IEEE Sensors Journal最新文献

{"title":"Dual-Branch Feature Learning and Hierarchical Domain Alignment for Multisource Adaptation Fault Diagnosis","authors":"Chenhui Zhao;Jiaming Wang;WaiYie Leong;Feifan Zhang;Jingya Wang","doi":"10.1109/JSEN.2026.3724354","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3724354","url":null,"abstract":"To address the significant distribution discrepancies among multiple source domains and the inefficient utilization of heterogeneous source information, a multisource adaptation fault diagnosis framework based on dual-branch feature learning and hierarchical domain alignment is proposed. A shared feature extractor and multiple domain-specific feature extractors are constructed to learn transferable common representations and complementary source-dependent discriminative information, respectively. The two branches are independently parameterized and guided by different optimization objectives, enabling the model to balance cross-domain transferability and source-domain individuality without assuming strict statistical disentanglement between the resulting feature spaces. A two-stage hierarchical domain alignment mechanism is further developed. The first stage performs intersource alignment in the shared feature space to reduce distribution discrepancies among multiple source domains, while the second stage conducts source–target alignment through domain-specific branches to enhance target-oriented feature transfer. In addition, a multiclassifier fusion strategy guided by the maximum-entropy principle is introduced to adaptively aggregate predictions from multiple sources, improving model stability under complex conditions and enhancing the recognition of uncertain samples. Experimental results demonstrate that the proposed method achieves superior diagnostic accuracy in multisource domain adaptation tasks and exhibits strong potential for practical engineering applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30737-30745"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"State-of-Health Estimation of Electric Vehicle Batteries From Multisource Onboard Sensor Data Using RGMamba","authors":"Chaolong Zhang;Kaixin Cheng;Kui Shao;Heng Li","doi":"10.1109/JSEN.2026.3732700","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3732700","url":null,"abstract":"Accurate state-of-health (SOH) estimation is essential for the safe operation and maintenance of electric vehicles. Battery management systems (BMSs) continuously record multisource onboard sensor signals, but measurement noise and varying operating conditions make it difficult to extract reliable indicators of battery degradation. This article proposes an SOH estimation framework that integrates cross-scale features extracted from multisource onboard data with a residual gating Mamba (RGMamba) network. During charging, incremental capacity features are extracted from cell-voltage and pack-current measurements to characterize battery aging. During discharging, vehicle-speed data are used to derive acceleration-based user-behavior features while charging temperature and accumulated mileage are incorporated to describe thermal and usage conditions. A feature-gating module adaptively weights the cross-scale features, the Mamba module captures long-term dependencies in battery degradation, and a residual module refines the nonlinear relationship between the learned features and SOH. The framework is evaluated using real-world operating data from seven electric vehicles, with the reference SOH derived from onboard state-of-charge (SOC) and current measurements. For the benchmark vehicle, RGMamba achieves a root-mean-square error (RMSE) of 0.0474 and a mean absolute percentage error (MAPE) of 4.40%. After fine-tuning with the first 20% of each target vehicle’s data, the MAPE values for the other six vehicles range from 3.17% to 6.54%. These results indicate that feature-level fusion of multisource onboard data, combined with target-vehicle fine-tuning, can support reference SOH estimation across vehicles equipped with the same battery system.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"31238-31248"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176164","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Personalized Blood Pressure Estimation With a Topological Prior Vector From Minimal Per-Subject Data","authors":"Myung-Kyu Yi;Jeyeon Lee;Jongshill Lee;In Young Kim","doi":"10.1109/JSEN.2026.3730879","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3730879","url":null,"abstract":"Cuffless blood pressure (BP) estimation from photoplethysmography (PPG) signals using deep learning (DL) has attracted significant attention for continuous, noninvasive health monitoring. However, traditional machine learning relies on handcrafted features that capture only limited aspects of the signal. While DL can learn richer representations from data, it often requires large-scale training to achieve stable generalization across subjects and conditions and produces models with limited interpretability, hindering clinical adoption in safety-critical applications. To address these issues, we propose a framework that integrates subject-specific topological priors derived from persistent homology of PPG morphology. We construct a 6-D topological prior vector (TPV) that summarizes per-beat global morphology and inject it into a dual-branch CNN via TPV-conditioned temporal gating, emphasizing waveform segments that are informative for BP. We evaluate our model on the MIMIC-III PPG dataset. Even when trained from scratch using only a small number of subject-specific segments, it achieves mean absolute errors (MAEs) of <inline-formula> <tex-math>${3}.{82} {,}~pm {,}~{0}.{53}$ </tex-math></inline-formula> mmHg for systolic BP (SBP) and <inline-formula> <tex-math>${2}.{63} {,}~pm {,}~{0}.{41}$ </tex-math></inline-formula> mmHg for diastolic BP (DBP), demonstrating strong performance under limited-data conditions. Our approach embeds morphological priors tailored to each subject. This design significantly improves both data efficiency and personalization, enabling more practical BP estimation in real-world scenarios.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30647-30662"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Six-Axis Force/Torque Sensor Based on Fiber Bragg Grating and Y-Shaped Crossbeam for Robotic Applications","authors":"Jun Huang;Zhirun Chen;Yu Peng;Yixin Hu;Mengying He","doi":"10.1109/JSEN.2026.3725417","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3725417","url":null,"abstract":"Six-axis force/torque (F/T) sensors enable robots, including industrial, collaborative, and humanoid robots, to perceive 6-D forces and torques for complex control and human–robot interaction. In this article, a six-axis F/T sensor utilizing fiber Bragg grating (FBG) and a Y-shaped crossbeam is presented for robotic applications. The sensing principles for six-axis forces and torques were investigated through theoretical calculations and finite element analysis (FEA). A Timoshenko beam model was used to analyze the strain of the Y-shaped crossbeam under six-axis forces and torques. A sensor prototype was developed, and calibration experiments (CEs) were conducted. Decoupling methods were investigated using back-propagation neural networks optimized by the sparrow search algorithm (SSA-BP), least squares (LS), and particle swarm optimization-BP neural networks (BP-PSO). Results show that the SSA-BP decoupling method has the lowest decoupling errors, with maximum type-I and type-II errors of 1.29% FS and 1.85% FS, respectively. The sensor’s maximum repeatability and hysteresis errors are 1.98% FS and 1.89% FS, respectively. The proposed six-axis F/T sensor has potential for robotic applications in harsh environments, including those with intense electromagnetic interference and the presence of explosive and flammable substances.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30350-30361"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176190","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Breathing to Heartbeat: Virtual Acoustic Enhancement for Cardiac Spike Extraction via Smartphone","authors":"Kounkou Vincent;Xiyue Tan;Ke Sun;Lyusha Cheng;Tao Gong","doi":"10.1109/JSEN.2026.3672895","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3672895","url":null,"abstract":"Continuous monitoring of vital signs is crucial for daily health management, enabling the early detection of conditions such as premature heartbeats (HBs), atrial fibrillation, and cardiac hyperactivity. However, subtle chest movements associated with cardiac activity are often obscured by respiratory artifacts and multipath interference, presenting significant challenges in tracking minute cardiac-induced vibrations using conventional acoustic sensing. This article introduces acoustic-augmented (AcAg), a novel noncontact monitoring framework that leverages smartphone-generated acoustic signals to extract cardiac signatures from respiratory-modulated chest movements. Our key innovation lies in virtual acoustic enhancement (VAcE), a paradigm-shifting approach that mathematically reconfigures multipath propagation to transform destructive interference into constructive patterns, revealing previously obscured cardiac features. We develop a comprehensive signal processing pipeline that combines this virtual enhancement with phase-coherent analysis to capture submillimeter cardiac displacements. Furthermore, we leverage AcAgNet, a variational autoencoder that corrects signal artifacts and recovers fine-grained waveform characteristics while providing a precise heart rate (Hr) estimate. Experimental results demonstrate a mean absolute error of 0.3431 breaths per minute (bpm) across our datasets, establishing a significant advancement over state-of-the-art acoustic methods for noncontact cardiac monitoring.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30671-30681"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Improving Speech Recognition in Noisy Environments With RF–Microphone Fusion for Next-Generation Hearing Aids","authors":"Michaela Reay;Hira Hameed;Balal Saleemi;Aisha Fatima;Muhammad Ali Imran;Qammer H. Abbasi","doi":"10.1109/JSEN.2026.3727324","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3727324","url":null,"abstract":"Hearing loss affects over 430 million people worldwide and is expected to exceed 700 million by 2050. While microphone-based hearing aids are widely used, their performance degrades in acoustically noisy environments, reducing speech understanding and leading to device abandonment. This work presents a proof-of-concept (PoC) multimodal hearing aid combining microphone and radio frequency (RF) speech sensing, where RF provides complementary speech information in noisy conditions. A ResNet18 network was used to classify speech samples across three dataset conditions: noise-less, postadded-noise, and live-added-noise. These datasets were created to evaluate model performance under baseline and acoustically degraded conditions, and to validate the postadded-noise modeling approach. The model was trained in three modality configurations: RF-only, microphone-only, and multimodal RF-microphone fusion. RF-only test accuracy was 93.92%. Noise-less microphone-only accuracy was 98.90%, but decreased progressively under postadded-noise conditions as the signal-to-noise ratio (SNR) was reduced, achieving 97.51% at 0 dB, 91.28% at −3 dB, 77.17% at −6 dB, and 24.89% at −9 dB. In contrast, multimodal accuracy was 99.80% under noise-less conditions and remained robust under postadded-noise conditions, achieving 94.19% at 0 dB, 92.15% at −3 dB, 91.05% at −6 dB, and 90.00% at −9 dB. For the live-added-noise dataset (LAND), evaluated only at −9 dB SNR, microphone-only accuracy was 23.07%, while multimodal accuracy was 83.84%. These results show that RF sensing enables robust speech classification in acoustically noisy environments and that postadded-noise datasets (PANDs) can effectively approximate live noisy speech conditions. Overall, the findings demonstrate the promise of RF-microphone multimodal fusion for next-generation noise-robust hearing aids, with the potential to improve speech intelligibility in challenging listening environments and reduce the user dissatisfaction and device abandonment associated with current-generation microphone-based systems.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"31320-31330"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176275","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-Time Thermal Infrared Object Tracking via Efficient Consistency-Reasoning Correlation Filters","authors":"Yueping Huang;Zehao Li;Ruitao Lu;Xiaofeng Li;Xiaogang Yang","doi":"10.1109/JSEN.2026.3728852","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3728852","url":null,"abstract":"Discriminative correlation filter (DCF)-based object tracking algorithms, characterized by their favorable tracking performance, high computational efficiency, and low memory consumption, have demonstrated significant appeal in the field of thermal infrared (TIR) object tracking. One of the core components is to learn a robust correlation filter. However, most existing approaches fail to adequately exploit historical information during the correlation filter learning process, resulting in model degradation in complex scenarios. To address this limitation, we propose a consistency reasoning method for correlation filter learning. In particular, by introducing forward–backward cycle consistency regularization and filter temporal consistency regularization, we establish a temporal reasoning mechanism across consecutive frames, enabling the effective utilization of historical information to train a robust correlation filter. Moreover, a dynamic threshold adaptive update strategy, designed based on historical response information, is proposed to further enhance the filter robustness. Extensive experiments on three widely used datasets demonstrate that the proposed method achieves competitive tracking performance with a running speed exceeding 60 fps, satisfying real-time tracking requirements.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30984-30995"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176293","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multichannel Radar Forward-Looking Space–Time Super-Resolution Imaging Considering Motion Error Estimation and Compensation","authors":"Wenchao Li;Zhe Zhang;Rui Chen;Yuhuan Han;Kun Zhang;Junjie Wu;Jianyu Yang","doi":"10.1109/JSEN.2026.3732669","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3732669","url":null,"abstract":"With coherent processing along the synthetic aperture dimension, multichannel radar offers the potential for high-resolution forward-looking imaging. However, it still faces the problem of poor angular resolution along the trajectory area and sensitivity to motion errors. In this article, a space–time super-resolution (SR) scheme including platform error estimation and compensation for multichannel radar forward-looking imaging is proposed. In the scheme, the procedure of space–time SR under ideal conditions is formulated first according to the signal model of multichannel radar forward-looking imaging. Then, based on the characteristic that motion errors will cause phase differences in the spatial-domain imaging results of the same target at different times, a linear system of equations is established, and the velocity errors and accelerations are estimated by solving the equations with the weighted least squares (WLS) method. Finally, these estimated values are used to correct the space–time steering matrix, and then SR processing is performed with the sparse Bayesian learning (SBL) algorithm. Both simulated and measured data are provided to validate the effectiveness of the proposed scheme in error estimation and imaging performance.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"31227-31237"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Two-Stage Adaptive Signal Decomposition and Noise Deconvolution Framework for Compound Fault Diagnosis of Rolling Bearings","authors":"Xiaoli Zhang;Guanju Huang;Wenbin Gong;Benkang Fu;Xin Luo","doi":"10.1109/JSEN.2026.3727484","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3727484","url":null,"abstract":"Rolling bearings are key components of rotating machinery, and compound faults usually contain coupled fault features and strong interference, which makes feature extraction difficult. To address this problem, this article proposes a two-stage compound fault diagnosis method based on feature mode decomposition optimized by adaptive sand cat swarm optimization (ASCSO-FMD) and adaptive minimum noise amplitude deconvolution (AMNAD). ASCSO is first used to optimize the key feature mode decomposition (FMD) parameters and decompose the compound fault signal into mode components. A feature evaluation index is then used to select components with significant fault information. For components with coupled or aliased features, AMNAD is further applied to enhance periodic impulses. In addition, the envelope harmonic product spectrum (EHPS) and the Gini index of the squared envelope spectrum (GISES) are used to adaptively determine the fault characteristic frequency and filter length of AMNAD. Simulation and measured compound fault analyses show that the proposed method can enhance and separate inner-race, outer-race, and rolling-element fault features under constant-speed conditions.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30870-30879"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149175969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radiation-Hardening Techniques and In-Orbit Robustness Validation for Trapezoidal Dual-Axis Sun Sensors in LEO Environments","authors":"Mang Ou-Yang;Ching-I Tai;Guan-Yu Huang;Tse-Yu Cheng;Chang-Hsun Liu;Yu-Siou Liu;Jin-Chern Chiou;Chen-Yu Chan;Tung-Yun Hsieh;Chen-Tsung Lin;Ying-Wen Jan;Chih-Hsun Lin;Yung-Jhe Yan","doi":"10.1109/JSEN.2026.3727622","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3727622","url":null,"abstract":"This work presents a sensor-level and algorithmic strategy for limiting radiation-induced angular drift in a previously developed trapezoidal dual-axis sun sensor without requiring dedicated radiation-hardened components. Radiation-induced degradation of the photodiodes is decomposed into additive dark-current offsets and multiplicative responsivity losses. To evaluate the proposed compensation strategy, the sensors were irradiated with 30-MeV protons to a cumulative dose exceeding <inline-formula> <tex-math>$mathrm {4735~krad(Si)}$ </tex-math></inline-formula>. This dose exceeds typical LEO mission requirements, enabling evaluation over an extended degradation range. Unbiased sensors were reserved for post-irradiation characterization, while biased units were operated in situ to collect photocurrent data. Experimental results show a 60-fold increase in dark current and a 30 % reduction in responsivity. Under nonirradiated conditions, the sensor had previously demonstrated angular errors within ±0.3°. Following proton irradiation, the uncompensated maximum angular deviation reached 4.8°. An embedded shielded dummy photodiode was used to compensate for dark-current drift, while ratio-based angular reconstruction was used to normalize responsivity degradation. Together, these methods reduced the maximum angular deviation to 0.53°. Low-dose-rate ⁶⁰Co gamma irradiation up to <inline-formula> <tex-math>$mathrm {30~krad(Si)}$ </tex-math></inline-formula> was further conducted to evaluate enhanced low-dose-rate sensitivity in the readout electronics, where common-mode gain drift was effectively suppressed through both circuit-topology selection and ratio-based cancellation. The system also maintained continuous functional operation throughout 221-MeV proton irradiation, with no destructive single-event latch-up observed within the monitored parameters. These results indicate that the proposed sensor-level and algorithmic approach has the potential to provide a viable, low-cost radiation-tolerant solution for long-duration LEO missions on resource-constrained small-satellite platforms.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 20","pages":"30438-30449"},"PeriodicalIF":4.5,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149175970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}