Pub Date : 2025-02-13DOI: 10.1109/TIM.2025.3538084
Zhihong Chen;Jun Zhu
Combat simulation has become crucial in military assessment in recent years due to the rapid development of information technology and artificial intelligence. However, the increasingly large volume of data and the credibility of information have heightened the difficulty of processing and analysis, subsequently affecting the quality of military decision-making. Extracting key features can clarify the core factors influencing the indicators, simplify model complexity, improve prediction accuracy, and save computation time. Quantifying uncertainty helps enhance decision quality and increases the system’s adaptability in uncertain environments. Accordingly, we propose a novel method for key feature selection and interval prediction to address specific regression tasks in combat simulation systems. First, our approach comprehensively considers the importance of features to the target variable, the interaction between features, and redundancy by integrating various feature selection methods, thereby precisely extracting key features. Second, we modify the output structure of traditional neural networks and design a new hybrid loss function to train the model. Furthermore, deep ensemble methods are utilized to enhance diversity and robustness, thus enabling uncertainty evaluation and interval prediction. The experimental results indicate that, after feature selection, the estimation achieved a mean squared error (mse) of only 0.151 and a prediction interval coverage probability (PICP) of 86.99%, providing crucial support for military decision-making.
{"title":"Intelligent Inference in Combat Simulation Systems Based on Key Feature Extraction and Uncertainty Interval Estimation","authors":"Zhihong Chen;Jun Zhu","doi":"10.1109/TIM.2025.3538084","DOIUrl":"https://doi.org/10.1109/TIM.2025.3538084","url":null,"abstract":"Combat simulation has become crucial in military assessment in recent years due to the rapid development of information technology and artificial intelligence. However, the increasingly large volume of data and the credibility of information have heightened the difficulty of processing and analysis, subsequently affecting the quality of military decision-making. Extracting key features can clarify the core factors influencing the indicators, simplify model complexity, improve prediction accuracy, and save computation time. Quantifying uncertainty helps enhance decision quality and increases the system’s adaptability in uncertain environments. Accordingly, we propose a novel method for key feature selection and interval prediction to address specific regression tasks in combat simulation systems. First, our approach comprehensively considers the importance of features to the target variable, the interaction between features, and redundancy by integrating various feature selection methods, thereby precisely extracting key features. Second, we modify the output structure of traditional neural networks and design a new hybrid loss function to train the model. Furthermore, deep ensemble methods are utilized to enhance diversity and robustness, thus enabling uncertainty evaluation and interval prediction. The experimental results indicate that, after feature selection, the estimation achieved a mean squared error (mse) of only 0.151 and a prediction interval coverage probability (PICP) of 86.99%, providing crucial support for military decision-making.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-12"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465717","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 : 2025-02-13DOI: 10.1109/TIM.2025.3541800
Jiancheng Yin;Wentao Sui;Xuye Zhuang;Yunlong Sheng;Jianjun Wang;Rujun Song;Yongbo Li
Lempel-Ziv (LZ) complexity has been widely applied in multiple fields, and there are numerous improvements in multiscale computation and encoding to enhance its ability to characterize signal changes. Based on the hierarchical analysis, this article proposes an improved LZ indicator based on multiscale decomposition and multiscale encoding, which is applied to the recognition of bearing failure severity. The signal is first decomposed into multiple scales through hierarchical analysis. Next, the decomposed node signal is further decomposed by coarse-grained methods. Then, the multiscale decomposed signal is further decomposed into low- and high-frequency components using hierarchical analysis and the multiscale encoding is performed based on the decomposed low- and high-frequency components. Finally, the LZ complexity is calculated based on multiscale encoding. The effectiveness of the proposed method is validated by three single-point bearing fault datasets with different failure severity. The proposed method can achieve a classification accuracy of over 97%. The proposed method can be effectively applied to classify the bearing failure severity.
{"title":"An Improved Lempel–Ziv Complexity Indicator Based on Multiscale Decomposition and Multiscale Encoding for Bearing Failure Severity Recognition","authors":"Jiancheng Yin;Wentao Sui;Xuye Zhuang;Yunlong Sheng;Jianjun Wang;Rujun Song;Yongbo Li","doi":"10.1109/TIM.2025.3541800","DOIUrl":"https://doi.org/10.1109/TIM.2025.3541800","url":null,"abstract":"Lempel-Ziv (LZ) complexity has been widely applied in multiple fields, and there are numerous improvements in multiscale computation and encoding to enhance its ability to characterize signal changes. Based on the hierarchical analysis, this article proposes an improved LZ indicator based on multiscale decomposition and multiscale encoding, which is applied to the recognition of bearing failure severity. The signal is first decomposed into multiple scales through hierarchical analysis. Next, the decomposed node signal is further decomposed by coarse-grained methods. Then, the multiscale decomposed signal is further decomposed into low- and high-frequency components using hierarchical analysis and the multiscale encoding is performed based on the decomposed low- and high-frequency components. Finally, the LZ complexity is calculated based on multiscale encoding. The effectiveness of the proposed method is validated by three single-point bearing fault datasets with different failure severity. The proposed method can achieve a classification accuracy of over 97%. The proposed method can be effectively applied to classify the bearing failure severity.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-13"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465557","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}
Fault reconstruction, a common approach for fault diagnosis, involves extracting fault subspaces from measurement data and using them to identify the type of online fault samples. These subspaces, however, are often challenged by insufficient fault representation and weak discriminability among different fault types. To address these issues, a semisupervised kernel independent component analysis (SsKICA) algorithm is proposed and applied in fault diagnosis. First, an innovative feature extraction method is developed that considers both the non-Gaussianity and discriminability of the features in parallel. Its objective function is reformulated into a dual-maximization criterion function that incorporates the supervised information of sample categories and independence into the same mathematical frame. Second, Newton’s methods and fixed-point iteration are employed to derive iterative solutions for this dual-maximization function, and the convergence of these iterative solutions is analyzed. Third, a fine-grained fault subspace extraction method is investigated by extending fault reconstruction to the Hilbert space. Finally, a complete fault diagnosis strategy based on SsKICA is designed that can provide feedback for unseen faults and easily interpretable diagnostics. The simulation results, on the Tennessee Eastman process (TEP) and the PROcess Networks Optimization (PRONTO), demonstrate that the proposed method provides effective feedback mechanisms for unseen faults and enhanced diagnostic accuracy for known faults.
{"title":"Semisupervised Kernel Independent Component Analysis and Its Application for Fault Diagnosis","authors":"Xiangyu Kong;Meizhi Liu;Qi Zhang;Jiayu Luo;Chen Zhang","doi":"10.1109/TIM.2025.3541712","DOIUrl":"https://doi.org/10.1109/TIM.2025.3541712","url":null,"abstract":"Fault reconstruction, a common approach for fault diagnosis, involves extracting fault subspaces from measurement data and using them to identify the type of online fault samples. These subspaces, however, are often challenged by insufficient fault representation and weak discriminability among different fault types. To address these issues, a semisupervised kernel independent component analysis (SsKICA) algorithm is proposed and applied in fault diagnosis. First, an innovative feature extraction method is developed that considers both the non-Gaussianity and discriminability of the features in parallel. Its objective function is reformulated into a dual-maximization criterion function that incorporates the supervised information of sample categories and independence into the same mathematical frame. Second, Newton’s methods and fixed-point iteration are employed to derive iterative solutions for this dual-maximization function, and the convergence of these iterative solutions is analyzed. Third, a fine-grained fault subspace extraction method is investigated by extending fault reconstruction to the Hilbert space. Finally, a complete fault diagnosis strategy based on SsKICA is designed that can provide feedback for unseen faults and easily interpretable diagnostics. The simulation results, on the Tennessee Eastman process (TEP) and the PROcess Networks Optimization (PRONTO), demonstrate that the proposed method provides effective feedback mechanisms for unseen faults and enhanced diagnostic accuracy for known faults.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-14"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489238","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 : 2025-02-13DOI: 10.1109/TIM.2025.3541777
Yonghui Hu;Yi Li;Junkai Wang;Yong Yan
Indoor human localization is of great significance in a variety of applications, including navigation, healthcare, security, and many other location-based services. This article presents a passive indoor localization method that exploits the varying electric fields naturally generated by human activities. An array of electrostatic sensors capable of passive, long-range sensing is developed using charge amplifiers. Human localization is formulated as an inverse problem that aims to reconstruct the charge distribution within the target area from sensor measurements. The spatial sensitivity matrix is preprocessed using QR factorization, and then, compressive sensing is used to find the sparse solution. Experiments were conducted in an office environment of $4.2times 4.2$ m. Results obtained show that the localization accuracy is location-dependent and a median error less than 0.26 m has been achieved. Although the sensor signals are vulnerable to a variety of factors, the localization method exhibits strong robustness against environmental and subject changes.
{"title":"Indoor Human Localization Using Electrostatic Sensors and Compressive Sensing Techniques","authors":"Yonghui Hu;Yi Li;Junkai Wang;Yong Yan","doi":"10.1109/TIM.2025.3541777","DOIUrl":"https://doi.org/10.1109/TIM.2025.3541777","url":null,"abstract":"Indoor human localization is of great significance in a variety of applications, including navigation, healthcare, security, and many other location-based services. This article presents a passive indoor localization method that exploits the varying electric fields naturally generated by human activities. An array of electrostatic sensors capable of passive, long-range sensing is developed using charge amplifiers. Human localization is formulated as an inverse problem that aims to reconstruct the charge distribution within the target area from sensor measurements. The spatial sensitivity matrix is preprocessed using QR factorization, and then, compressive sensing is used to find the sparse solution. Experiments were conducted in an office environment of <inline-formula> <tex-math>$4.2times 4.2$ </tex-math></inline-formula> m. Results obtained show that the localization accuracy is location-dependent and a median error less than 0.26 m has been achieved. Although the sensor signals are vulnerable to a variety of factors, the localization method exhibits strong robustness against environmental and subject changes.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-10"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489259","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 : 2025-02-13DOI: 10.1109/TIM.2025.3533957
{"title":"2023 List of Reviewers","authors":"","doi":"10.1109/TIM.2025.3533957","DOIUrl":"https://doi.org/10.1109/TIM.2025.3533957","url":null,"abstract":"","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"73 ","pages":"1-124"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10884969","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143403795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-13DOI: 10.1109/TIM.2025.3534225
Kun Wu;Hongtao Zhang;Maosheng Hou;Jianli Zheng
The laser scanning projection systems scan to determine the position of the center of a cooperative target and then calculate the coordinate transformation between the projected surface and the projection instrument. This enables the laser projection system to accurately project the positions and contours of components. To solve the problem of ambient light interference during the calibration of a laser scanning projection system, this study proposes a detection module that uses correlated double sampling (CDS) technology. Experimental verification shows that this detection module can accurately identify whether the laser spot is located in the highly reflective region of the cooperative target, even when the signal interference ratio is −29.5 dB. In accordance with the features of the CDS technique, a noncontinuous scanning method was developed to quickly determine the center position of the cooperative target. The proposed methods were used to detect and scan cooperative targets at a distance of 5000 mm using a self-constructed laser scanning projection system. The results show that the noncontinuous scanning method reduces the number of sampling points by 97.5% compared with the raster scanning method, with a positioning deviation of 0.052 mm.
{"title":"Noncontinuous Scanning Method for Cooperative Targets Based on Correlated Double Sampling","authors":"Kun Wu;Hongtao Zhang;Maosheng Hou;Jianli Zheng","doi":"10.1109/TIM.2025.3534225","DOIUrl":"https://doi.org/10.1109/TIM.2025.3534225","url":null,"abstract":"The laser scanning projection systems scan to determine the position of the center of a cooperative target and then calculate the coordinate transformation between the projected surface and the projection instrument. This enables the laser projection system to accurately project the positions and contours of components. To solve the problem of ambient light interference during the calibration of a laser scanning projection system, this study proposes a detection module that uses correlated double sampling (CDS) technology. Experimental verification shows that this detection module can accurately identify whether the laser spot is located in the highly reflective region of the cooperative target, even when the signal interference ratio is −29.5 dB. In accordance with the features of the CDS technique, a noncontinuous scanning method was developed to quickly determine the center position of the cooperative target. The proposed methods were used to detect and scan cooperative targets at a distance of 5000 mm using a self-constructed laser scanning projection system. The results show that the noncontinuous scanning method reduces the number of sampling points by 97.5% compared with the raster scanning method, with a positioning deviation of 0.052 mm.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-9"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143455346","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 : 2025-02-13DOI: 10.1109/TIM.2025.3541784
Junfeng Chen;Hao Li;Ke Ai;Shuolong Zhu;Cunzheng Fan;Zhijun Yan;Qizhen Sun
Since the superiorities of large-scale and distributed measurement, the fiber-optic distributed acoustic sensing (DAS) technique can collect plenty of distributed acoustic signals. Therefore, it has been gradually applied to perimeter security applications recently. However, the multisource near-field 2-D and 3-D localization and signal enhancement without aliasing noise for large-scale DAS has not been achieved yet, which are highly necessary for large-scale perimeter security applications. So, in this work, the DAS assisted with near-field array signal processing (ASP) is further developed and demonstrated, in which the near-field multiple signal classification (NF-MUSIC) algorithm is developed to achieve the near-field multisource localization. Then, based on the obtained target source position, the minimum variance distortionless response beamformer (MVDR-BF) is further proposed to enhance target signal in DAS, including multisource aliasing noise. In lab test, the maximum localization error no more than 0.03 m and 0.3° is demonstrated by using the proposed ASP-assisted DAS. And the acoustic target tracking is also achieved in underwater field test. Then, the signal to noise ratio (SNR) of the target source in lab and underwater test is averagely enhanced about 9.7 and 16.8 dB by, respectively, using 5 and 20 channels. To the best of our knowledge, it is the first time to achieve the near-field multisource localization and signal enhancement for fiber-optic DAS. Substantially, this work has been greatly verified the proposed scheme that can further expand and promote the large-scale DAS practical applications, including the low-altitude and underwater multitarget detection, localization, and recognition.
{"title":"Near-Field Multisource Localization and Signal Enhancement With ASP-Assisted DAS","authors":"Junfeng Chen;Hao Li;Ke Ai;Shuolong Zhu;Cunzheng Fan;Zhijun Yan;Qizhen Sun","doi":"10.1109/TIM.2025.3541784","DOIUrl":"https://doi.org/10.1109/TIM.2025.3541784","url":null,"abstract":"Since the superiorities of large-scale and distributed measurement, the fiber-optic distributed acoustic sensing (DAS) technique can collect plenty of distributed acoustic signals. Therefore, it has been gradually applied to perimeter security applications recently. However, the multisource near-field 2-D and 3-D localization and signal enhancement without aliasing noise for large-scale DAS has not been achieved yet, which are highly necessary for large-scale perimeter security applications. So, in this work, the DAS assisted with near-field array signal processing (ASP) is further developed and demonstrated, in which the near-field multiple signal classification (NF-MUSIC) algorithm is developed to achieve the near-field multisource localization. Then, based on the obtained target source position, the minimum variance distortionless response beamformer (MVDR-BF) is further proposed to enhance target signal in DAS, including multisource aliasing noise. In lab test, the maximum localization error no more than 0.03 m and 0.3° is demonstrated by using the proposed ASP-assisted DAS. And the acoustic target tracking is also achieved in underwater field test. Then, the signal to noise ratio (SNR) of the target source in lab and underwater test is averagely enhanced about 9.7 and 16.8 dB by, respectively, using 5 and 20 channels. To the best of our knowledge, it is the first time to achieve the near-field multisource localization and signal enhancement for fiber-optic DAS. Substantially, this work has been greatly verified the proposed scheme that can further expand and promote the large-scale DAS practical applications, including the low-altitude and underwater multitarget detection, localization, and recognition.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-11"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143489237","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 : 2025-02-13DOI: 10.1109/TIM.2025.3541778
Zhexin Cui;Haichuan Liu;Jiguang Yue;Chenhao Wu
Electrohydraulic actuators (EHAs) are one of the most widely applied high-power-density equipments in mechatronic systems. Anomaly detection is essential for EHA prototypes to avoid potential design errors and ensure the reliability of the final design. However, current data-driven anomaly detection methods rely on extensive previous samples under complete design anomaly modes. This premise is impractical in industry applications, where sufficient intrusive physical anomaly experiment data may not be available or introduce unaffordable design costs. This article develops a twin data-driven design anomaly detection method to address the aforementioned problem. First, digital twins (DTs) of the EHA system are established to broadly simulate dynamic responses to design anomalies. Physics-informed parameter estimation ensures twin model fidelity and data availability. Besides, a twin data-driven domain-adversarial long short-term memory (LSTM) network (TD-DALN) is proposed to facilitate domain-invariant and discriminative feature extraction and cross-domain knowledge transfer for accurate design anomaly classification. Correspondingly, domain reconstruction is designed to bridge initial distribution differences between virtual and physical domains caused by the imbalance of anomaly and dynamic conditions. The experimental results demonstrate the effectiveness of the proposed method and its advantages over the competitors.
{"title":"TD-DALN: A Twin Data-Driven Design Anomaly Detection Method for Electrohydraulic Actuators","authors":"Zhexin Cui;Haichuan Liu;Jiguang Yue;Chenhao Wu","doi":"10.1109/TIM.2025.3541778","DOIUrl":"https://doi.org/10.1109/TIM.2025.3541778","url":null,"abstract":"Electrohydraulic actuators (EHAs) are one of the most widely applied high-power-density equipments in mechatronic systems. Anomaly detection is essential for EHA prototypes to avoid potential design errors and ensure the reliability of the final design. However, current data-driven anomaly detection methods rely on extensive previous samples under complete design anomaly modes. This premise is impractical in industry applications, where sufficient intrusive physical anomaly experiment data may not be available or introduce unaffordable design costs. This article develops a twin data-driven design anomaly detection method to address the aforementioned problem. First, digital twins (DTs) of the EHA system are established to broadly simulate dynamic responses to design anomalies. Physics-informed parameter estimation ensures twin model fidelity and data availability. Besides, a twin data-driven domain-adversarial long short-term memory (LSTM) network (TD-DALN) is proposed to facilitate domain-invariant and discriminative feature extraction and cross-domain knowledge transfer for accurate design anomaly classification. Correspondingly, domain reconstruction is designed to bridge initial distribution differences between virtual and physical domains caused by the imbalance of anomaly and dynamic conditions. The experimental results demonstrate the effectiveness of the proposed method and its advantages over the competitors.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-15"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465558","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}
A parallel Fabry-Perot interferometers (FPIs) fiber-optic sensor based on magnetic fluid (MF) and Vernier effect is proposed. The proposed sensor consists of two parallel FPIs, fabricated by splicing a section of hole-assisted one-core fiber (HAOCF) onto an Au-plated single-mode fiber (APSMF). The side holes of the HAOCF in two FPIs are filled with MF and polydimethylsiloxane, respectively. This configuration allows the sensor to utilize the Vernier effect to enhance magnetic field detection sensitivity while achieving temperature self-compensation within the operating range. Additionally, the asymmetric structure of the sensor produces different spectral responses to varying magnetic field directions. The spectral data under different magnetic field directions are collected and used to train the designed convolutional neural network (CNN). Combined with the trained CNN, the sensor overcomes the limitations of traditional wavelength demodulation methods and realizes the accurate identification of magnetic field direction in the range of 0°–360°. Experimental results show that the magnetic field sensitivity of the sensor reaches −1.27 nm/mT within the 0–7-mT range, which is 5.52 times higher than that of FPI1. The temperature crosstalk of the sensor is only $5.08times 10^{-3}$ mT/°C, a reduction of 15.87 times compared to FPI1. The sensor achieved a prediction error of less than 0.37° for the magnetic field direction on the testing dataset. This work offers a novel methodology for optical fiber sensing in vector magnetic field detection applications.
{"title":"A Temperature Self-Compensating Fiber-Optic Fabry–Perot Sensor for High-Sensitive Vector Magnetic Field Measurement","authors":"Rui Pan;Chaopeng Wang;Wenlong Yang;Ji Liu;Liuyang Zhang;Shuang Yu;Haibin Wu;Mingze Zhang","doi":"10.1109/TIM.2025.3541650","DOIUrl":"https://doi.org/10.1109/TIM.2025.3541650","url":null,"abstract":"A parallel Fabry-Perot interferometers (FPIs) fiber-optic sensor based on magnetic fluid (MF) and Vernier effect is proposed. The proposed sensor consists of two parallel FPIs, fabricated by splicing a section of hole-assisted one-core fiber (HAOCF) onto an Au-plated single-mode fiber (APSMF). The side holes of the HAOCF in two FPIs are filled with MF and polydimethylsiloxane, respectively. This configuration allows the sensor to utilize the Vernier effect to enhance magnetic field detection sensitivity while achieving temperature self-compensation within the operating range. Additionally, the asymmetric structure of the sensor produces different spectral responses to varying magnetic field directions. The spectral data under different magnetic field directions are collected and used to train the designed convolutional neural network (CNN). Combined with the trained CNN, the sensor overcomes the limitations of traditional wavelength demodulation methods and realizes the accurate identification of magnetic field direction in the range of 0°–360°. Experimental results show that the magnetic field sensitivity of the sensor reaches −1.27 nm/mT within the 0–7-mT range, which is 5.52 times higher than that of FPI1. The temperature crosstalk of the sensor is only <inline-formula> <tex-math>$5.08times 10^{-3}$ </tex-math></inline-formula> mT/°C, a reduction of 15.87 times compared to FPI1. The sensor achieved a prediction error of less than 0.37° for the magnetic field direction on the testing dataset. This work offers a novel methodology for optical fiber sensing in vector magnetic field detection applications.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-8"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143471817","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 : 2025-02-13DOI: 10.1109/TIM.2025.3538086
Xingbin Shi;Haiyan Wang;Baojiang Li;Yuxin Qin;Cheng Peng;Yifan Lu
Brain–computer interface (BCI) is an important way of human-computer interaction, with the ability to monitor brain states, and it has become an increasingly significant research direction. Single-modal noninvasive brain signals have limitations, such as low spatial resolution or low temporal resolution, while multimodal brain signal acquisition and processing can overcome these limitations. Electroencephalogram and functional near-infrared spectroscopy (EEG-fNIRS) is a method with advantages in multimodal brain signal processing, but current fusion methods mostly use manual feature extraction or channel selection, which may lead to the loss of important information during the feature extraction or channel selection process in real-time BCI systems. In order to solve this issue, this article proposes an innovative fusion analysis method for EEG-fNIRS multimodal brain signals, using a hybrid algorithm that combines convolutional neural network (CNN) and Attention mechanisms for signal classification. The method first preprocesses the EEG and fNIRS signals separately, then extracts features using spatial-temporal convolutional layers, and finally merges them to obtain the classification results through dual attention calculation. Our method is validated on two publicly available mixed EEG-fNIRS BCI datasets, including three types of experimental tasks that do not involve actual movement: motor imagery (MI), mental arithmetic, and word generation (WG). The accuracy rates for each task reached 92.2% for MI, 98.6% for mental arithmetic, and 95.2% for WG, respectively. These rates have surpassed all the current methods. This indicates that our proposed method achieves better classification performance in non-actual movement classification tasks under the premise of lightweight. The method proposed in this study can be applied to the field of rapid and efficient identification of brain signals.
{"title":"Fusion Analysis of EEG-fNIRS Multimodal Brain Signals: A Multitask Classification Algorithm Incorporating Spatial-Temporal Convolution and Dual Attention Mechanisms","authors":"Xingbin Shi;Haiyan Wang;Baojiang Li;Yuxin Qin;Cheng Peng;Yifan Lu","doi":"10.1109/TIM.2025.3538086","DOIUrl":"https://doi.org/10.1109/TIM.2025.3538086","url":null,"abstract":"Brain–computer interface (BCI) is an important way of human-computer interaction, with the ability to monitor brain states, and it has become an increasingly significant research direction. Single-modal noninvasive brain signals have limitations, such as low spatial resolution or low temporal resolution, while multimodal brain signal acquisition and processing can overcome these limitations. Electroencephalogram and functional near-infrared spectroscopy (EEG-fNIRS) is a method with advantages in multimodal brain signal processing, but current fusion methods mostly use manual feature extraction or channel selection, which may lead to the loss of important information during the feature extraction or channel selection process in real-time BCI systems. In order to solve this issue, this article proposes an innovative fusion analysis method for EEG-fNIRS multimodal brain signals, using a hybrid algorithm that combines convolutional neural network (CNN) and Attention mechanisms for signal classification. The method first preprocesses the EEG and fNIRS signals separately, then extracts features using spatial-temporal convolutional layers, and finally merges them to obtain the classification results through dual attention calculation. Our method is validated on two publicly available mixed EEG-fNIRS BCI datasets, including three types of experimental tasks that do not involve actual movement: motor imagery (MI), mental arithmetic, and word generation (WG). The accuracy rates for each task reached 92.2% for MI, 98.6% for mental arithmetic, and 95.2% for WG, respectively. These rates have surpassed all the current methods. This indicates that our proposed method achieves better classification performance in non-actual movement classification tasks under the premise of lightweight. The method proposed in this study can be applied to the field of rapid and efficient identification of brain signals.","PeriodicalId":13341,"journal":{"name":"IEEE Transactions on Instrumentation and Measurement","volume":"74 ","pages":"1-12"},"PeriodicalIF":5.6,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480749","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：