. The experimental findings demonstrate that FMANet obtains 87.38% F1-score and 79.68% mIoU on the CCD test set, which surpasses the other comparison models. The ablation experiment shows the Mamba module’s contribution to the model’s 36.89% improvement in PI value. This work offers an effective way to gather fracture data and automatically segment fractures in complicated settings.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"69 ","pages":"Article 103808"},"PeriodicalIF":9.9,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049731","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}