Pub Date : 2026-01-12DOI: 10.1016/j.measurement.2026.120420
Anastasia Medvedeva , Aleksandra Titova , Anna Kharkova , Aleksey Efremov , Andrey Kulikov , Adele Latypova , Roman Perchikov , Vyacheslav Arlyapov
This article presents the development of an electrochemical biosensor based on conductive polymers and microorganisms for rapid and sensitive detection of surfactants in aqueous media. The study includes a systematic comparative investigation of several conductive polymers (poly(neutral red) pNR, poly(thionine) pTN, polypyrrole PPy, polyaniline PANI, and PEDOT:PSS) and selection of the optimal conductive polymer (pNR) and microorganism (Pseudomonas putida VKM B-973) as well as investigation of their interaction rates and electrochemical properties. Modification of electrodes with carbon nanomaterials such as single-walled carbon nanotubes (SWCNT) has been shown to significantly improve the sensitivity (lower limit of detectable concentration is 0.061 mg/dm3) and stability (microbial sensor can function for 15 days with relative standard deviation of analytical signal being 5.4 %) of the biosensor. The device is used for detecting anionic surfactant in river water samples and the results obtained are statistically insignificant from those obtained by the conventional method of analysis.
{"title":"Enhanced surfactant detection using microbial biosensor: new applications of conducting polymers and their nanocomposites","authors":"Anastasia Medvedeva , Aleksandra Titova , Anna Kharkova , Aleksey Efremov , Andrey Kulikov , Adele Latypova , Roman Perchikov , Vyacheslav Arlyapov","doi":"10.1016/j.measurement.2026.120420","DOIUrl":"10.1016/j.measurement.2026.120420","url":null,"abstract":"<div><div>This article presents the development of an electrochemical biosensor based on conductive polymers and microorganisms for rapid and sensitive detection of surfactants in aqueous media. The study includes a systematic comparative investigation of several conductive polymers (poly(neutral red) pNR, poly(thionine) pTN, polypyrrole PPy, polyaniline PANI, and PEDOT:PSS) and selection of the optimal conductive polymer (pNR) and microorganism (<em>Pseudomonas putida</em> VKM B-973) as well as investigation of their interaction rates and electrochemical properties. Modification of electrodes with carbon nanomaterials such as single-walled carbon nanotubes (SWCNT) has been shown to significantly improve the sensitivity (lower limit of detectable concentration is 0.061 mg/dm<sup>3</sup>) and stability (microbial sensor can function for 15 days with relative standard deviation of analytical signal being 5.4 %) of the biosensor. The device is used for detecting anionic surfactant in river water samples and the results obtained are statistically insignificant from those obtained by the conventional method of analysis.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120420"},"PeriodicalIF":5.6,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980081","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 : 2026-01-12DOI: 10.1016/j.measurement.2026.120425
Chunlei Xie , Xianglong Li , Ze Zhang , Yaqian Dong , Qingkai Yan , Andrei Zhang
Permafrost degradation driven by regional warming increasingly threatens the stability of transportation infrastructure. Existing studies have documented surface deformation associated with permafrost change, yet the underlying mechanisms and their variability across different permafrost regimes remain poorly elucidated. This work established an integrated space-ground-underground monitoring framework that integrates Sentinel-1 satellite observations, meteorological observations, and borehole measurements, and applies Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR), seasonal surface deformation model, the GeoDetector model, and correlation analysis to systematically investigate surface deformation along the Genhe to Labudalin (G332) Highway, which traverses both sporadic permafrost and isolated patch permafrost areas. Results show that surface deformation along the highway is pronounced, dominated by subsidence and exhibiting clear seasonal variability. Permafrost thaw driven by regional warming is the primary cause of surface deformation, while spatial heterogeneity is shaped by local factors. Distance from the highway (DFH) influences long-term deformation rate, whereas aspect influences seasonal surface deformation. Surface deformation is notably stronger in sporadic permafrost than in isolated patch permafrost areas, indicating more advanced permafrost degradation. By integrating InSAR measurements with ground meteorological records and underground borehole observations, this study provides a quantitative assessment of permafrost degradation and its expression in surface deformation across sporadic permafrost and isolated patch permafrost areas, offering critical guidance for monitoring and maintaining infrastructure in permafrost regions.
{"title":"Measurement of surface deformation along the Genhe–Labudalin highway in Northeast China using time-series InSAR and ground observations","authors":"Chunlei Xie , Xianglong Li , Ze Zhang , Yaqian Dong , Qingkai Yan , Andrei Zhang","doi":"10.1016/j.measurement.2026.120425","DOIUrl":"10.1016/j.measurement.2026.120425","url":null,"abstract":"<div><div>Permafrost degradation driven by regional warming increasingly threatens the stability of transportation infrastructure. Existing studies have documented surface deformation associated with permafrost change, yet the underlying mechanisms and their variability across different permafrost regimes remain poorly elucidated. This work established an integrated space-ground-underground monitoring framework that integrates Sentinel-1 satellite observations, meteorological observations, and borehole measurements, and applies Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR), seasonal surface deformation model, the GeoDetector model, and correlation analysis to systematically investigate surface deformation along the Genhe to Labudalin (G332) Highway, which traverses both sporadic permafrost and isolated patch permafrost areas. Results show that surface deformation along the highway is pronounced, dominated by subsidence and exhibiting clear seasonal variability. Permafrost thaw driven by regional warming is the primary cause of surface deformation, while spatial heterogeneity is shaped by local factors. Distance from the highway (DFH) influences long-term deformation rate, whereas aspect influences seasonal surface deformation. Surface deformation is notably stronger in sporadic permafrost than in isolated patch permafrost areas, indicating more advanced permafrost degradation. By integrating InSAR measurements with ground meteorological records and underground borehole observations, this study provides a quantitative assessment of permafrost degradation and its expression in surface deformation across sporadic permafrost and isolated patch permafrost areas, offering critical guidance for monitoring and maintaining infrastructure in permafrost regions.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120425"},"PeriodicalIF":5.6,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980082","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 : 2026-01-12DOI: 10.1016/j.measurement.2026.120313
Jingshan Hong , Haigen Hu , Huihuang Zhang , Kangkang Song , Qianwei Zhou
Due to the diverse and complex processes in the manufacturing of Carbon Fiber-Reinforced Thermoplastic Prepreg(CFRTP), its surface is prone to generating defects, such as yarn feathers and wrinkles, leading to performance degradation. Currently, it is still a challenge to accurately and effectively detect these common surface defects. In this work, we propose DD-Net, a visual inspection-based defect detection model designed for CFRTP surface quality assessment. Specifically, an Efficient Re-parameter Aggregation Module (ERAM) is introduced to enhance feature extraction and inference speed, while a lightweight multi-scale pooling module (CCSPPF) is designed to improve multi-scale feature fusion efficiency. In addition, an attention-based downsampling module (DS-A) is proposed to strengthen small defect perception. Finally, a lightweight decoupled detection head is proposed to balance detection accuracy and speed by improving localization and classification precision. Extensive experiments demonstrate that DD-Net achieves superior performance compared with mainstream detection methods, reaching an [email protected] of 95.2% on the CFRTP dataset. Furthermore, comprehensive interpretability and ablation analyses validate the effectiveness of each proposed module and provide deeper insights into how the model captures and distinguishes key defect characteristics.
{"title":"DD-Net: A defect detection model for carbon fiber-reinforce thermoplastic prepreg surface","authors":"Jingshan Hong , Haigen Hu , Huihuang Zhang , Kangkang Song , Qianwei Zhou","doi":"10.1016/j.measurement.2026.120313","DOIUrl":"10.1016/j.measurement.2026.120313","url":null,"abstract":"<div><div>Due to the diverse and complex processes in the manufacturing of Carbon Fiber-Reinforced Thermoplastic Prepreg(CFRTP), its surface is prone to generating defects, such as yarn feathers and wrinkles, leading to performance degradation. Currently, it is still a challenge to accurately and effectively detect these common surface defects. In this work, we propose DD-Net, a visual inspection-based defect detection model designed for CFRTP surface quality assessment. Specifically, an Efficient Re-parameter Aggregation Module (ERAM) is introduced to enhance feature extraction and inference speed, while a lightweight multi-scale pooling module (CCSPPF) is designed to improve multi-scale feature fusion efficiency. In addition, an attention-based downsampling module (DS-A) is proposed to strengthen small defect perception. Finally, a lightweight decoupled detection head is proposed to balance detection accuracy and speed by improving localization and classification precision. Extensive experiments demonstrate that DD-Net achieves superior performance compared with mainstream detection methods, reaching an [email protected] of 95.2% on the CFRTP dataset. Furthermore, comprehensive interpretability and ablation analyses validate the effectiveness of each proposed module and provide deeper insights into how the model captures and distinguishes key defect characteristics.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120313"},"PeriodicalIF":5.6,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145981976","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 : 2026-01-11DOI: 10.1016/j.measurement.2026.120412
Jie Zhang , Jian Chen , Jintao Pan , Zhiyong Ji , Gongbiao Yang , Yimin Xia
Disc cutters on Tunnel Boring Machines (TBMs) fracture rock by rotating passively under the combined action of thrust and cutterhead motion. Consequently, their rotational speed is a key measurand for fault diagnosis and wear prediction. However, in harsh tunneling environments, this signal is heavily contaminated by noise and transient shocks; such contamination obscures weak but informative features. To address this, an adaptive denoising method, ICEEMDAN-EK-WSTD is proposed. It integrates improved complete ensemble empirical mode decomposition (ICEEMDAN), envelope kurtosis (EK), and wavelet soft-threshold denoising (WSTD). An EK-based intrinsic mode function (IMF) selection strategy is introduced to adaptively identify noise-dominated IMFs, thereby replacing fixed empirical thresholds. The selected IMFs are then denoised via WSTD and recombined to reconstruct the signal. Experiments are conducted on a laboratory-scale linear rock-cutting platform to evaluate performance under four typical health states—normal, evenly wear, chipping, and uneven wear. Across these states, the proposed method achieved an average signal-to-noise ratio (SNR) of 29.12 dB and a root-mean-square error (RMSE) as low as 0.1381, yielding up to a 10.73 dB improvement in SNR (average 3.35 dB) relative to baseline methods. These results demonstrate the method’s feasibility at a laboratory scale. The resulting higher-fidelity rotational-speed signal enables more accurate cutter-wear estimation and fault identification, thereby strengthening TBM condition monitoring. In practice, more reliable measurements can facilitate earlier maintenance decisions, reduce unplanned downtime, and enhance logging of rock–machine interaction for construction planning. As next steps, we will embed the sensing module into a full-scale disc cutter and conduct short-duration in situ tests on an operational TBM to assess method performance under production conditions.
{"title":"An adaptive denoising method based on ICEEMDAN-EK-WSTD for cutter rotational speed signals in tunnel boring Machines","authors":"Jie Zhang , Jian Chen , Jintao Pan , Zhiyong Ji , Gongbiao Yang , Yimin Xia","doi":"10.1016/j.measurement.2026.120412","DOIUrl":"10.1016/j.measurement.2026.120412","url":null,"abstract":"<div><div>Disc cutters on Tunnel Boring Machines (TBMs) fracture rock by rotating passively under the combined action of thrust and cutterhead motion. Consequently, their rotational speed is a key measurand for fault diagnosis and wear prediction. However, in harsh tunneling environments, this signal is heavily contaminated by noise and transient shocks; such contamination obscures weak but informative features. To address this, an adaptive denoising method, ICEEMDAN-EK-WSTD is proposed. It integrates improved complete ensemble empirical mode decomposition (ICEEMDAN), envelope kurtosis (EK), and wavelet soft-threshold denoising (WSTD). An EK-based intrinsic mode function (IMF) selection strategy is introduced to adaptively identify noise-dominated IMFs, thereby replacing fixed empirical thresholds. The selected IMFs are then denoised via WSTD and recombined to reconstruct the signal. Experiments are conducted on a laboratory-scale linear rock-cutting platform to evaluate performance under four typical health states—normal, evenly wear, chipping, and uneven wear. Across these states, the proposed method achieved an average signal-to-noise ratio (SNR) of 29.12 dB and a root-mean-square error (RMSE) as low as 0.1381, yielding up to a 10.73 dB improvement in SNR (average 3.35 dB) relative to baseline methods. These results demonstrate the method’s feasibility at a laboratory scale. The resulting higher-fidelity rotational-speed signal enables more accurate cutter-wear estimation and fault identification, thereby strengthening TBM condition monitoring. In practice, more reliable measurements can facilitate earlier maintenance decisions, reduce unplanned downtime, and enhance logging of rock–machine interaction for construction planning. As next steps, we will embed the sensing module into a full-scale disc cutter and conduct short-duration in situ tests on an operational TBM to assess method performance under production conditions.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120412"},"PeriodicalIF":5.6,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980033","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 : 2026-01-11DOI: 10.1016/j.measurement.2026.120415
Yangjian Lin , Peiyang Chen , Jun Xu , Xiaolong Zhang , Jianchao Yu , Heng Du , Zijun Li
Tires are the only vehicle components in contact with the road surface, and their dynamic state directly determines overall driving safety and handling stability. Therefore, high-precision measurement of tire longitudinal force is essential for enhancing the dynamic response performance of vehicle active safety control systems. However, under extreme driving conditions involving rapid transitions between driving and braking, sensor noise increases sharply, while the tire longitudinal force exhibits abrupt nonlinear variations, making high-precision force measurement extremely challenging. Accordingly, a novel measurement system is constructed by integrating wireless intelligent tires with a Multi-Granularity Hierarchical Cooperative Network (MGHCN). Specifically, the wireless intelligent tire signal acquisition and preprocessing system is used to obtain high-quality dynamic acceleration signals; a cross-channel dual-level multi-granularity embedding-attention mechanism enables deep fusion of multi-scale acceleration features; on this basis, the hierarchical collaborative estimation network significantly improves the accuracy of longitudinal force estimation under extreme driving conditions. Experimental results demonstrate that the proposed method achieves a Normalized Root Mean Square Error (NRMSE) of 3.3172% and a coefficient of determination (R2) of 0.9980 under complex driving conditions. With an Average Single-Sample Inference Time (ASSIT) of merely 0.30 ms, it exhibits comprehensive advantages in estimation accuracy and real-time performance. Therefore, this research offers a novel solution for intelligent tire longitudinal force measurement in extreme driving conditions, with significant implications for vehicle active safety control.
{"title":"Novel intelligent tire longitudinal force measurement system based on multi-granularity hierarchical collaborative networks","authors":"Yangjian Lin , Peiyang Chen , Jun Xu , Xiaolong Zhang , Jianchao Yu , Heng Du , Zijun Li","doi":"10.1016/j.measurement.2026.120415","DOIUrl":"10.1016/j.measurement.2026.120415","url":null,"abstract":"<div><div>Tires are the only vehicle components in contact with the road surface, and their dynamic state directly determines overall driving safety and handling stability. Therefore, high-precision measurement of tire longitudinal force is essential for enhancing the dynamic response performance of vehicle active safety control systems. However, under extreme driving conditions involving rapid transitions between driving and braking, sensor noise increases sharply, while the tire longitudinal force exhibits abrupt nonlinear variations, making high-precision force measurement extremely challenging. Accordingly, a novel measurement system is constructed by integrating wireless intelligent tires with a Multi-Granularity Hierarchical Cooperative Network (MGHCN). Specifically, the wireless intelligent tire signal acquisition and preprocessing system is used to obtain high-quality dynamic acceleration signals; a cross-channel dual-level multi-granularity embedding-attention mechanism enables deep fusion of multi-scale acceleration features; on this basis, the hierarchical collaborative estimation network significantly improves the accuracy of longitudinal force estimation under extreme driving conditions. Experimental results demonstrate that the proposed method achieves a Normalized Root Mean Square Error (NRMSE) of 3.3172% and a coefficient of determination (R<sup>2</sup>) of 0.9980 under complex driving conditions. With an Average Single-Sample Inference Time (ASSIT) of merely 0.30 ms, it exhibits comprehensive advantages in estimation accuracy and real-time performance. Therefore, this research offers a novel solution for intelligent tire longitudinal force measurement in extreme driving conditions, with significant implications for vehicle active safety control.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120415"},"PeriodicalIF":5.6,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979881","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 : 2026-01-11DOI: 10.1016/j.measurement.2026.120374
Qianhai Lu , Lingfei Kong , Hongbo Dong , Longlong Li , Jie Liu , Jin Sun
To address the critical technical challenges in the automated drill pipe thread make-up & break-out operations of the Underground Drilling Robot in Coal Mines, including multi-physics coupled interference, the inability to perform online detection of thread fastening status, and the lack of fault self-recovery mechanisms, this study innovatively proposes a dynamic thread state identification and adaptive make-up & break-out control method based on spatiotemporal fusion of multi-source heterogeneous information. By constructing a multi-dimensional sensor data fusion framework, a dynamic evolution model is established. Quantitative metrics for effective make-up & break-out lengths are proposed, overcoming the limitations of traditional threshold control:① A displacement-feed pressure-rotational speed joint analysis algorithm is developed to detect effective make-up length, achieving precise localisation of thread contact states through discrete wavelet multi-scale decomposition.② A displacement-rotational speed-feed velocity collaborative monitoring model is designed to calculate effective break-out length, enhancing robustness under complex working conditions by incorporating a sliding window dynamic optimization mechanism. To address the challenge of self-recovery during make-up & break-out failures, A stepwise back-off control method is established to resolve axial deviation issues during drill pipe thread make-up. A dual-mode torque regulation strategy is constructed to dynamically respond to abnormal preload conditions during drill pipe thread break-out. Industrial experiments demonstrate that the system achieves thread state recognition accuracies of 96.38% (make-up) and 97.55% (break-out) within a drilling inclination range of −90° to 90°. Large-scale underground validations confirm fault self-recovery success rates of 82.67% (make-up) and 70.69% (break-out). The “perception-decision-execution” closed-loop control methodology advances the automation level of drill pipe handling processes.
{"title":"Research on intelligent make-up & break-out methods for drill pipe threads of underground drilling robot in coal mine based on multi-source information fusion","authors":"Qianhai Lu , Lingfei Kong , Hongbo Dong , Longlong Li , Jie Liu , Jin Sun","doi":"10.1016/j.measurement.2026.120374","DOIUrl":"10.1016/j.measurement.2026.120374","url":null,"abstract":"<div><div>To address the critical technical challenges in the automated drill pipe thread make-up & break-out operations of the Underground Drilling Robot in Coal Mines, including multi-physics coupled interference, the inability to perform online detection of thread fastening status, and the lack of fault self-recovery mechanisms, this study innovatively proposes a dynamic thread state identification and adaptive make-up & break-out control method based on spatiotemporal fusion of multi-source heterogeneous information. By constructing a multi-dimensional sensor data fusion framework, a dynamic evolution model is established. Quantitative metrics for effective make-up & break-out lengths are proposed, overcoming the limitations of traditional threshold control:① A displacement-feed pressure-rotational speed joint analysis algorithm is developed to detect effective make-up length, achieving precise localisation of thread contact states through discrete wavelet multi-scale decomposition.② A displacement-rotational speed-feed velocity collaborative monitoring model is designed to calculate effective break-out length, enhancing robustness under complex working conditions by incorporating a sliding window dynamic optimization mechanism. To address the challenge of self-recovery during make-up & break-out failures, A stepwise back-off control method is established to resolve axial deviation issues during drill pipe thread make-up. A dual-mode torque regulation strategy is constructed to dynamically respond to abnormal preload conditions during drill pipe thread break-out. Industrial experiments demonstrate that the system achieves thread state recognition accuracies of 96.38% (make-up) and 97.55% (break-out) within a drilling inclination range of −90° to 90°. Large-scale underground validations confirm fault self-recovery success rates of 82.67% (make-up) and 70.69% (break-out). The “perception-decision-execution” closed-loop control methodology advances the automation level of drill pipe handling processes.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120374"},"PeriodicalIF":5.6,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980599","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 : 2026-01-11DOI: 10.1016/j.measurement.2026.120409
Chun Wang , Songtao Ye , Xiandao Lei , Dou An , Huan Xi
Accurate monitoring of State of Charge (SOC), State of Health (SOH), and State of Temperature (SOT) is indispensable for ensuring the operational safety and efficiency of Battery Management Systems (BMS) in electric vehicles and large-scale energy storage. However, conventional data-driven approaches often isolate these states, overlooking their intrinsic physical coupling, which inevitably leads to suboptimal estimation accuracy. To address this challenge, this paper introduces a novel integrated co-estimation framework, the Attention Mechanism-Parallel Temporal Convolutional Neural Network (AM-PTCN). By leveraging a parallel feature extraction structure combined with a physically interpretable attention mechanism, the model dynamically identifies and re-weights influential factors within multivariate time-series inputs, effectively disentangling the complex interdependencies among battery states. Furthermore, estimation uncertainty is quantified to provide a probabilistic assessment of system reliability. The optimized model achieves Mean Absolute Errors (MAE) of 1.3323% for SOC, 1.8901% for SOH, and 0.2208 for SOT. Crucially, rigorous ablation studies and comparative experiments validate the specific contributions of the attention-based parallel architecture, demonstrating superior accuracy and robustness over existing machine learning approaches. Validated on noise-contaminated datasets, this novel joint estimation framework provides a valuable reference for advanced battery monitoring.
{"title":"Battery joint state estimation with uncertainty based on feature independent representation and re-weighted","authors":"Chun Wang , Songtao Ye , Xiandao Lei , Dou An , Huan Xi","doi":"10.1016/j.measurement.2026.120409","DOIUrl":"10.1016/j.measurement.2026.120409","url":null,"abstract":"<div><div>Accurate monitoring of State of Charge (SOC), State of Health (SOH), and State of Temperature (SOT) is indispensable for ensuring the operational safety and efficiency of Battery Management Systems (BMS) in electric vehicles and large-scale energy storage. However, conventional data-driven approaches often isolate these states, overlooking their intrinsic physical coupling, which inevitably leads to suboptimal estimation accuracy. To address this challenge, this paper introduces a novel integrated co-estimation framework, the Attention Mechanism-Parallel Temporal Convolutional Neural Network (AM-PTCN). By leveraging a parallel feature extraction structure combined with a physically interpretable attention mechanism, the model dynamically identifies and re-weights influential factors within multivariate time-series inputs, effectively disentangling the complex interdependencies among battery states. Furthermore, estimation uncertainty is quantified to provide a probabilistic assessment of system reliability. The optimized model achieves Mean Absolute Errors (MAE) of 1.3323% for SOC, 1.8901% for SOH, and 0.2208 for SOT. Crucially, rigorous ablation studies and comparative experiments validate the specific contributions of the attention-based parallel architecture, demonstrating superior accuracy and robustness over existing machine learning approaches. Validated on noise-contaminated datasets, this novel joint estimation framework provides a valuable reference for advanced battery monitoring.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120409"},"PeriodicalIF":5.6,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980026","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 : 2026-01-11DOI: 10.1016/j.measurement.2026.120383
Mingyue Zhang , Yang Han , Yongchao Sun , Huaiyu Zhang , Fan Yang , Congling Wang
Accurate ultra-short-term forecasting of wind power is crucial for the reliable integration of renewable energy. This study proposes a comprehensive forecasting framework that integrates spatial preprocessing, signal decomposition, hybrid deep learning, transfer learning, and interval prediction. An enhanced Kriging interpolation (EKI) method is first introduced to refine the spatial resolution of numerical weather prediction data. A two-stage clustering approach combining self-organizing maps (SOM) with K-means is employed to group wind farms with similar spatiotemporal characteristics, thereby facilitating knowledge transfer. Wind power series are subsequently decomposed through variational mode decomposition (VMD) optimized by an improved Aquila optimizer (IAO), while key input features are selected via Pearson correlation analysis. For each subsequence, a hybrid deep learning network integrating bidirectional temporal convolutional networks (BiTCN) and bidirectional gated recurrent units (BiGRU) with attention mechanism (AM) is constructed to forecast using multi-scale temporal features. Transfer learning is then applied to adapt the model to new sites. Finally, a heteroscedastic Gaussian process regression (HGPR) module is further employed to generate reliable interval forecasts. Case studies on 13 wind farms in Sichuan Province, China, evaluated against twelve comparative models from four perspectives, show that the framework improves forecasting precision, enhances uncertainty quantification, and reduces training time by over 90%.
{"title":"Power forecasting for distributed wind farms using a hybrid deep learning model with spatiotemporal clustering and feature mining","authors":"Mingyue Zhang , Yang Han , Yongchao Sun , Huaiyu Zhang , Fan Yang , Congling Wang","doi":"10.1016/j.measurement.2026.120383","DOIUrl":"10.1016/j.measurement.2026.120383","url":null,"abstract":"<div><div>Accurate ultra-short-term forecasting of wind power is crucial for the reliable integration of renewable energy. This study proposes a comprehensive forecasting framework that integrates spatial preprocessing, signal decomposition, hybrid deep learning, transfer learning, and interval prediction. An enhanced Kriging interpolation (EKI) method is first introduced to refine the spatial resolution of numerical weather prediction data. A two-stage clustering approach combining self-organizing maps (SOM) with K-means is employed to group wind farms with similar spatiotemporal characteristics, thereby facilitating knowledge transfer. Wind power series are subsequently decomposed through variational mode decomposition (VMD) optimized by an improved Aquila optimizer (IAO), while key input features are selected via Pearson correlation analysis. For each subsequence, a hybrid deep learning network integrating bidirectional temporal convolutional networks (BiTCN) and bidirectional gated recurrent units (BiGRU) with attention mechanism (AM) is constructed to forecast using multi-scale temporal features. Transfer learning is then applied to adapt the model to new sites. Finally, a heteroscedastic Gaussian process regression (HGPR) module is further employed to generate reliable interval forecasts. Case studies on 13 wind farms in Sichuan Province, China, evaluated against twelve comparative models from four perspectives, show that the framework improves forecasting precision, enhances uncertainty quantification, and reduces training time by over 90%.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120383"},"PeriodicalIF":5.6,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980030","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 : 2026-01-10DOI: 10.1016/j.measurement.2026.120370
Delu Che , Liangchun Hua , Pengyu Hou , Li-Ta Hsu , Fei Ye , Baocheng Zhang
Receiver–satellite-dependent code biases, referred to here as ambiguity-like code biases (ALCBs) due to their formulation being comparable to that of carrier-phase ambiguities, represent major error sources in Global Navigation Satellite System (GNSS) observation equations and can cause several-meter ranging errors. These biases are traditionally regarded as stable over the long term and are calibrated using post-processed estimates; however, such calibration becomes invalid when the biases vary in real-time data processing. To address this limitation, this study proposes a new single-frequency ionosphere-weighted (IW) real-time kinematic (RTK) model that explicitly parameterizes and estimates ALCBs. The proposed model not only facilitates the extraction of ALCBs for calibration purposes but also enables their real-time estimation, ensuring robustness even when ALCBs exhibit variations. The single-frequency case is emphasized in this study because it allows a clearer isolation and analysis of ALCB impacts and is more susceptible to errors than multi-frequency cases. Four baselines, including two ultra-short and two medium-long baselines, are selected to evaluate the performance of the proposed model. ALCBs extracted from the two ultra-short baselines exhibit discontinuities, which render calibration approaches ineffective for real-time data processing and further underscore the necessity of real-time estimation. Positioning experiments on the four baselines demonstrate that the proposed model significantly improves positioning performance compared with the traditional model that neglects ALCBs. For the 30 km baseline, the proposed model improves the accuracy of the ambiguity-fixed positioning solution by 22.97%, 17.24%, and 44.25% in the East, North, and Up components, respectively, and shortens the time to first fix (TTFF) by 70.56%. In addition, the applicability of the proposed model to dual-frequency and multi-constellation cases is examined, further extending its practical scope.
{"title":"Single-frequency RTK positioning in the presence of ambiguity-like code biases","authors":"Delu Che , Liangchun Hua , Pengyu Hou , Li-Ta Hsu , Fei Ye , Baocheng Zhang","doi":"10.1016/j.measurement.2026.120370","DOIUrl":"10.1016/j.measurement.2026.120370","url":null,"abstract":"<div><div>Receiver–satellite-dependent code biases, referred to here as ambiguity-like code biases (ALCBs) due to their formulation being comparable to that of carrier-phase ambiguities, represent major error sources in Global Navigation Satellite System (GNSS) observation equations and can cause several-meter ranging errors. These biases are traditionally regarded as stable over the long term and are calibrated using post-processed estimates; however, such calibration becomes invalid when the biases vary in real-time data processing. To address this limitation, this study proposes a new single-frequency ionosphere-weighted (IW) real-time kinematic (RTK) model that explicitly parameterizes and estimates ALCBs. The proposed model not only facilitates the extraction of ALCBs for calibration purposes but also enables their real-time estimation, ensuring robustness even when ALCBs exhibit variations. The single-frequency case is emphasized in this study because it allows a clearer isolation and analysis of ALCB impacts and is more susceptible to errors than multi-frequency cases. Four baselines, including two ultra-short and two medium-long baselines, are selected to evaluate the performance of the proposed model. ALCBs extracted from the two ultra-short baselines exhibit discontinuities, which render calibration approaches ineffective for real-time data processing and further underscore the necessity of real-time estimation. Positioning experiments on the four baselines demonstrate that the proposed model significantly improves positioning performance compared with the traditional model that neglects ALCBs. For the 30 km baseline, the proposed model improves the accuracy of the ambiguity-fixed positioning solution by 22.97%, 17.24%, and 44.25% in the East, North, and Up components, respectively, and shortens the time to first fix (TTFF) by 70.56%. In addition, the applicability of the proposed model to dual-frequency and multi-constellation cases is examined, further extending its practical scope.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"266 ","pages":"Article 120370"},"PeriodicalIF":5.6,"publicationDate":"2026-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145969257","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 : 2026-01-10DOI: 10.1016/j.measurement.2026.120402
Yifei Ding , Yudong Cao , Qiuhua Miao , Xiaoli Zhao , Yang Ge , Chaobin Hu
Online monitoring of critical equipment is essential for modern industrial systems. One-class learning has shown promise for intelligent fault detection, but two key challenges remain: (a) enabling unsupervised online detection with few samples, and (b) ensuring robustness against unknown anomalies that may distort decision boundaries. To address these issues, this paper proposes an unsupervised few-shot one-class classification (UFOCC) framework for online incipient fault detection of rolling bearings. The framework incorporates a self-calibration contrast (SCC) module, designed to adaptively regulate uncertain predictions and mitigate the adverse impact of anomaly contamination, and a local anomaly augmentation (LAA) module, which enriches normality representation under limited data through customized perturbations. By integrating SCC and LAA, the proposed UFOCC framework facilitates contamination-resilient and anomaly-aware one-class learning, leading to a more stable and robust characterization of normal operating behavior. Comprehensive experiments conducted on multiple run-to-failure bearing datasets demonstrate that UFOCC can accurately detect the onset of incipient faults during the degradation process of critical components, achieving a 5.9%–17.4% improvement in advance warning capability. Furthermore, it showcases substantial advantages in detection accuracy, training efficiency, and response speed.
{"title":"UFOCC: Contamination-Robust few-shot one-class learning for online detection of Bearing incipient faults","authors":"Yifei Ding , Yudong Cao , Qiuhua Miao , Xiaoli Zhao , Yang Ge , Chaobin Hu","doi":"10.1016/j.measurement.2026.120402","DOIUrl":"10.1016/j.measurement.2026.120402","url":null,"abstract":"<div><div>Online monitoring of critical equipment is essential for modern industrial systems. One-class learning has shown promise for intelligent fault detection, but two key challenges remain: (a) enabling unsupervised online detection with few samples, and (b) ensuring robustness against unknown anomalies that may distort decision boundaries. To address these issues, this paper proposes an unsupervised few-shot one-class classification (UFOCC) framework for online incipient fault detection of rolling bearings. The framework incorporates a self-calibration contrast (SCC) module, designed to adaptively regulate uncertain predictions and mitigate the adverse impact of anomaly contamination, and a local anomaly augmentation (LAA) module, which enriches normality representation under limited data through customized perturbations. By integrating SCC and LAA, the proposed UFOCC framework facilitates contamination-resilient and anomaly-aware one-class learning, leading to a more stable and robust characterization of normal operating behavior. Comprehensive experiments conducted on multiple run-to-failure bearing datasets demonstrate that UFOCC can accurately detect the onset of incipient faults during the degradation process of critical components, achieving a 5.9%–17.4% improvement in advance warning capability. Furthermore, it showcases substantial advantages in detection accuracy, training efficiency, and response speed.</div></div>","PeriodicalId":18349,"journal":{"name":"Measurement","volume":"265 ","pages":"Article 120402"},"PeriodicalIF":5.6,"publicationDate":"2026-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979878","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