Structural Health Monitoring最新文献

{"title":"Damage identification of steel bridge based on data augmentation and adaptive optimization neural network","authors":"Minshui Huang, Jianwei Zhang, Jun Li, Zhihang Deng, Jin Luo","doi":"10.1177/14759217241255042","DOIUrl":"https://doi.org/10.1177/14759217241255042","url":null,"abstract":"With the advancement of deep learning, data-driven structural damage identification (SDI) has shown considerable development. However, collecting vibration signals related to structural damage poses certain challenges, which can undermine the accuracy of the identification results produced by data-driven SDI methods in scenarios where data is scarce. This paper introduces an innovative approach to bridge SDI in a few-shot context by integrating an adaptive simulated annealing particle swarm optimization-convolutional neural network (ASAPSO-CNN) as the foundational framework, augmented by data enhancement techniques. Firstly, three specific types of noise are introduced to augment the source signals used for training. Subsequently, the source signals and augmented signals are recombined to construct a four-dimensional matrix as the input to the CNN, while defining the damage feature vector as the output. Secondly, a CNN is constructed to establish the mapping relationship between the input and output. Then, an adaptive fitness function is proposed that simultaneously considers the accuracy of SDI, model complexity, and training efficiency. The ASAPSO is employed to adaptively optimize the hyperparameters of the CNN. The proposed method is validated on an experimental model of a three-span continuous beam. It is compared with four other data-driven methods, demonstrating good effectiveness and robustness of SDI under cases of scarce data. Finally, the effectiveness of this SDI method is validated in a real-world case of a steel truss bridge.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"57 5","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141805920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adaptive time-domain impact extraction method for multi-source impact vibration signal of diesel engine","authors":"Nanyang Zhao, Chao Liu, Dongxiang Jiang, Jinjie Zhang, Zhinong Jiang","doi":"10.1177/14759217241264927","DOIUrl":"https://doi.org/10.1177/14759217241264927","url":null,"abstract":"Diesel engines are widely used in fields such as ships, vehicles, and nuclear power. The vibration signals of a diesel engine’s casing exhibit a characteristic of intermittent distribution of multi-source impacts. In response to the challenges faced by existing feature extraction methods in identifying and localizing impact signals, this paper proposes the adaptive time-domain impact extraction (ATDIE) method, which is based on the characteristics of impact signals exhibiting local high-energy distribution in the time domain and rapid amplitude decay. The purpose of the ATDIE method is to extract various impact components from multi-source impact signals. The ATDIE method constructs a solution model with the goal of minimizing the signal’s multi-order amplitude central moments. By using the residual iterative solution method, the number of extracted components is adaptively determined. Then, an impact window optimization function is established to enhance the adaptability. Finally, the test results with both simulation signals and diesel engine signals demonstrate that the ATDIE method possesses good capabilities for impact extraction and computational efficiency.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"3 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141803033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent segmentation and quantification of tunnel lining cracks via computer vision","authors":"Yong Feng, Xiao-Lei Zhang, Shi-Jin Feng, Wei Zhang, Kan Hu, Yue-Wu Da","doi":"10.1177/14759217241254748","DOIUrl":"https://doi.org/10.1177/14759217241254748","url":null,"abstract":"Aiming to automatically, precisely, and rapidly detect tunnel lining cracks from images and extract geometric information for structural condition assessment, this study proposes a novel tunnel lining crack segmentation network (TCSegNet) and establishes a framework for calculating key geometric parameters of cracks. A tunnel lining crack segmentation dataset is first built by conducting on-site inspections of metro tunnels and collecting open-sourced tunnel images. Afterward, the TCSegNet, conforming to the encoder–decoder architectural paradigm, is designed to separate cracks from lining images pixel-to-pixel. An improved ConvNeXt and developed efficient atrous spatial pyramid pooling module constitute the encoder. The skip connections, upsampling modules, and tailored segmentation head form the decoder. Upon the segmentation results of TCSegNet, a computing framework integrating multiple digital image processing techniques is proposed to obtain the length, average width, and maximum width of cracks. The experimental results show that the TCSegNet achieves leading results among several dominant models, with 70.78% mean intersection over union (mIoU) and 57.43% F1 score. Furthermore, the TCSegNet has 32.01 million parameters, requires 55.13 billion floating point operations, and gets 107.28 frames per second, proving that it has low time and space complexities and implements real-time segmentation. Also, the rationality and effectiveness of TCSegNet in alleviating the crack disjoint problem and preserving crack edge details are verified through comparative experiments. In addition, the TCSegNet achieves 71.99%, 70.45%, and 70.23% mIoU in high-resolution image segmentation, robustness, and generalization tests, respectively, demonstrating that it is competent for detecting high-resolution lining images, has a solid resistance to illumination variations, and can be well generalized to other tunnel lining image datasets. Finally, the applicability of the crack quantification framework is validated by practical application examples. The developed approaches in this study provide pixel-level segmentation results and detailed measurements of concrete lining cracks to assess tunnel structural safety status.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"74 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141802288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Segment anything model-based crack segmentation using low-rank adaption fine-tuning","authors":"Yapeng Guo, Yang Xu, Hongtao Cui, Minghao Dang, Shunlong Li","doi":"10.1177/14759217241261089","DOIUrl":"https://doi.org/10.1177/14759217241261089","url":null,"abstract":"High-precision crack segmentation is crucial for analyzing and maintaining the apparent state of structures. The introduction of large vision models, such as the segment anything model (SAM), has brought significant advancements in object segmentation due to their remarkable generalization capabilities. However, SAM cannot be directly used for the purpose of automatic crack segmentation. This study introduces a novel approach that fine-tunes SAM specifically for crack segmentation by incorporating low-rank adaptation (LoRA). This method involves adding a dedicated crack segmentation head to SAM, enabling automatic crack segmentation. Additionally, the application of LoRA technology facilitates the readjustment of SAM’s features without incurring the substantial costs typically associated with fine-tuning entire networks. A comparative analysis with current leading crack segmentation models demonstrated a significant increase in accuracy across eight different crack datasets. This study offers guidelines for the application of large vision models for crack identification.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"21 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141803120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A semisupervised fault frequency analysis method for rotating machinery based on restricted self-attention network","authors":"Huaqin Zhang, Jichao Hong, Haixu Yang, Xinyang Zhang, Fengwei Liang, Chi Zhang, Zhongguo Huang","doi":"10.1177/14759217241262956","DOIUrl":"https://doi.org/10.1177/14759217241262956","url":null,"abstract":"With the development of informatization and digitalization, condition monitoring has been applied to industrial equipment such as rotating machinery. Collecting and storing large amounts of equipment operating data enable the detection of mechanical equipment faults using historical operational data. This article proposes a semisupervised data-driven approach to analyze the fault frequencies of rotating machinery. Frequency band information and the degree of association with faults are obtained through the variance of attention values. To address the inherent issue of decoupling information between data segments in deep learning, restrictive layers are proposed. These layers prevent the flow of information between data segments from rendering interpretable information ineffective. Bearing and gearbox datasets are used to validate the proposed method. The fault frequencies extracted by this method correspond to actual faults. The preferred deep learning framework achieves an accuracy exceeding 99% on both datasets. The method is compared with various signal processing methods and identifies fault frequencies that are difficult to identify using traditional methods. Furthermore, the unreliability of traditional deep learning in fault diagnosis is also exposed. In this study, semisupervised deep learning fault frequency extraction is achieved for the first time.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"32 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141806137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An explainable variational autoencoder model for three-dimensional acoustic emission source localization in hollow cylindrical structures","authors":"Guan-Wei Lee, S. Livadiotis, S. Salamone","doi":"10.1177/14759217241260254","DOIUrl":"https://doi.org/10.1177/14759217241260254","url":null,"abstract":"We introduce an explainable variational autoencoder for three-dimensional (3D) localization of acoustic emission sources in hollow cylindrical structures, with an unsupervised approach. This research capitalizes on multi-arrival waveforms generated by helical path propagation in cylindrical geometries to enable efficient two-receiver localization. By integrating the modal characteristics of Lamb modes under multi-path conditions, we demonstrate that two sets of time-of-arrival differences and peak amplitudes extracted from one receiver can serve as effective localization features. This initial approach identifies four potential source locations, highlighting the feasibility of two-receiver source localization using traditional feature extraction methods. However, direct extraction can be challenging when mode overlaps occur, complicating the localization process. To address this, our work proposes a novel waveform-based method. This method leverages the consistent dispersion characteristics within isotropic materials, where each unique combination of mode arrival times and peak amplitudes constructs a distinct waveform. This distinctiveness overcomes the ambiguities associated with mode overlaps, significantly enhancing the method’s precision and robustness. Our approach adopts a data-driven strategy for waveform-based localization using variational autoencoder (VAE). VAE discerns waveform patterns for localization, while also addressing data uncertainties. The VAE’s encoder and decoder networks capture the localization process and the source’s influence on waveform generation, respectively, guiding latent variables to segregate waveforms by source in the latent space. The design of the learning process focuses on specific localization characteristics to enhance result explainability. Localization predictions are generated by projecting test waveforms, not included in the training set, onto a trained latent space. The prediction is determined using a nearest-neighbor approach based on the closest latent representation of a source. Validation with pencil-lead-break tests on a metallic pipe confirmed our method’s effectiveness, achieving an averaged 3D localization accuracy of 0.84.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"7 2","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141803359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep hybrid neural network-aided electromechanical impedance method for automated damage detection of lining concrete under freeze-thaw cycling","authors":"Chuan Zhang, Q. Yan, Xiaolong Liao, Yunhui Qiu, Yifeng Zhang, Ping Wang","doi":"10.1177/14759217241259955","DOIUrl":"https://doi.org/10.1177/14759217241259955","url":null,"abstract":"Cold regional tunnels extensively suffer from severe damage in concrete linings under cyclic freeze-thaw environment. Therefore, accurate detection and evaluation of cyclic freeze-thaw damage within lining concrete is of great significance to help grasp structural health state and guarantee timely maintenance. This study pioneered the application of electromechanical impedance (EMI) method to monitor the freeze-thaw damage in bended concrete beams. The mass loss and flexural strength degradation of concrete beams under two different bending loads were thoroughly assessed to quantify the evolution of cyclic freeze-thaw damage. Moreover, the conductance signatures driven by d31 and d33 modes were analyzed, respectively. It was found that the variation in the d31 mode-dominated signal well agreed with the progressive damage characterized by the flexural strength degradation. The key innovation of this study is that a deep hybrid neural network DenseNet–GRU was constructed and well trained to predict the cyclic freeze-thaw damage from augmented EMI data. The results indicated that the proposed model achieved excellent performance with determination coefficients exceeding 0.997 for both bending scenarios. Additionally, DenseNet–GRU outperformed conventional baseline machine or deep learning models in prediction accuracy and noise-resistance capacity. Notably, it demonstrated good adaptability when trained with limited data samples. In summary, the proposed methodology enabled automated detection and accurate forecasting of the cyclic freeze-thaw damage in lining concrete without hand-crafted features.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"44 18","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141805375","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multiscale ECA network: a rotation mechanical domain adaptation method with minimal class confusion","authors":"Xueyi Li, Tianyu Yu, Kaiyu Su, Peng Yuan, Zhijie Xie","doi":"10.1177/14759217241261155","DOIUrl":"https://doi.org/10.1177/14759217241261155","url":null,"abstract":"Unsupervised rotation mechanical fault diagnosis methods have become popular, but existing unsupervised methods still have some issues. For example, it is challenging to capture vibration signal features at different scales and to address partial class confusion. To improve the diagnostic performance, this study introduces a multiscale, efficient channel attention (ECA) attention mechanism, joint adaptation network (JAN), and minimum class confusion (MCC) for addressing the aforementioned issues. First, the authors design a multiscale fault feature extraction module to capture discriminative information at different scales in vibration signals. Second, the authors introduce the ECA mechanism to weight the extracted features at the channel level, enhancing useful features and suppressing redundant features. Then, the authors employ the JAN method to establish local maximum mean discrepancy, enabling adaptation between corresponding subdomains of the source and target domains, avoiding the problem of being too close. Finally, the authors use MCC as the loss function to reduce prediction confusion between correct and ambiguous categories in target samples, thus improving transfer performance. Experimental results demonstrate that the proposed method exhibits excellent performance in unsupervised rotation mechanical fault diagnosis tasks.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"73 6","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141802684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel intelligent multicross domain fault diagnosis of servo motor-bearing system based on Domain Generalized Graph Convolution Autoencoder","authors":"Xiaoli Zhao, Yuanhao Hu, Jiahui Liu, Jianyong Yao, W. Deng, Jian Hu, Zhuanzhe Zhao, Xiaoan Yan","doi":"10.1177/14759217241262722","DOIUrl":"https://doi.org/10.1177/14759217241262722","url":null,"abstract":"The data measured by the servo motor-bearing system under complex working conditions will present problems such as amplitude fluctuations, unequal impact intervals, and significant differences in data distribution, and so forth. However, the most intelligent fault diagnosis focus on deep learning or transfer learning, which cannot complement knowledge transfer and generalized diagnosis with the structural neighbor relationship under unknown conditions or cross-machine samples. By utilizing Domain Generalized Graph Convolution Autoencoder (DGGCAE), a novel intelligent multicross domain fault diagnosis method for servo-motor bearing systems can be developed. Specifically, the Dirichlet Mixup and Distilled augmentations are first employed to augment the domain data of the feature and label layer for model training. Accordingly, graph representation learning on multisource domain data is mainly performed for the developed algorithm. Afterward, the graph convolutional autoencoder is employed to extract enough generalized high-dimensional features. Furthermore, DGGCAE’s classification loss and domain discrimination loss can be calculated to narrow the distribution gap among multisource domains. Finally, the fault simulation test bench (called servo motor-Cylindrical roller bearing system from Nanjing University of Science and Technology) has validated the development of the diagnostic method.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"51 47","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141805040","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Damage identification of a jacket platform based on a hybrid deep learning framework","authors":"Su Xin, Zhang Qi, Li Yang, Huang Yi, Ziguang Jia","doi":"10.1177/14759217241262558","DOIUrl":"https://doi.org/10.1177/14759217241262558","url":null,"abstract":"Given the complex operational environment of offshore platforms, accurate identification of structural damage has become a crucial aspect of structural health monitoring. However, accurately pinpointing the damage locations based on vibration data under load, particularly for intricate platform structures, is a challenging task. Existing damage-identification methods, particularly those rooted in deep learning frameworks, often encounter difficulties when applied to marine platforms. Therefore, this study proposes an innovative approach. The accuracy of damage identification for marine platforms operating under unique service conditions was enhanced by introducing a deconvolutional parallel processing module and an auxiliary loss function processing module into the core ResNet50 network. This enhancement improved the accuracy of the model in detecting damage within complex marine structures. Information processing is enriched by fusing the vibration data acquired from the measurement points across different domains: time, frequency, and recurrence plots. The results of this approach were remarkable. When the algorithm model, validated through model experiments, is extended to a digital twin established based on real marine platforms, simulations and loading under real loads were performed on a refined high-fidelity finite-element model, yielding dynamic response information that closely mirrored real-world conditions. A corresponding damage-recognition database was established to support the digital twin system. For the eight different directions, the model accuracy ranged from a minimum of 87.38% to a maximum of 92.27%. This represents a significant advancement compared to the performance of the original network. Empirical experiments substantiated the efficacy of the improved algorithm, demonstrating an impressive recognition accuracy of 93.75%. This achievement underscores the potential of this method to revolutionize damage identification for marine platforms, particularly under the distinctive conditions that these structures encounter. The integration of specialized modules and enhanced processing methodologies further bolster the accuracy of deep-learning-based damage identification and makes the building of digital twin models of offshore platforms feasible.","PeriodicalId":515545,"journal":{"name":"Structural Health Monitoring","volume":"39 24","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141805887","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}