Pub Date : 2026-01-14DOI: 10.1016/j.measurement.2026.120473
Shahab Faiz Minhas , Maqsood Hussain Shah
In this paper, we address the challenge of improving the accuracy of mine detection systems in highly mineralized soils, which has severe impact on detection accuracy. Mineral-rich soils not only limit detection depth but also obscure mines with low conductive metal content. We introduce novel algorithms that employ adaptive filters to effectively mitigate the impact of soil mineralization, enabling the precise detection of metal content at sub-gram levels. The proposed adaptive filters utilize supervised learning with stochastic gradient descent to dynamically learn and counteract the soil’s mineralization response, achieving 98% accuracy in removing mineralization interference. The effectiveness of proposed adaptive filter-based compensation (AFC) algorithms is validated in practical mineralized environments, demonstrating robust performance with real-time parameters and enabling accurate mine detection where conventional systems fail.
{"title":"Adaptive filtering technique for mitigating soil mineralization effects in pulse-induced metallic mine detectors","authors":"Shahab Faiz Minhas , Maqsood Hussain Shah","doi":"10.1016/j.measurement.2026.120473","DOIUrl":"10.1016/j.measurement.2026.120473","url":null,"abstract":"<div><div>In this paper, we address the challenge of improving the accuracy of mine detection systems in highly mineralized soils, which has severe impact on detection accuracy. Mineral-rich soils not only limit detection depth but also obscure mines with low conductive metal content. We introduce novel algorithms that employ adaptive filters to effectively mitigate the impact of soil mineralization, enabling the precise detection of metal content at sub-gram levels. The proposed adaptive filters utilize supervised learning with stochastic gradient descent to dynamically learn and counteract the soil’s mineralization response, achieving <span><math><mo>></mo></math></span>98% accuracy in removing mineralization interference. The effectiveness of proposed adaptive filter-based compensation (AFC) algorithms is validated in practical mineralized environments, demonstrating robust performance with real-time parameters and enabling accurate mine detection where conventional systems fail.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120473"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036277","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}
Pub Date : 2026-01-14DOI: 10.1016/j.measurement.2026.120478
Jiaxuan Li , Dongkai Dai , Ying Yu , Liheng Ma , Shiqiao Qin
LEO satellites exhibit high angular velocities relative to Earth-based observers, causing the satellite motion to spread optical signals across multiple pixels, which leads to an overestimated brightness using the apparent magnitude model. To address this challenge, this paper proposes the refined effective magnitude model (REMM) and the brightest-pixel grayscale model, both based on a Gaussian point spread function, which accurately quantify variations in satellites’ photometric characteristics under different angular velocities and exposure times. The proposed models have been validated through numerical simulations, controlled laboratory experiments using a star simulator, and field observations of Starlink satellites and OneWeb satellites, and have shown superior performance compared to conventional apparent magnitude models. Results demonstrate that the proposed models effectively capture the photometric attenuation behavior, with the derived analytical expression for trailing-induced signal variation showing strong agreement across validation phases. The proposed framework clearly delineates the influence of angular velocity, detection conditions, and exposure time on satellite photometric performance, providing a reliable theoretical foundation for high-precision photometry and characterizing space objects.
{"title":"Refined modeling of effective visible magnitudes for optical observations of low Earth orbit satellites","authors":"Jiaxuan Li , Dongkai Dai , Ying Yu , Liheng Ma , Shiqiao Qin","doi":"10.1016/j.measurement.2026.120478","DOIUrl":"10.1016/j.measurement.2026.120478","url":null,"abstract":"<div><div>LEO satellites exhibit high angular velocities relative to Earth-based observers, causing the satellite motion to spread optical signals across multiple pixels, which leads to an overestimated brightness using the apparent magnitude model. To address this challenge, this paper proposes the refined effective magnitude model (REMM) and the brightest-pixel grayscale model, both based on a Gaussian point spread function, which accurately quantify variations in satellites’ photometric characteristics under different angular velocities and exposure times. The proposed models have been validated through numerical simulations, controlled laboratory experiments using a star simulator, and field observations of Starlink satellites and OneWeb satellites, and have shown superior performance compared to conventional apparent magnitude models. Results demonstrate that the proposed models effectively capture the photometric attenuation behavior, with the derived analytical expression for trailing-induced signal variation showing strong agreement across validation phases. The proposed framework clearly delineates the influence of angular velocity, detection conditions, and exposure time on satellite photometric performance, providing a reliable theoretical foundation for high-precision photometry and characterizing space objects.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120478"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036546","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}
Pub Date : 2026-01-14DOI: 10.1016/j.measurement.2026.120368
Bumsoo Park , Julius Mauch , Hyeokjin Kweon , Jochen Kriegseis , Seungchul Lee , Hyoungsoo Kim
Multicomponent droplet evaporation generates inherently three-dimensional, solutal-Marangoni flows that challenge single-camera velocimetry. We present STAR-APTV (Segmentation and Tracking Anything-based Robust Astigmatic Particle Tracking Velocimetry), a zero-shot, deep-learning-assisted, astigmatic particle tracking framework for time-resolved 3D-3C flow reconstruction with minimal optical hardware. We leverage zero-shot segmentation using SAM to detect particles in microscopic images without any task-specific labels or training. To characterize each detected particle under optical aberration, we combine shape-aware refinement using elliptic Fourier descriptors with intensity-based features within the refined mask region. We then estimate depth using an uncertainty-aware deep learning model, in which the estimated 3D trajectories are stabilized with a multi-object tracking algorithm and Kalman filter. Against a representative baseline (DefocusTracker), STAR-APTV detects up to six times more particles at high seeding density, while maintaining temporally coherent tracks, and preserving positional accuracy of particles in the presence of noise. Through synthetic validation, the proposed algorithm exhibited AEE = 0.077 px/frame and AAE = 1.45° in a known analytical flow field reconstruction. Experimental validation in two droplet regimes confirms robustness in complex, refractive samples and cross-setup transfer without any task-specific training. Among these flows, in the more challenging flow with relatively dense particle seeding, the detection rate was increased by nearly 70%, with increased retention rates and extended trajectories by almost three times compared to the conventional method. These results altogether demonstrate high-fidelity, single-camera, volumetric velocimetry in refractive, densely seeded environments, extending defocusing/astigmatic PTV toward complex droplet flows.
{"title":"STAR-APTV: Deep learning-enabled 3D flow reconstruction in evaporating multicomponent droplets","authors":"Bumsoo Park , Julius Mauch , Hyeokjin Kweon , Jochen Kriegseis , Seungchul Lee , Hyoungsoo Kim","doi":"10.1016/j.measurement.2026.120368","DOIUrl":"10.1016/j.measurement.2026.120368","url":null,"abstract":"<div><div>Multicomponent droplet evaporation generates inherently three-dimensional, solutal-Marangoni flows that challenge single-camera velocimetry. We present STAR-APTV (Segmentation and Tracking Anything-based Robust Astigmatic Particle Tracking Velocimetry), a zero-shot, deep-learning-assisted, astigmatic particle tracking framework for time-resolved 3D-3C flow reconstruction with minimal optical hardware. We leverage zero-shot segmentation using SAM to detect particles in microscopic images without any task-specific labels or training. To characterize each detected particle under optical aberration, we combine shape-aware refinement using elliptic Fourier descriptors with intensity-based features within the refined mask region. We then estimate depth using an uncertainty-aware deep learning model, in which the estimated 3D trajectories are stabilized with a multi-object tracking algorithm and Kalman filter. Against a representative baseline (DefocusTracker), STAR-APTV detects up to six times more particles at high seeding density, while maintaining temporally coherent tracks, and preserving positional accuracy of particles in the presence of noise. Through synthetic validation, the proposed algorithm exhibited AEE = 0.077 px/frame and AAE = 1.45° in a known analytical flow field reconstruction. Experimental validation in two droplet regimes confirms robustness in complex, refractive samples and cross-setup transfer without any task-specific training. Among these flows, in the more challenging flow with relatively dense particle seeding, the detection rate was increased by nearly 70%, with increased retention rates and extended trajectories by almost three times compared to the conventional method. These results altogether demonstrate high-fidelity, single-camera, volumetric velocimetry in refractive, densely seeded environments, extending defocusing/astigmatic PTV toward complex droplet flows.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120368"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036443","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}
Deep learning algorithms have become prevalent in equipment Prognostics and Health Management (PHM) modeling, yet their deployment in safety–critical applications remains constrained by inherent epistemic uncertainty. This study advances uncertainty calibration methodologies for two fundamental PHM tasks: classification and time-series prediction. First, we proposed a regularization-free calibration framework that dynamically adjust labels to achieve accuracy-uncertainty alignment in classification models. Building on this, we presented the first comprehensive uncertainty calibration framework for time-series prediction models, along with pioneering evaluation metrics specifically designed for temporal uncertainty assessment. To validate our methodology, we acquired comprehensive fault and degradation datasets from aircraft actuators. The evaluation framework encompassed two critical dimensions: (1) classification calibration across four distinct neural architectures on three benchmark datasets, and (2) degradation calibration evaluated across three neural architectures utilizing eight publicly available datasets. Our experimental results consistently demonstrated statistically significant enhancements in calibration performance metrics. The complete implementation is publicly available at https://github.com/ppqweasd/uncertainty-calibration.
{"title":"Uncertainty calibration in PHM model from classification to time-series prediction","authors":"Jinxin Pan, Xiaoxuan Jiao, Bo Jing, Shenglong Wang, Xiangzhen Meng, Zhe Liang","doi":"10.1016/j.measurement.2026.120433","DOIUrl":"10.1016/j.measurement.2026.120433","url":null,"abstract":"<div><div>Deep learning algorithms have become prevalent in equipment Prognostics and Health Management (PHM) modeling, yet their deployment in safety–critical applications remains constrained by inherent epistemic uncertainty. This study advances uncertainty calibration methodologies for two fundamental PHM tasks: classification and time-series prediction. First, we proposed a regularization-free calibration framework that dynamically adjust labels to achieve accuracy-uncertainty alignment in classification models. Building on this, we presented the first comprehensive uncertainty calibration framework for time-series prediction models, along with pioneering evaluation metrics specifically designed for temporal uncertainty assessment. To validate our methodology, we acquired comprehensive fault and degradation datasets from aircraft actuators. The evaluation framework encompassed two critical dimensions:<!--> <!-->(1) classification calibration across four distinct neural architectures on three benchmark datasets, and (2) degradation calibration evaluated across three neural architectures utilizing eight publicly available datasets. Our experimental results consistently demonstrated statistically significant enhancements in calibration performance metrics. The complete implementation is publicly available at <span><span>https://github.com/ppqweasd/uncertainty-calibration</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120433"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036215","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}
Accurate flow-rate measurement underpins effective water-resources management. We develop a measurement-driven, machine-learning “virtual flowmeter” for estimating discharge (Q) through semi-circular flap gates (SCFG) in circular conduits. Five advanced tabular AI models are evaluated—TabPFN (Table-based Prior-Data Fitted Network), SAINT (Self- and Inter-sample Attention Transformer), GMDH (Group Method of Data Handling), LightGBM, and CatBoost. Because hyperparameter choice strongly affects metrological performance, we first compare four metaheuristics—Sparrow Search Algorithm (SSA), Moth-Flame Optimization (MFO), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA)—using GMDH as a baseline. Considering error metrics, convergence speed, stability, and computational cost, MFO yields the most effective hyperparameter optimization; we therefore build hybrid models (e.g., TabPFN-MFO). Dimensional analysis and ANOVA identify the upstream depth-to-diameter ratio (y/D) and normalized gate mass (m/√ρD3) as key predictors. Sensitivity analysis via Explainable Boosting Machine and SHAP confirms y/D as the dominant factor. Model ranking uses Taylor diagrams and Normalized Discrepancy Analysis. During the training stage, the TabPFN-MFO model exhibited the highest predictive accuracy, achieving the largest coefficient of determination and the lowest associated error metrics among all evaluated models. During the testing stage, the performance of the TabPFN-MFO model was again better than other methods, resulting in values of R2 = 0.995, RMSE = 0.0024, sMAPE = 3.8879 %, SI = 0.0320, WMAPE = 3.2072 %, and MAE = 0.0018. The results of uncertainty analysis using the R-Factor further supported that the TabPFN-MFO model possesses less predictive uncertainty than other methods, which signifies improved robustness and reliability of the model.
{"title":"Metaheuristic-tuned tabular models for data-driven flow measurement through semi-circular flap gates","authors":"Ma Chuxin , Ehsan Afaridegan , Pourya Nejatipour , Zhaoliang Zang","doi":"10.1016/j.measurement.2026.120477","DOIUrl":"10.1016/j.measurement.2026.120477","url":null,"abstract":"<div><div>Accurate flow-rate measurement underpins effective water-resources management. We develop a measurement-driven, machine-learning “virtual flowmeter” for estimating discharge (<em>Q</em>) through semi-circular flap gates (SCFG) in circular conduits. Five advanced tabular AI models are evaluated—TabPFN (Table-based Prior-Data Fitted Network), SAINT (Self- and Inter-sample Attention Transformer), GMDH (Group Method of Data Handling), LightGBM, and CatBoost. Because hyperparameter choice strongly affects metrological performance, we first compare four metaheuristics—Sparrow Search Algorithm (SSA), Moth-Flame Optimization (MFO), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA)—using GMDH as a baseline. Considering error metrics, convergence speed, stability, and computational cost, MFO yields the most effective hyperparameter optimization; we therefore build hybrid models (e.g., TabPFN-MFO). Dimensional analysis and ANOVA identify the upstream depth-to-diameter ratio (<em>y/D</em>) and normalized gate mass (<em>m</em>/√<em>ρD</em><sup>3</sup>) as key predictors. Sensitivity analysis via Explainable Boosting Machine and SHAP confirms <em>y/D</em> as the dominant factor. Model ranking uses Taylor diagrams and Normalized Discrepancy Analysis. During the training stage, the TabPFN-MFO model exhibited the highest predictive accuracy, achieving the largest coefficient of determination and the lowest associated error metrics among all evaluated models. During the testing stage, the performance of the TabPFN-MFO model was again better than other methods, resulting in values of <em>R</em><sup>2</sup> = 0.995, RMSE = 0.0024, sMAPE = 3.8879 %, SI = 0.0320, WMAPE = 3.2072 %, and MAE = 0.0018. The results of uncertainty analysis using the R-Factor further supported that the TabPFN-MFO model possesses less predictive uncertainty than other methods, which signifies improved robustness and reliability of the model.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"267 ","pages":"Article 120477"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146043315","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}
Pub Date : 2026-01-14DOI: 10.1016/j.measurement.2026.120479
Zheng Fang , Yuao Gao , Yuheng Cai , Wei Liang
The escalating threat of terrorist attacks demands rapid and non-destructive explosives detection technologies for security checks. Recognizing the limitations of current approaches, namely Raman and infrared spectroscopy, whose testing depth rarely exceeds 5 mm. Mass spectrometry and chromatography also demand tight environmental control and expert operators. To address these drawbacks, we developed a portable broadband lightsource X-ray absorption spectroscopy (BL-XAS) system integrated with a novel deep-learning classifier. The hardware combines a 128-channel CdTe photon-counting detector with a tungsten-target X-ray source. We propose the Parallelized Retention Encoder PR-Encoder that places gated multi-scale retention and multi-layer perceptron modules on parallel computation paths to reduce per-layer latency and accelerate inference. Trained on 2000 spectra from 10 explosive materials, the PR-Encoder was evaluated against two baseline models. Transformer baselines achieved 88.5% classification accuracy with a per-spectrum inference latency of 13.1 ms, while Retention encoders reached 90.1% accuracy with 12.5 ms latency. In contrast, the PR-Encoder attained the highest performance — 93.4% accuracy under ten-fold cross-validation, with an average inference latency of approximately 10.1 ms per spectrum, demonstrating superior accuracy and computational efficiency. Integrating portable BL-XAS instrumentation with retention-based deep learning provides a real-time and non-destructive solution for explosive security screening.
随着恐怖袭击威胁的不断升级，安全检查需要快速、非破坏性的爆炸物检测技术。认识到当前方法的局限性，即拉曼光谱和红外光谱，其测试深度很少超过5毫米。质谱法和色谱法也需要严格的环境控制和专业的操作人员。为了解决这些缺点，我们开发了一种便携式宽带光源x射线吸收光谱（BL-XAS）系统，该系统集成了一种新型深度学习分类器。硬件结合了128通道CdTe光子计数探测器和钨靶x射线源。我们提出了并行保留编码器PR-Encoder，它将门控多尺度保留和多层感知器模块放置在并行计算路径上，以减少每层延迟并加速推理。PR-Encoder对来自10种爆炸材料的2000个光谱进行了训练，并对两个基线模型进行了评估。Transformer基线的分类准确率为88.5%，每频谱推断延迟为13.1 ms，而Retention编码器的准确率为90.1%，延迟为12.5 ms。相比之下，PR-Encoder在10倍交叉验证下获得了最高的性能- 93.4%的准确率，平均推理延迟约为10.1 ms /谱，显示出卓越的准确性和计算效率。将便携式BL-XAS仪器与基于保留的深度学习相结合，为爆炸物安全筛查提供了实时、非破坏性的解决方案。
{"title":"A retention-based method for explosive classification using broadband lightsource X-ray absorption spectroscopy (BL-XAS)","authors":"Zheng Fang , Yuao Gao , Yuheng Cai , Wei Liang","doi":"10.1016/j.measurement.2026.120479","DOIUrl":"10.1016/j.measurement.2026.120479","url":null,"abstract":"<div><div>The escalating threat of terrorist attacks demands rapid and non-destructive explosives detection technologies for security checks. Recognizing the limitations of current approaches, namely Raman and infrared spectroscopy, whose testing depth rarely exceeds 5 mm. Mass spectrometry and chromatography also demand tight environmental control and expert operators. To address these drawbacks, we developed a portable broadband lightsource X-ray absorption spectroscopy (BL-XAS) system integrated with a novel deep-learning classifier. The hardware combines a 128-channel CdTe photon-counting detector with a tungsten-target X-ray source. We propose the Parallelized Retention Encoder PR-Encoder that places gated multi-scale retention and multi-layer perceptron modules on parallel computation paths to reduce per-layer latency and accelerate inference. Trained on 2000 spectra from 10 explosive materials, the PR-Encoder was evaluated against two baseline models. Transformer baselines achieved 88.5% classification accuracy with a per-spectrum inference latency of 13.1 ms, while Retention encoders reached 90.1% accuracy with 12.5 ms latency. In contrast, the PR-Encoder attained the highest performance — 93.4% accuracy under ten-fold cross-validation, with an average inference latency of approximately 10.1 ms per spectrum, demonstrating superior accuracy and computational efficiency. Integrating portable BL-XAS instrumentation with retention-based deep learning provides a real-time and non-destructive solution for explosive security screening.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120479"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979818","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}
Pub Date : 2026-01-14DOI: 10.1016/j.measurement.2026.120460
Fengyun Wang , Guolin Liu , Mengyue Zhang , Fei Wang , Yang Chen , Fengyun Zhang
This paper introduces a metrological approach to enhance the accuracy and reliability of Synthetic Aperture Radar (SAR) radiometric measurements. A hybrid-regularization sparse reconstruction based on the non-subsampled contourlet transform (NSCT) is proposed as a dedicated radiometric estimator for calibrated single-look complex (SLC) data, aiming to reduce speckle-induced measurement uncertainty while preserving spatial resolution and radiometric calibration. Structurally, L1-norm regularization stabilises the estimation of low-frequency components that carry the primary radiometric information, while Frobenius-norm shrinkage denoises high-frequency subbands that are critical for edge and texture fidelity. An interpretable and auditable solution is derived via proximal iterations (Gradient Descent/Iterative Shrinkage-Thresholding Algorithm). Quantitative evaluation under a GUM-consistent uncertainty framework shows that the proposed method increases the equivalent number of looks (ENL) in homogeneous regions by up to a factor of 1.73, while significantly reducing the expanded uncertainty of regional backscatter estimates. Compared with matched filtering and several sparse SAR reconstruction methods, the proposed approach consistently achieves improved radiometric stability, enhanced edge and texture preservation, and effective suppression of sidelobes, noise, and clutter. This NSCT-regularized sparse SAR reconstruction provides a traceable and metrologically sound pathway for obtaining higher-quality SAR-derived geophysical quantities.
{"title":"NSCT-regularized radiometric reconstruction of complex-valued SAR images","authors":"Fengyun Wang , Guolin Liu , Mengyue Zhang , Fei Wang , Yang Chen , Fengyun Zhang","doi":"10.1016/j.measurement.2026.120460","DOIUrl":"10.1016/j.measurement.2026.120460","url":null,"abstract":"<div><div>This paper introduces a metrological approach to enhance the accuracy and reliability of Synthetic Aperture Radar (SAR) radiometric measurements. A hybrid-regularization sparse reconstruction based on the non-subsampled contourlet transform (NSCT) is proposed as a dedicated radiometric estimator for calibrated single-look complex (SLC) data, aiming to reduce speckle-induced measurement uncertainty while preserving spatial resolution and radiometric calibration. Structurally, L1-norm regularization stabilises the estimation of low-frequency components that carry the primary radiometric information, while Frobenius-norm shrinkage denoises high-frequency subbands that are critical for edge and texture fidelity. An interpretable and auditable solution is derived via proximal iterations (Gradient Descent/Iterative Shrinkage-Thresholding Algorithm). Quantitative evaluation under a GUM-consistent uncertainty framework shows that the proposed method increases the equivalent number of looks (ENL) in homogeneous regions by up to a factor of 1.73, while significantly reducing the expanded uncertainty of regional backscatter estimates. Compared with matched filtering and several sparse SAR reconstruction methods, the proposed approach consistently achieves improved radiometric stability, enhanced edge and texture preservation, and effective suppression of sidelobes, noise, and clutter. This NSCT-regularized sparse SAR reconstruction provides a traceable and metrologically sound pathway for obtaining higher-quality SAR-derived geophysical quantities.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120460"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145981975","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}
Pub Date : 2026-01-14DOI: 10.1016/j.measurement.2026.120436
Mingyu Yuan , Shilu Mai , Jin Wang , Xinyang Li , Jinlong Liu , Yaqi Wu , Zheng Li , Wei Liu , Zongjin Ren
The aero vector engine is the core backbone for the high agility, post-stall maneuverability, and short takeoff/landing capabilities of next-generation fighter aircraft. During its development and finalization phases, precise measurement of the generated vector thrust parameters is indispensable to establish the correlation between thrust magnitude, angle, and control parameters—this precise measurement ensures the accurate execution of the fighter’s tactical maneuvers. To overcome the limitation of inadequate dynamic response in traditional strain-based distributed measurement methods, this study proposes a high-dynamic vector thrust distributed testing system for aero engines, with piezoelectric sensors as force-sensitive elements. To achieve high-precision measurement, the research follows a logical workflow: first, the influence of different spatial arrangements of piezoelectric force-sensitive units (PFSU) on test performance was investigated, and a theoretical mechanical model was established to correlate the system’s overall force with the three-dimensional force outputs of individual PFSUs; based on the theoretical model, ANSYS was utilized for static and modal simulation analysis to confirm that the system’s structural strength and natural frequency meet design requirements; after validating the feasibility via simulation, three-way orthogonal experiments were conducted using a calibration loading device to obtain the calibration coefficient matrix and decoupling compensation matrix; finally, single/double vector angle simulation loading experiments are performed to verify the reliability of the system’s performance. Experimental results show that the AE-VTTS exhibits excellent static and dynamic performance, with good linearity and repeatability. The measurement error of force and angle in each direction is ≤1.5%, and the first-order natural frequency reaches 20.02 Hz (higher than the 10 Hz of the strain-based system in the laboratory). This study provides an innovative solution for the high-precision static and dynamic measurement of aero engine vector thrust.
{"title":"Design of a distributed testing system for high-dynamic vector thrust of aero-engines based on piezoelectric force-sensitive units","authors":"Mingyu Yuan , Shilu Mai , Jin Wang , Xinyang Li , Jinlong Liu , Yaqi Wu , Zheng Li , Wei Liu , Zongjin Ren","doi":"10.1016/j.measurement.2026.120436","DOIUrl":"10.1016/j.measurement.2026.120436","url":null,"abstract":"<div><div>The aero vector engine is the core backbone for the high agility, post-stall maneuverability, and short takeoff/landing capabilities of next-generation fighter aircraft. During its development and finalization phases, precise measurement of the generated vector thrust parameters is indispensable to establish the correlation between thrust magnitude, angle, and control parameters—this precise measurement ensures the accurate execution of the fighter’s tactical maneuvers. To overcome the limitation of inadequate dynamic response in traditional strain-based distributed measurement methods, this study proposes a high-dynamic vector thrust distributed testing system for aero engines, with piezoelectric sensors as force-sensitive elements. To achieve high-precision measurement, the research follows a logical workflow: first, the influence of different spatial arrangements of piezoelectric force-sensitive units (PFSU) on test performance was investigated, and a theoretical mechanical model was established to correlate the system’s overall force with the three-dimensional force outputs of individual PFSUs; based on the theoretical model, ANSYS was utilized for static and modal simulation analysis to confirm that the system’s structural strength and natural frequency meet design requirements; after validating the feasibility via simulation, three-way orthogonal experiments were conducted using a calibration loading device to obtain the calibration coefficient matrix and decoupling compensation matrix; finally, single/double vector angle simulation loading experiments are performed to verify the reliability of the system’s performance. Experimental results show that the AE-VTTS exhibits excellent static and dynamic performance, with good linearity and repeatability. The measurement error of force and angle in each direction is ≤1.5%, and the first-order natural frequency reaches 20.02 Hz (higher than the 10 Hz of the strain-based system in the laboratory). This study provides an innovative solution for the high-precision static and dynamic measurement of aero engine vector thrust.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120436"},"PeriodicalIF":5.6,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036262","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}
Pub Date : 2026-01-13DOI: 10.1016/j.measurement.2026.120432
Jun Zhang , Yuqian Zhao , Jinlong Liu , Yongwang Cao , Shijia Li , Hengkun Shao , Yaming Jin , Zongjin Ren , Wei Liu
To address the technical challenges of poor flexibility in sensor placement and difficulties in force measurement and calculation for condition monitoring of complex structures, this study proposes a piezoelectric force-measuring device with an irregular sensor arrangement and a dual-dimensional performance detection method, breaking through the space constraints of traditional symmetric arrangements in irregular surfaces and enabling high-precision measurement of force values and action points within asymmetric surface areas. An asymmetric mechanical model was established, and combined with the force-position matching relationship, the corresponding relationship between the force to be measured and the output voltage was revealed. The finite element method (FEM) was employed to perform parameter optimization, modal analysis, and stiffness analysis of the measuring device, thereby completing the structural design of key components in the measurement system. A self-developed multi-dimensional force calibration platform was utilized to conduct static calibration, natural frequency testing, and investigation on variable loading points of the force-measuring system, and the output variations of the force-measuring system under different loading positions were obtained. The results indicate that the output performance of the proposed device is highly consistent with the theoretical results, exhibiting high precision and a high natural frequency. Compared with the traditional force-measuring device with four-point symmetric arrangement, the proposed device features significantly enhanced flexibility in sensor placement, which can adapt to the irregular installation spaces of complex structures. Specifically, its linearity and repeatability errors are both less than 0.4 %, and its first-order natural frequency reaches 2124 Hz; additionally, the outputs of the device at different loading positions within the asymmetric surface area show good consistency.
{"title":"Research on piezoelectric force measurement methods with Asymmetrical sensor arrangement","authors":"Jun Zhang , Yuqian Zhao , Jinlong Liu , Yongwang Cao , Shijia Li , Hengkun Shao , Yaming Jin , Zongjin Ren , Wei Liu","doi":"10.1016/j.measurement.2026.120432","DOIUrl":"10.1016/j.measurement.2026.120432","url":null,"abstract":"<div><div>To address the technical challenges of poor flexibility in sensor placement and difficulties in force measurement and calculation for condition monitoring of complex structures, this study proposes a piezoelectric force-measuring device with an irregular sensor arrangement and a dual-dimensional performance detection method, breaking through the space constraints of traditional symmetric arrangements in irregular surfaces and enabling high-precision measurement of force values and action points within asymmetric surface areas. An asymmetric mechanical model was established, and combined with the force-position matching relationship, the corresponding relationship between the force to be measured and the output voltage was revealed. The finite element method (FEM) was employed to perform parameter optimization, modal analysis, and stiffness analysis of the measuring device, thereby completing the structural design of key components in the measurement system. A self-developed multi-dimensional force calibration platform was utilized to conduct static calibration, natural frequency testing, and investigation on variable loading points of the force-measuring system, and the output variations of the force-measuring system under different loading positions were obtained. The results indicate that the output performance of the proposed device is highly consistent with the theoretical results, exhibiting high precision and a high natural frequency. Compared with the traditional force-measuring device with four-point symmetric arrangement, the proposed device features significantly enhanced flexibility in sensor placement, which can adapt to the irregular installation spaces of complex structures. Specifically, its linearity and repeatability errors are both less than 0.4 %, and its first-order natural frequency reaches 2124 Hz; additionally, the outputs of the device at different loading positions within the asymmetric surface area show good consistency.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120432"},"PeriodicalIF":5.6,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036442","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}
Accurate 3D measurement of the inner surface is essential for quality control in the precise assembly of cylindrical cavity components. Considering the restricted space within a cylindrical cavity, current geometric measurement methods are unable to effectively capture panoramic information (a 360° cross-sectional profile of the current location) of the inner surface. To this end, this paper proposes a geometric strategy for panoramic measurement combining circular structured light and parallel binocular cameras, which is able to realize the accurate measurement of the panoramic information on the inner surface of a cylindrical cavity. First, the circular structured light in the probe is calibrated using a blank planar plate to improve the measurement accuracy of the inner surface of the cylindrical cavity; second, the circular structured stripes are projected onto the inner surface of the cylindrical cavity, and the panoramic information of the inner surface of the cylindrical cavity is captured by a parallel binocular camera. Finally, the 3D reconstruction of the inner wall of the cylinder cavity is completed by carrying the probe on an arbitrary moving platform. Measurement results indicate that the developed probe maintains an inner diameter error within ± 13 μm and an RMS error within 15 μm when measuring standard ring gauges of varying inner diameters. Furthermore, 3D measurements of motor housings can be completed in just 6 to 8 s. Additionally, the developed probe exhibits significant potential for industrial applications in restricted spaces, such as artillery barrels and motor housings.
{"title":"Panoramic geometry measurement of inner surfaces in cylindrical restricted spaces","authors":"Zuo Zhang , Huining Zhao , Minghui Duan , Haojie Xia","doi":"10.1016/j.measurement.2026.120424","DOIUrl":"10.1016/j.measurement.2026.120424","url":null,"abstract":"<div><div>Accurate 3D measurement of the inner surface is essential for quality control in the precise assembly of cylindrical cavity components. Considering the restricted space within a cylindrical cavity, current geometric measurement methods are unable to effectively capture panoramic information (a 360° cross-sectional profile of the current location) of the inner surface. To this end, this paper proposes a geometric strategy for panoramic measurement combining circular structured light and parallel binocular cameras, which is able to realize the accurate measurement of the panoramic information on the inner surface of a cylindrical cavity. First, the circular structured light in the probe is calibrated using a blank planar plate to improve the measurement accuracy of the inner surface of the cylindrical cavity; second, the circular structured stripes are projected onto the inner surface of the cylindrical cavity, and the panoramic information of the inner surface of the cylindrical cavity is captured by a parallel binocular camera. Finally, the 3D reconstruction of the inner wall of the cylinder cavity is completed by carrying the probe on an arbitrary moving platform. Measurement results indicate that the developed probe maintains an inner diameter error within ± 13 μm and an RMS error within 15 μm when measuring standard ring gauges of varying inner diameters. Furthermore, 3D measurements of motor housings can be completed in just 6 to 8 s. Additionally, the developed probe exhibits significant potential for industrial applications in restricted spaces, such as artillery barrels and motor housings.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120424"},"PeriodicalIF":5.6,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146036491","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号