Measurement最新文献_第3页

Comprehensive explainable model using low-frequency vibration characteristics for leakage detection in water pipelines 基于低频振动特性的水管道泄漏检测综合可解释模型

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120701

Yulong Yang , Shuhui Zhou , Jie Chen , Zheyu Huang , Xiaojun Ding , Jun Qian , Kang Wang

Although low-frequency vibration analysis can potentially be used to detect leakage in water pipelines, its practical application and performance are underexplored. Herein, a feature space describing the physical signatures of leakage and enabling its identification in cast iron, steel, and polyethylene (PE) pipelines is constructed, and the effectiveness of utilizing the low-frequency (0–300 Hz) characteristics of leakage-induced vibrations is validated using laboratory-scale pipeline data. A feature extraction method based on these low-frequency characteristics is proposed, and five types of machine learning models are used to achieve recognition accuracies of 97.26%–99.32%. The developed method is shown to outperform the two-dimensional convolution neural network (2D-CNN) model through the comparison of features extracted using both approaches. Interpretable feature analysis is performed using the Shapley additive explanation (SHAP) method, confirming the suitability of using low-frequency vibrations for leakage detection. The number of features is reduced from 19 to 7 features via SHAP-based feature importance analysis, and the model with the highest accuracy (stacking model) is selected to validate the optimized feature space. The established method is applied to cast iron, steel, and PE pipes and shown to be suitable for detecting leaks therein. Finally, by comparing model accuracy and SHAP analysis results under conditions with and without pump interference, the robustness and stability of the proposed method were validated.

虽然低频振动分析可以潜在地用于检测输水管道的泄漏，但其实际应用和性能尚未得到充分的探索。本文构建了一个描述泄漏物理特征的特征空间，并使其能够在铸铁、钢和聚乙烯（PE）管道中进行识别，并使用实验室规模的管道数据验证了利用低频（0-300 Hz）泄漏诱发振动特征的有效性。提出了一种基于这些低频特征的特征提取方法，并利用5种机器学习模型实现了97.26% ~ 99.32%的识别准确率。通过对两种方法提取的特征进行比较，表明该方法优于二维卷积神经网络（2D-CNN）模型。使用Shapley加性解释（SHAP）方法进行可解释特征分析，确认使用低频振动进行泄漏检测的适用性。通过基于shap的特征重要性分析，将特征数量从19个减少到7个，并选择精度最高的模型（叠加模型）对优化后的特征空间进行验证。所建立的方法适用于铸铁、钢管和聚乙烯管，并证明适用于检测其中的泄漏。最后，通过对比存在和不存在泵干扰情况下的模型精度和SHAP分析结果，验证了所提方法的鲁棒性和稳定性。

{"title":"Comprehensive explainable model using low-frequency vibration characteristics for leakage detection in water pipelines","authors":"Yulong Yang , Shuhui Zhou , Jie Chen , Zheyu Huang , Xiaojun Ding , Jun Qian , Kang Wang","doi":"10.1016/j.measurement.2026.120701","DOIUrl":"10.1016/j.measurement.2026.120701","url":null,"abstract":"<div><div>Although low-frequency vibration analysis can potentially be used to detect leakage in water pipelines, its practical application and performance are underexplored. Herein, a feature space describing the physical signatures of leakage and enabling its identification in cast iron, steel, and polyethylene (PE) pipelines is constructed, and the effectiveness of utilizing the low-frequency (0–300 Hz) characteristics of leakage-induced vibrations is validated using laboratory-scale pipeline data. A feature extraction method based on these low-frequency characteristics is proposed, and five types of machine learning models are used to achieve recognition accuracies of 97.26%–99.32%. The developed method is shown to outperform the two-dimensional convolution neural network (2D-CNN) model through the comparison of features extracted using both approaches. Interpretable feature analysis is performed using the Shapley additive explanation (SHAP) method, confirming the suitability of using low-frequency vibrations for leakage detection. The number of features is reduced from 19 to 7 features via SHAP-based feature importance analysis, and the model with the highest accuracy (stacking model) is selected to validate the optimized feature space. The established method is applied to cast iron, steel, and PE pipes and shown to be suitable for detecting leaks therein. Finally, by comparing model accuracy and SHAP analysis results under conditions with and without pump interference, the robustness and stability of the proposed method were validated.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"269 ","pages":"Article 120701"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146147328","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust subspace intersection method for source bearing estimation in uncertain shallow water 不确定浅水源方位估计的鲁棒子空间交会方法

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120704

Jingrong Wu, Chao Sun, Mingyang Li

Traditional bearing estimation methods assuming plane wave propagation suffer significant measurement errors due to multi-modal propagation in ocean waveguides, particularly for sources near the endfire of a horizontal line array (HLA). The HLA signal wavefront is a linear combination of modal steering vectors, characterized by modal horizontal wavenumbers, spanning the modal subspace. The subspace intersection (SI) method matches this signal subspace with replica modal subspaces for measuring source bearing but requires precise wavenumber knowledge. Uncertainty in environmental parameters renders wavenumbers unknown, causing SI performance degradation. In this work, we propose a robust SI method for uncertain shallow water. We suggest to incorporate the value range of modal horizontal wavenumbers resulting from the uncertain environment to construct a high-dimensional modal subspace. By using an optimization criterion derived from minimizing global bearing errors, we present a simple metric to truncate redundant dimensions to achieve the effective modal span. The resulting processor, termed the effective-modal-subspace-based SI (EM-SI), exhibits better performance than the SI. Simulations in a benchmark uncertain waveguide confirm that the EM-SI outperforms the SI and traditional methods in bearing measurement errors and resolution. Real data from the SWellEx96 sea trial further demonstrates its effectiveness.

传统的假设平面波传播的方位估计方法由于海洋波导中的多模态传播而存在显著的测量误差，特别是对于水平线阵列（HLA）末端附近的波源。HLA信号波前是模态转向矢量的线性组合，以模态水平波数为特征，跨越模态子空间。子空间交集（SI）方法将该信号子空间与复制模态子空间相匹配以测量源方位，但需要精确的波数知识。环境参数的不确定性使波数未知，导致SI性能下降。在这项工作中，我们提出了一种不确定浅水的鲁棒SI方法。我们建议将不确定环境引起的模态水平波数的取值范围纳入到高维模态子空间中。利用最小化全局方位误差的优化准则，提出了一种截断冗余尺寸以实现有效模态跨度的简单度量。由此产生的处理器，称为基于有效模态子空间的SI (EM-SI)，表现出比SI更好的性能。在基准不确定波导中进行的仿真验证了EM-SI在轴承测量误差和分辨率方面优于SI和传统方法。welllex96海上试验的真实数据进一步证明了其有效性。

{"title":"Robust subspace intersection method for source bearing estimation in uncertain shallow water","authors":"Jingrong Wu, Chao Sun, Mingyang Li","doi":"10.1016/j.measurement.2026.120704","DOIUrl":"10.1016/j.measurement.2026.120704","url":null,"abstract":"<div><div>Traditional bearing estimation methods assuming plane wave propagation suffer significant measurement errors due to multi-modal propagation in ocean waveguides, particularly for sources near the endfire of a horizontal line array (HLA). The HLA signal wavefront is a linear combination of modal steering vectors, characterized by modal horizontal wavenumbers, spanning the modal subspace. The subspace intersection (SI) method matches this signal subspace with replica modal subspaces for measuring source bearing but requires precise wavenumber knowledge. Uncertainty in environmental parameters renders wavenumbers unknown, causing SI performance degradation. In this work, we propose a robust SI method for uncertain shallow water. We suggest to incorporate the value range of modal horizontal wavenumbers resulting from the uncertain environment to construct a high-dimensional modal subspace. By using an optimization criterion derived from minimizing global bearing errors, we present a simple metric to truncate redundant dimensions to achieve the effective modal span. The resulting processor, termed the effective-modal-subspace-based SI (EM-SI), exhibits better performance than the SI. Simulations in a benchmark uncertain waveguide confirm that the EM-SI outperforms the SI and traditional methods in bearing measurement errors and resolution. Real data from the SWellEx96 sea trial further demonstrates its effectiveness.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"269 ","pages":"Article 120704"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146147313","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A normalized spectral test for assessing non-flat spectrum near a given frequency 在给定频率附近评估非平坦谱的一种归一化谱测试

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120598

Aluizio D’Affonsêca Netto , Tiago Zanotelli , Carlos J. Tierra-Criollo , Leonardo B. Felix , David M. Simpson , Antonio M.F.L. Miranda de Sá

A new statistical test is proposed for detecting isolated spectral peaks in regions with flat spectra. The need for such tests typically arises in hearing tests based on the analysis of EEG (electroencephalography) signals with auditory steady state responses. Commonly, a local F-test is used as a statistical technique in these cases. The current work proposes a modification of this method, which bounds the parameter to the range 0 to 1. The aim is to facilitate monitoring the strength of responses by keeping values bounded to facilitate their interpretation. The probability density function of the new detector was derived for both the cases of lack and presence of evoked responses, which allowed obtaining the confidence limits and the probability of detection. Simulation studies were carried out to confirm the theoretical results, and the technique was applied to the electroencephalogram during auditory stimulation by amplitude-modulated tones. The proposed technique has detection power similar to other local detectors. However, it has the advantage of making the detector time-evolution visualization easier since it is always bounded between zero and one.

提出了一种新的统计检验方法，用于检测平坦光谱区域的孤立光谱峰。这种测试的需求通常出现在基于对具有听觉稳态反应的脑电图（EEG）信号分析的听力测试中。在这些情况下，通常使用局部f检验作为统计技术。目前的工作提出了对该方法的修改，将参数限定在0到1的范围内。其目的是通过保持数值的界限以方便对其进行解释，从而便于监测响应的强度。推导了该检测器在缺乏和存在诱发反应情况下的概率密度函数，从而获得了置信度限和检测概率。为了验证理论结果，进行了仿真研究，并将该技术应用于调幅音听觉刺激时的脑电图。该方法具有与其他局部检测器相似的检测能力。然而，它的优点是使探测器的时间演化可视化更容易，因为它总是在0和1之间有界。

{"title":"A normalized spectral test for assessing non-flat spectrum near a given frequency","authors":"Aluizio D’Affonsêca Netto , Tiago Zanotelli , Carlos J. Tierra-Criollo , Leonardo B. Felix , David M. Simpson , Antonio M.F.L. Miranda de Sá","doi":"10.1016/j.measurement.2026.120598","DOIUrl":"10.1016/j.measurement.2026.120598","url":null,"abstract":"<div><div>A new statistical test is proposed for detecting isolated spectral peaks in regions with flat spectra. The need for such tests typically arises in hearing tests based on the analysis of EEG (electroencephalography) signals with auditory steady state responses. Commonly, a local F-test is used as a statistical technique in these cases. The current work proposes a modification of this method, which bounds the parameter to the range 0 to 1. The aim is to facilitate monitoring the strength of responses by keeping values bounded to facilitate their interpretation. The probability density function of the new detector was derived for both the cases of lack and presence of evoked responses, which allowed obtaining the confidence limits and the probability of detection. Simulation studies were carried out to confirm the theoretical results, and the technique was applied to the electroencephalogram during auditory stimulation by amplitude-modulated tones. The proposed technique has detection power similar to other local detectors. However, it has the advantage of making the detector time-evolution visualization easier since it is always bounded between zero and one.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"268 ","pages":"Article 120598"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146171926","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Low-light contrast enhancement based on weighted-preprocessing polarization distance model 基于加权预处理偏振距离模型的弱光对比度增强

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120723

Xin Wang , Junfeng Xu , Xiyun Zeng , Jun Gao

In addressing challenges related to low contrast and high noise stemming from low illumination, this study aims to enhance target detection in low-light environments. To achieve this, we preprocess captured low-light polarization images, leveraging a combination of the four-channel polarization distance model and the HSI color space. We introduce a weighted preprocessing polarization distance (WPDI) model, which incorporates weighted fusion preprocessing across different expressions of each component of the Stokes vector. Additionally, we utilize polarization angle information to establish threshold ranges, merging polarization distance information with original light intensity data to create a new intensity channel. This channel is then fused with the original tone and saturation channels to produce the final WPDI model mapping result. The experimental findings indicate that, in comparison with three existing target contrast enhancement models, the model proposed in this paper yields markedly enhanced qualitative and quantitative outcomes. Moreover, it effectively enhances the distinction between the target and the background in low-light conditions.

为了解决低照度导致的低对比度和高噪声问题，本研究旨在增强低照度环境下的目标检测。为了实现这一目标，我们利用四通道偏振距离模型和HSI色彩空间的组合，对捕获的低光偏振图像进行预处理。引入加权预处理极化距离（WPDI）模型，对Stokes矢量各分量的不同表达式进行加权融合预处理。此外，我们利用偏振角信息建立阈值范围，将偏振距离信息与原始光强数据合并，形成新的光强通道。然后将该通道与原始色调和饱和度通道融合以产生最终的WPDI模型映射结果。实验结果表明，与现有的三种目标对比度增强模型相比，本文模型的定性和定量结果均有显著提高。在弱光条件下，有效地增强了目标与背景的区分。

{"title":"Low-light contrast enhancement based on weighted-preprocessing polarization distance model","authors":"Xin Wang , Junfeng Xu , Xiyun Zeng , Jun Gao","doi":"10.1016/j.measurement.2026.120723","DOIUrl":"10.1016/j.measurement.2026.120723","url":null,"abstract":"<div><div>In addressing challenges related to low contrast and high noise stemming from low illumination, this study aims to enhance target detection in low-light environments. To achieve this, we preprocess captured low-light polarization images, leveraging a combination of the four-channel polarization distance model and the HSI color space. We introduce a weighted preprocessing polarization distance (WPDI) model, which incorporates weighted fusion preprocessing across different expressions of each component of the Stokes vector. Additionally, we utilize polarization angle information to establish threshold ranges, merging polarization distance information with original light intensity data to create a new intensity channel. This channel is then fused with the original tone and saturation channels to produce the final WPDI model mapping result. The experimental findings indicate that, in comparison with three existing target contrast enhancement models, the model proposed in this paper yields markedly enhanced qualitative and quantitative outcomes. Moreover, it effectively enhances the distinction between the target and the background in low-light conditions.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"268 ","pages":"Article 120723"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Nanoparticle-doped fibers combined with machine learning methods for 3-dimensional shape reconstruction systems 纳米粒子掺杂纤维与机器学习相结合的三维形状重建系统

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120733

Arnaldo Leal-Junior , Leandro Macedo , Jan Nedoma , Radek Martinek , Wilfried Blanc

The development of nanoparticle (NP) doped optical fibers provides new possibilities in distributed sensing and provides important advantages in spatially distributed system due to the tailoring of the backscattering signal, which enables the development of new approaches in shape reconstruction systems. For this reason, this paper presents the application of a distributed sensor system based on optical frequency domain reflectometry (OFDR) using the NP-doped optical fiber as the sensing medium in which there is an increase in sensitivity and spatial resolution due to the increase in backscaterred signal. The fiber is embedded in a rubber strip to increase the measurement range and the sensor robustness due to the nitrile rubber flexibility, since the rubber can provide strain transfer to the optical fiber, leading to higher strain limits, which can withstand the complex 3D loadings for the shape reconstruction. The sensor system is tested under different mechanical loadings conditions and the contribution of this work is a readily available sensing system using specialty optical fiber with high spatial resolution for 3D shape sensing approaches. The results indicate the feasibility of the proposed approach for shape sensing monitoring through a determination coefficient (

R^{2}

) higher than 0.99 for all cases. In addition, the strain distribution along the fiber was also estimated for all loading conditions as well as the combination of them with an RMSE of around 1 mm considering the amplitude estimation of all loadings. The 3D shape reconstruction resulted in a mean error of around 2.0% considering all 3 axes of the cartesian plane. Therefore, the proposed sensor system is a feasible option for shape sensing with sub-centimeter spatial resolution and high accuracy using the NP-doped optical fiber as enhanced backscattered medium.

纳米粒子（NP）掺杂光纤的发展为分布式传感提供了新的可能性，并且由于后向散射信号的剪裁，在空间分布式系统中提供了重要的优势，这使得形状重建系统的新方法的发展成为可能。为此，本文提出了一种基于光频域反射（OFDR）的分布式传感器系统的应用，该系统采用掺杂np的光纤作为传感介质，由于后向散射信号的增加，灵敏度和空间分辨率都有所提高。由于丁腈橡胶的灵活性，将光纤嵌入橡胶条中以增加测量范围和传感器的鲁棒性，因为橡胶可以向光纤提供应变传递，从而产生更高的应变极限，可以承受复杂的3D载荷进行形状重建。传感器系统在不同的机械载荷条件下进行了测试，这项工作的贡献是使用具有高空间分辨率的专用光纤用于3D形状传感方法的易于获得的传感系统。结果表明，该方法在所有情况下的决定系数（R2）均大于0.99，是一种可行的形状传感监测方法。此外，还估计了所有加载条件下沿纤维的应变分布以及它们的组合，考虑到所有加载的振幅估计，RMSE约为1 mm。考虑到直角平面的所有3个轴，三维形状重建的平均误差约为2.0%。因此，利用掺np光纤作为增强后向散射介质，该传感器系统是实现亚厘米空间分辨率和高精度形状传感的可行选择。

{"title":"Nanoparticle-doped fibers combined with machine learning methods for 3-dimensional shape reconstruction systems","authors":"Arnaldo Leal-Junior , Leandro Macedo , Jan Nedoma , Radek Martinek , Wilfried Blanc","doi":"10.1016/j.measurement.2026.120733","DOIUrl":"10.1016/j.measurement.2026.120733","url":null,"abstract":"<div><div>The development of nanoparticle (NP) doped optical fibers provides new possibilities in distributed sensing and provides important advantages in spatially distributed system due to the tailoring of the backscattering signal, which enables the development of new approaches in shape reconstruction systems. For this reason, this paper presents the application of a distributed sensor system based on optical frequency domain reflectometry (OFDR) using the NP-doped optical fiber as the sensing medium in which there is an increase in sensitivity and spatial resolution due to the increase in backscaterred signal. The fiber is embedded in a rubber strip to increase the measurement range and the sensor robustness due to the nitrile rubber flexibility, since the rubber can provide strain transfer to the optical fiber, leading to higher strain limits, which can withstand the complex 3D loadings for the shape reconstruction. The sensor system is tested under different mechanical loadings conditions and the contribution of this work is a readily available sensing system using specialty optical fiber with high spatial resolution for 3D shape sensing approaches. The results indicate the feasibility of the proposed approach for shape sensing monitoring through a determination coefficient (<span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span>) higher than 0.99 for all cases. In addition, the strain distribution along the fiber was also estimated for all loading conditions as well as the combination of them with an RMSE of around 1 mm considering the amplitude estimation of all loadings. The 3D shape reconstruction resulted in a mean error of around 2.0% considering all 3 axes of the cartesian plane. Therefore, the proposed sensor system is a feasible option for shape sensing with sub-centimeter spatial resolution and high accuracy using the NP-doped optical fiber as enhanced backscattered medium.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"268 ","pages":"Article 120733"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multiple tapping method for non-destructive testing of defects in metal components 金属构件缺陷无损检测的多次攻丝法

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120751

Dima Bykhovsky , Shalev Neuman , Oz Golan , Moshe Dror Kobo , Yael Ashkenaz , Michael Zolotih , Strokin Evgeny , Shai Essel , Oshrit Hoffer

Non-destructive testing (NDT) of additively manufactured (AM) metal components typically relies on costly imaging or ultrasonic systems. We introduce a low-cost mechanical tapping device combined with a machine learning (ML)-based acoustic measurement workflow, and demonstrate its superiority as a measurement system compared to standard fundamental frequency analysis approaches. We treat the classification pipeline as a calibrated “soft sensor” that outputs a defect probability. While maintaining a very simple mechanical tapping system, we hypothesize that introducing Type A measurement uncertainty and fusing multiple probabilistic outputs significantly improves decision accuracy. Furthermore, we demonstrate that varying the tap location constitutes a distinct measurement modality, offering validation beyond the benefits accrued from mere repetition. Using a dataset of 80 Ti–6Al–4V specimens with defects validated by computed tomography (CT), we experimentally show that the proposed multiple mechanical tapping fusion method significantly reduces the probability of error.

增材制造（AM）金属部件的无损检测（NDT）通常依赖于昂贵的成像或超声波系统。我们介绍了一种低成本的机械敲击装置，结合了基于机器学习（ML）的声学测量工作流程，并证明了其作为测量系统与标准基频分析方法相比的优势。我们将分类管道视为一个输出缺陷概率的校准“软传感器”。在维持一个非常简单的机械攻丝系统的同时，我们假设引入a型测量不确定度和融合多个概率输出可以显著提高决策精度。此外，我们证明了改变水龙头位置构成了一种独特的测量方式，提供了超越单纯重复所带来的好处的验证。利用80个带有缺陷的Ti-6Al-4V样品的计算机断层扫描（CT）验证数据集，我们实验表明，所提出的多次机械攻丝融合方法显著降低了误差概率。

{"title":"Multiple tapping method for non-destructive testing of defects in metal components","authors":"Dima Bykhovsky , Shalev Neuman , Oz Golan , Moshe Dror Kobo , Yael Ashkenaz , Michael Zolotih , Strokin Evgeny , Shai Essel , Oshrit Hoffer","doi":"10.1016/j.measurement.2026.120751","DOIUrl":"10.1016/j.measurement.2026.120751","url":null,"abstract":"<div><div>Non-destructive testing (NDT) of additively manufactured (AM) metal components typically relies on costly imaging or ultrasonic systems. We introduce a low-cost mechanical tapping device combined with a machine learning (ML)-based acoustic measurement workflow, and demonstrate its superiority as a measurement system compared to standard fundamental frequency analysis approaches. We treat the classification pipeline as a calibrated “soft sensor” that outputs a defect probability. While maintaining a very simple mechanical tapping system, we hypothesize that introducing Type A measurement uncertainty and fusing multiple probabilistic outputs significantly improves decision accuracy. Furthermore, we demonstrate that varying the tap location constitutes a distinct measurement modality, offering validation beyond the benefits accrued from mere repetition. Using a dataset of 80 Ti–6Al–4V specimens with defects validated by computed tomography (CT), we experimentally show that the proposed multiple mechanical tapping fusion method significantly reduces the probability of error.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"269 ","pages":"Article 120751"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146147305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ADAMNet: Improving imbalanced defect classification with Anomaly-Driven Attention Maps 用异常驱动的注意图改进不平衡缺陷分类

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120537

Jun-Hui Liang , Y.S. Gan , Sze-Teng Liong , Shih-Yuan Wang , Yu-Ting Sheng , Lit-Ken Tan

In response to the increasing demand for high-quality leather products and the limitations of manual visual inspections, there is a growing need for automated and accurate defect classification systems. This paper presents ADAMNet (Anomaly-Driven Attention Map Network), a novel deep learning framework designed to enhance multi-class leather defect classification under data imbalance conditions. ADAMNet integrates anomaly detection with attention-based classification by leveraging a modified PatchCore backbone. Specifically, pretrained Wide ResNet-50 features with adaptive pooling are used to generate fine-grained anomaly maps that highlight defective regions. These anomaly maps are then incorporated into a CNN as attention masks, guiding the model’s focus toward informative areas and improving classification performance. To evaluate the effectiveness of the proposed method, extensive experiments were conducted on a real-world leather defect dataset comprising 375 images across three categories: non-defective, wrinkle defects, and black line defects. ADAMNet achieved over 90% classification accuracy on highly imbalanced data and outperformed several baseline models by a margin of 15.2%. Furthermore, we conducted ablation studies and both quantitative and qualitative analyses to validate the role of anomaly-guided attention in improving model performance. Overall, ADAMNet provides a flexible, interpretable, and effective solution for fine-grained defect recognition, demonstrating strong potential for deployment in industrial quality control workflows and broader defect inspection tasks

由于对高质量皮革产品的需求不断增加，以及人工目视检查的局限性，对自动化和准确的缺陷分类系统的需求日益增长。本文提出了一种新的深度学习框架ADAMNet (Anomaly-Driven Attention Map Network)，旨在增强数据不平衡条件下多类皮革缺陷的分类能力。ADAMNet集成异常检测与基于注意力的分类，利用改进的PatchCore主干。具体来说，使用预训练的宽ResNet-50特征和自适应池来生成细粒度异常图，突出缺陷区域。然后将这些异常图作为注意力掩模合并到CNN中，引导模型关注信息区域并提高分类性能。为了评估所提出方法的有效性，在一个真实的皮革缺陷数据集上进行了大量的实验，该数据集包含375张图像，分为三大类：无缺陷、皱纹缺陷和黑线缺陷。ADAMNet在高度不平衡的数据上实现了超过90%的分类准确率，并且比几个基线模型高出15.2%。此外，我们进行了消融研究，并进行了定量和定性分析，以验证异常引导注意力在提高模型性能方面的作用。总之，ADAMNet为细粒度缺陷识别提供了灵活的、可解释的和有效的解决方案，展示了在工业质量控制工作流程和更广泛的缺陷检查任务中部署的强大潜力

{"title":"ADAMNet: Improving imbalanced defect classification with Anomaly-Driven Attention Maps","authors":"Jun-Hui Liang , Y.S. Gan , Sze-Teng Liong , Shih-Yuan Wang , Yu-Ting Sheng , Lit-Ken Tan","doi":"10.1016/j.measurement.2026.120537","DOIUrl":"10.1016/j.measurement.2026.120537","url":null,"abstract":"<div><div>In response to the increasing demand for high-quality leather products and the limitations of manual visual inspections, there is a growing need for automated and accurate defect classification systems. This paper presents ADAMNet (Anomaly-Driven Attention Map Network), a novel deep learning framework designed to enhance multi-class leather defect classification under data imbalance conditions. ADAMNet integrates anomaly detection with attention-based classification by leveraging a modified PatchCore backbone. Specifically, pretrained Wide ResNet-50 features with adaptive pooling are used to generate fine-grained anomaly maps that highlight defective regions. These anomaly maps are then incorporated into a CNN as attention masks, guiding the model’s focus toward informative areas and improving classification performance. To evaluate the effectiveness of the proposed method, extensive experiments were conducted on a real-world leather defect dataset comprising 375 images across three categories: non-defective, wrinkle defects, and black line defects. ADAMNet achieved over 90% classification accuracy on highly imbalanced data and outperformed several baseline models by a margin of 15.2%. Furthermore, we conducted ablation studies and both quantitative and qualitative analyses to validate the role of anomaly-guided attention in improving model performance. Overall, ADAMNet provides a flexible, interpretable, and effective solution for fine-grained defect recognition, demonstrating strong potential for deployment in industrial quality control workflows and broader defect inspection tasks</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"268 ","pages":"Article 120537"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146171996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An IoT based novel data logger and controller and normalized gain calculator for gaseous detector 一种基于物联网的新型数据记录仪和控制器以及气体检测器的归一化增益计算器

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-05 DOI: 10.1016/j.measurement.2026.120693

Shradhha Suman Panigrahi , Akankshya Nayak , Sanjib Kumar Sahu , P.K. Sahu

Maintaining a clean and controlled environment in the Detector Laboratory is critical for optimal performance of gaseous and solid-state detectors. An IoT based environmental data logger and controller has a wide functional advantage over the traditional data loggers. The system described in this paper allows for real-time monitoring of these parameters as well as tracking particle concentrations, including mass and number concentrations for various particles sizes. Additionally, it controls the air-conditioner, dehumidifier, etc. using an IoT Cloud based concept, which is based on an operating micro-controller system over WiFi. The local controls and data logging of different parameters are based on the LabVIEW interface. The data acquisition circuits enable data collection for experiments such as gaseous detectors as well as solid-state detectors. This advancement provides a robust comprehensive and remotely accessible solution for laboratory environmental monitoring. The developed DAQ system is not only suitable for environmental monitoring but also for determining the temperature (T) and pressure (p) - dependent gain (G) of GEM detectors. For demonstration purpose, a quad-GEM detector has been used to perform an experiment and T/p and the corresponding G(T/p) is calculated in real-time using the proposed DAQ system, in an attempt to implement the T/p compensation in the detector gain. At an ambient temperature of

T = 294.462 K

and a pressure of

p = 1010.150 mbar

(

0.996 a t m

), the corresponding

T / p

ratio is

295.644 K/atm

. Under these conditions, the quad-GEM detector exhibited a gain of (

4.326 \pm 0.067

)

\times 1 0^{4}

, as estimated by the DAQ software, with a particulate matter concentration of

P M 2.5 = 3.6 μ {g/cm}^{3}

.

在探测器实验室中保持清洁和受控的环境对于气体和固态探测器的最佳性能至关重要。基于物联网的环境数据记录仪和控制器与传统数据记录仪相比具有广泛的功能优势。本文中描述的系统允许实时监测这些参数以及跟踪颗粒浓度，包括各种颗粒尺寸的质量和数量浓度。此外，它还使用基于物联网云的概念来控制空调，除湿机等，该概念基于WiFi上的操作微控制器系统。不同参数的本地控制和数据记录是基于LabVIEW接口实现的。数据采集电路使数据采集实验，如气体探测器以及固态探测器。这一进步为实验室环境监测提供了一个强大的全面和远程访问的解决方案。所开发的DAQ系统不仅适用于环境监测，而且适用于测定GEM探测器的温度(T)和压力(p)相关增益(G)。为了演示目的，使用四极板探测器进行实验，并使用所提出的DAQ系统实时计算T/p和相应的G(T/p)，试图在探测器增益中实现T/p补偿。在环境温度T=294.462K，压力p=1010.150mbar （0.996atm）时，对应的T/p比值为295.644K/atm。在此条件下，经DAQ软件估算，四极板探测器的增益为（4.326±0.067）×104, PM2.5颗粒物浓度为3.6μg/cm3。

{"title":"An IoT based novel data logger and controller and normalized gain calculator for gaseous detector","authors":"Shradhha Suman Panigrahi , Akankshya Nayak , Sanjib Kumar Sahu , P.K. Sahu","doi":"10.1016/j.measurement.2026.120693","DOIUrl":"10.1016/j.measurement.2026.120693","url":null,"abstract":"<div><div>Maintaining a clean and controlled environment in the Detector Laboratory is critical for optimal performance of gaseous and solid-state detectors. An IoT based environmental data logger and controller has a wide functional advantage over the traditional data loggers. The system described in this paper allows for real-time monitoring of these parameters as well as tracking particle concentrations, including mass and number concentrations for various particles sizes. Additionally, it controls the air-conditioner, dehumidifier, etc. using an IoT Cloud based concept, which is based on an operating micro-controller system over WiFi. The local controls and data logging of different parameters are based on the LabVIEW interface. The data acquisition circuits enable data collection for experiments such as gaseous detectors as well as solid-state detectors. This advancement provides a robust comprehensive and remotely accessible solution for laboratory environmental monitoring. The developed DAQ system is not only suitable for environmental monitoring but also for determining the temperature (T) and pressure (p) - dependent gain (G) of GEM detectors. For demonstration purpose, a quad-GEM detector has been used to perform an experiment and T/p and the corresponding G(T/p) is calculated in real-time using the proposed DAQ system, in an attempt to implement the T/p compensation in the detector gain. At an ambient temperature of <span><math><mrow><mi>T</mi><mo>=</mo><mn>294</mn><mo>.</mo><mn>462</mn><mspace></mspace><mi>K</mi></mrow></math></span> and a pressure of <span><math><mrow><mi>p</mi><mo>=</mo><mn>1010</mn><mo>.</mo><mn>150</mn><mspace></mspace><mi>mbar</mi></mrow></math></span> (<span><math><mrow><mn>0</mn><mo>.</mo><mn>996</mn><mspace></mspace><mi>a</mi><mi>t</mi><mi>m</mi></mrow></math></span>), the corresponding <span><math><mrow><mi>T</mi><mo>/</mo><mi>p</mi></mrow></math></span> ratio is <span><math><mrow><mn>295</mn><mo>.</mo><mn>644</mn><mspace></mspace><mi>K/atm</mi></mrow></math></span>. Under these conditions, the quad-GEM detector exhibited a gain of (<span><math><mrow><mn>4</mn><mo>.</mo><mn>326</mn><mo>±</mo><mn>0</mn><mo>.</mo><mn>067</mn></mrow></math></span>) <span><math><mrow><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mn>4</mn></mrow></msup></mrow></math></span>, as estimated by the DAQ software, with a particulate matter concentration of <span><math><mrow><mi>P</mi><mi>M</mi><mn>2</mn><mo>.</mo><mn>5</mn><mo>=</mo><mn>3</mn><mo>.</mo><mn>6</mn><mspace></mspace><mi>μ</mi><msup><mrow><mi>g/cm</mi></mrow><mrow><mn>3</mn></mrow></msup></mrow></math></span>.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"268 ","pages":"Article 120693"},"PeriodicalIF":5.6,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172135","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An uncertainty-weighted fusion method of PS-InSAR, GNSS, and leveling for millimeter-level deformation monitoring in sensitive observatory environments 敏感观测环境下PS-InSAR、GNSS和水准测量的不确定加权融合方法

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-04 DOI: 10.1016/j.measurement.2026.120660

Xiangrui Kong, Xilong Wang, Ziyi Chen, Mingwei Huang

Achieving millimeter-level measurement accuracy at high-sensitivity seismic observation stations is extremely challenging, especially when urban engineering construction is carried out in their vicinity, causing non-tectonic ground deformation. Such complex environments expose the limitations of traditional multi-sensor deformation monitoring methods, which are specifically manifested in insufficient precision calibration, lack of registration of a unified reference framework, and insufficient in-depth analysis of the attribution of mechanical mechanisms. These problems restrict the reliable separation of deformation signals caused by tectonic and non-tectonic causes. In response to the above problems, this study proposes a multi-sensor fusion framework based on uncertainty weighting. The framework integrates three observation methods, namely Permanent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR), Global Navigation Satellite System (GNSS), and high-precision leveling, and combines the three-dimensional irregular load model to carry out the attribution analysis of the deformation mechanism. The results show that in the application practice of the Jinzhou seismic station from 2021 to 2024, the framework controls the deviation of results between different monitoring technologies within ± 0.5 mm, clarifies that more than 90% of the peak settlement is caused by engineering construction activities, and realizes the effective separation of deformation signals caused by tectonic and non-tectonic causes.

在高灵敏度地震观测站实现毫米级的测量精度是极具挑战性的，特别是当城市工程建设在其附近进行时，会造成非构造地面变形。这种复杂的环境暴露了传统的多传感器变形监测方法的局限性，具体表现在精度标定不够、缺乏统一参考框架的配准、对力学机制属性分析不够深入等方面。这些问题限制了构造和非构造引起的变形信号的可靠分离。针对上述问题，本研究提出了一种基于不确定性加权的多传感器融合框架。该框架整合了永久散射体干涉合成孔径雷达（PS-InSAR）、全球卫星导航系统（GNSS）和高精度平准三种观测方法，结合三维不规则载荷模型对变形机理进行归因分析。结果表明：在2021 - 2024年锦州地震台站的应用实践中，该框架将不同监测技术之间的结果偏差控制在±0.5 mm以内，明确了90%以上的峰值沉降是由工程建设活动引起的，实现了构造与非构造原因引起的变形信号的有效分离。

{"title":"An uncertainty-weighted fusion method of PS-InSAR, GNSS, and leveling for millimeter-level deformation monitoring in sensitive observatory environments","authors":"Xiangrui Kong, Xilong Wang, Ziyi Chen, Mingwei Huang","doi":"10.1016/j.measurement.2026.120660","DOIUrl":"10.1016/j.measurement.2026.120660","url":null,"abstract":"<div><div>Achieving millimeter-level measurement accuracy at high-sensitivity seismic observation stations is extremely challenging, especially when urban engineering construction is carried out in their vicinity, causing non-tectonic ground deformation. Such complex environments expose the limitations of traditional multi-sensor deformation monitoring methods, which are specifically manifested in insufficient precision calibration, lack of registration of a unified reference framework, and insufficient in-depth analysis of the attribution of mechanical mechanisms. These problems restrict the reliable separation of deformation signals caused by tectonic and non-tectonic causes. In response to the above problems, this study proposes a multi-sensor fusion framework based on uncertainty weighting. The framework integrates three observation methods, namely Permanent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR), Global Navigation Satellite System (GNSS), and high-precision leveling, and combines the three-dimensional irregular load model to carry out the attribution analysis of the deformation mechanism. The results show that in the application practice of the Jinzhou seismic station from 2021 to 2024, the framework controls the deviation of results between different monitoring technologies within ± 0.5 mm, clarifies that more than 90% of the peak settlement is caused by engineering construction activities, and realizes the effective separation of deformation signals caused by tectonic and non-tectonic causes.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"268 ","pages":"Article 120660"},"PeriodicalIF":5.6,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172140","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Measurement-based temperature field characteristics of CRTS II slab ballastless track structure in construction period 基于实测的CRTS II型平板无砟轨道结构施工期间温度场特征

IF 5.6 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Measurement

Pub Date : 2026-02-04 DOI: 10.1016/j.measurement.2026.120681

Xiaodan Sun , Yasen Zhang , Yu Liu , Zheying Ji , Yuting Dai , An He , Yang Xu

This study addresses the lack of data on the early temperature field of ballastless track structures during the construction period. The CRTS II slab ballastless track on the Beijing-Shanghai High-speed Railway was investigated. A 72-hour on-site temperature monitoring following the completion of CA mortar grouting was conducted for the first time. Three conditions were compared: Summer-Sunny, Summer-Rainfall, and Autumn-Sunny. The spatiotemporal evolution of the early temperature field was revealed. The results show that: the early temperature field of the ballastless track exhibits pronounced spatial non-uniformity. The non-uniformity of both the temperature and the temperature gradient at of the track slab is the highest in Summer-Sunny condition. The lag time of temperature variations on the top and bottom surfaces of the track slab is prolonged, and the positive temperature gradient of the track slab is reduced by the hydration heat of CA mortar. However, the negative temperature gradient of the track slab is significantly increased, and its duration is extended. During the construction period, the temperature gradient of the track slab during the nighttime period should be given attention. The results also show that construction under Summer-Rainfall condition has significant advantages. Compared with other sunny conditions, both the temperature gradient and its uniformity coefficient are notably low.

本研究解决了无砟轨道结构施工初期温度场数据不足的问题。对京沪高速铁路CRTSⅱ型平板无砟轨道进行了研究。在CA砂浆灌浆完成后，首次进行了72小时的现场温度监测。比较了三种条件：夏季-晴天、夏季-降雨和秋季-晴天。揭示了早期温度场的时空演化过程。结果表明：无砟轨道初期温度场表现出明显的空间非均匀性。在夏季-日照条件下，轨道板温度和温度梯度的不均匀性最大。CA砂浆的水化热延长了轨道板上下表面温度变化的滞后时间，减小了轨道板的正温度梯度。但轨道板负温度梯度明显增大，持续时间延长。施工期间，应注意夜间轨道板的温度梯度。结果还表明，夏季降雨条件下的施工具有显著的优势。与其他阳光条件相比，温度梯度和均匀度系数都明显较低。

{"title":"Measurement-based temperature field characteristics of CRTS II slab ballastless track structure in construction period","authors":"Xiaodan Sun , Yasen Zhang , Yu Liu , Zheying Ji , Yuting Dai , An He , Yang Xu","doi":"10.1016/j.measurement.2026.120681","DOIUrl":"10.1016/j.measurement.2026.120681","url":null,"abstract":"<div><div>This study addresses the lack of data on the early temperature field of ballastless track structures during the construction period. The CRTS II slab ballastless track on the Beijing-Shanghai High-speed Railway was investigated. A 72-hour on-site temperature monitoring following the completion of CA mortar grouting was conducted for the first time. Three conditions were compared: Summer-Sunny, Summer-Rainfall, and Autumn-Sunny. The spatiotemporal evolution of the early temperature field was revealed. The results show that: the early temperature field of the ballastless track exhibits pronounced spatial non-uniformity. The non-uniformity of both the temperature and the temperature gradient at of the track slab is the highest in Summer-Sunny condition. The lag time of temperature variations on the top and bottom surfaces of the track slab is prolonged, and the positive temperature gradient of the track slab is reduced by the hydration heat of CA mortar. However, the negative temperature gradient of the track slab is significantly increased, and its duration is extended. During the construction period, the temperature gradient of the track slab during the nighttime period should be given attention. The results also show that construction under Summer-Rainfall condition has significant advantages. Compared with other sunny conditions, both the temperature gradient and its uniformity coefficient are notably low.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"269 ","pages":"Article 120681"},"PeriodicalIF":5.6,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146147311","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0