Pub Date : 2026-01-12DOI: 10.1109/JSEN.2025.3650455
Chun-Chi Lai;Bo-Jun Yang;Chia-Jen Lin
This study proposes a hybrid control framework that integrates a deep Q-network (DQN), adaptive proportional–integral–derivative (PID) control, and multisensor fusion via an extended Kalman filter (EKF) to enhance the accuracy, stability, and adaptability of autonomous mobile robots (AMRs) during docking tasks in complex indoor environments. A neural network dynamically tunes PID parameters based on the robot’s state, combining the robustness of classical control with the flexibility of learningbased methods. For localization, AprilTag visual markers are fused with multisensor data through EKF, yielding more accurate state estimation. A task-specific reward function incorporates target distance, angular deviation, collision penalties, and docking incentives, guiding the learning process toward smooth and efficient trajectories. Cosine-based angular velocity modulation and a LiDAR-triggered mode selector enable seamless switching between DQN–PID control and a modified DQN policy with smoother motion and faster reward convergence. While conventional DQN suffers from unsmooth motion and slower reward convergence, experimental results in both simulated and real-world environments show that the proposed switching framework achieves nearly 100% docking success, greatly surpassing the DQN-only approach, which gained only 59%. These results demonstrate clear advantages in convergence speed, trajectory smoothness, and robustness, confirming the framework’s suitability for real-world autonomous docking applications.
{"title":"A Hybrid DQN–PID Control Framework With Multisensor Fusion for Enhanced Docking Performance of Autonomous Mobile Robots in Complex Environments","authors":"Chun-Chi Lai;Bo-Jun Yang;Chia-Jen Lin","doi":"10.1109/JSEN.2025.3650455","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3650455","url":null,"abstract":"This study proposes a hybrid control framework that integrates a deep Q-network (DQN), adaptive proportional–integral–derivative (PID) control, and multisensor fusion via an extended Kalman filter (EKF) to enhance the accuracy, stability, and adaptability of autonomous mobile robots (AMRs) during docking tasks in complex indoor environments. A neural network dynamically tunes PID parameters based on the robot’s state, combining the robustness of classical control with the flexibility of learningbased methods. For localization, AprilTag visual markers are fused with multisensor data through EKF, yielding more accurate state estimation. A task-specific reward function incorporates target distance, angular deviation, collision penalties, and docking incentives, guiding the learning process toward smooth and efficient trajectories. Cosine-based angular velocity modulation and a LiDAR-triggered mode selector enable seamless switching between DQN–PID control and a modified DQN policy with smoother motion and faster reward convergence. While conventional DQN suffers from unsmooth motion and slower reward convergence, experimental results in both simulated and real-world environments show that the proposed switching framework achieves nearly 100% docking success, greatly surpassing the DQN-only approach, which gained only 59%. These results demonstrate clear advantages in convergence speed, trajectory smoothness, and robustness, confirming the framework’s suitability for real-world autonomous docking applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6438-6449"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162195","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}
Discharge plasma-based treatments for different applications, such as sterilization, surface decontamination, food processing, water, and air pollution control, have many advantages over conventional technologies. Nonthermal plasma (NTP) consists of energetic electrons, which in turn generates free radicals through interactions and triggers important chemical reactions. As a result, NTP is being preferred in a variety of domains but there are several technical challenges for its successful implementation. Researchers from different backgrounds: physics, chemistry, electrical, and chemical engineering, are working to understand various aspects of NTP treatments with different domain-specific objectives. It is required to form multidisciplinary research groups and study various procedures to increase the potency of these treatments. This article is intended to stand as a complete guide for budding researchers working on NTP applications, simultaneously attracting new researchers to this area, by providing total information regarding various high-voltage (HV) sources, discharge techniques, and reactor configurations available in the literature for different NTP treatments, along with their details: designs, fabrications, and performances.
{"title":"High-Voltage Sources and Reactors for Nonthermal Plasma Applications: A Review of Designs, Fabrications, and Their Performances","authors":"Sushma Balanagu;Srikanth Allamsetty;Ambrish Devanshu","doi":"10.1109/TPS.2025.3650163","DOIUrl":"https://doi.org/10.1109/TPS.2025.3650163","url":null,"abstract":"Discharge plasma-based treatments for different applications, such as sterilization, surface decontamination, food processing, water, and air pollution control, have many advantages over conventional technologies. Nonthermal plasma (NTP) consists of energetic electrons, which in turn generates free radicals through interactions and triggers important chemical reactions. As a result, NTP is being preferred in a variety of domains but there are several technical challenges for its successful implementation. Researchers from different backgrounds: physics, chemistry, electrical, and chemical engineering, are working to understand various aspects of NTP treatments with different domain-specific objectives. It is required to form multidisciplinary research groups and study various procedures to increase the potency of these treatments. This article is intended to stand as a complete guide for budding researchers working on NTP applications, simultaneously attracting new researchers to this area, by providing total information regarding various high-voltage (HV) sources, discharge techniques, and reactor configurations available in the literature for different NTP treatments, along with their details: designs, fabrications, and performances.","PeriodicalId":450,"journal":{"name":"IEEE Transactions on Plasma Science","volume":"54 2","pages":"801-818"},"PeriodicalIF":1.5,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146199208","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1109/JSEN.2026.3650823
Shailendra Chauhan;Rajeev Trehan;Ravi Pratap Singh;Vishal S. Sharma
This research presents an integrated and systematically validated framework for predicting tool wear in milling Inconel X750 using multisensor fusion. In this study, an accelerometer and a dynamometer are integrated to achieve sensor fusion, along with cryogenically treated cutting tool inserts with different edge radii. Experiments were designed to analyze tool wear, with results evaluated using analysis of variance (ANOVA) tests. The study employs Savitsky Golay (S-Golay) filtered Stationary Wavelet Transform and the largest Lyapunov exponent (LLE) to extract features from vibration and cutting force signals, enhancing prediction accuracy. Explainable artificial intelligence (XAI) ensures model transparency, while the extreme learning machine (ELM) effectively manages complex data relationships, yielding robust predictions. By combining sensor fusion with XAI, the study enhances interpretability and trust in AI-based decisions, making predictive maintenance more actionable for industrial applications. Results show the depth of cut has the highest mean Shapley values, achieving accurate metrics for tool inserts T1 and T2. Furthermore, the study achieves comparable accuracy metrics for cutting tool inserts T1 and T2, with a root mean square error (RMSE) of 2.27%, a mean absolute error (MAE) of 1.47%, and $left|R_{95 %}right|$ of 4.61% for cutting tool T1 and an RMSE of 3.14%, an MAE of 1.95%, and $left|R_{95 %}right|$ of 5.1% for cutting tool T2. This research enhances machining practices, particularly in aerospace, improving tool life and efficiency.
{"title":"Intelligent Tool Wear Prediction for Enhanced Sustainability in Milling of Ni-Based Superalloy","authors":"Shailendra Chauhan;Rajeev Trehan;Ravi Pratap Singh;Vishal S. Sharma","doi":"10.1109/JSEN.2026.3650823","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3650823","url":null,"abstract":"This research presents an integrated and systematically validated framework for predicting tool wear in milling Inconel X750 using multisensor fusion. In this study, an accelerometer and a dynamometer are integrated to achieve sensor fusion, along with cryogenically treated cutting tool inserts with different edge radii. Experiments were designed to analyze tool wear, with results evaluated using analysis of variance (ANOVA) tests. The study employs Savitsky Golay (S-Golay) filtered Stationary Wavelet Transform and the largest Lyapunov exponent (LLE) to extract features from vibration and cutting force signals, enhancing prediction accuracy. Explainable artificial intelligence (XAI) ensures model transparency, while the extreme learning machine (ELM) effectively manages complex data relationships, yielding robust predictions. By combining sensor fusion with XAI, the study enhances interpretability and trust in AI-based decisions, making predictive maintenance more actionable for industrial applications. Results show the depth of cut has the highest mean Shapley values, achieving accurate metrics for tool inserts T1 and T2. Furthermore, the study achieves comparable accuracy metrics for cutting tool inserts T1 and T2, with a root mean square error (RMSE) of 2.27%, a mean absolute error (MAE) of 1.47%, and <inline-formula> <tex-math>$left|R_{95 %}right|$ </tex-math></inline-formula> of 4.61% for cutting tool T1 and an RMSE of 3.14%, an MAE of 1.95%, and <inline-formula> <tex-math>$left|R_{95 %}right|$ </tex-math></inline-formula> of 5.1% for cutting tool T2. This research enhances machining practices, particularly in aerospace, improving tool life and efficiency.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6344-6352"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162183","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}
Pub Date : 2026-01-12DOI: 10.1109/JSEN.2025.3649820
Jiguang Yang;Jiuyuan Huo;Fang Cao;Cong Mu
The node deployment optimization of heterogeneous wireless sensor networks (HWSNs) in elongated structural spaces faces complex multiobjective tradeoffs. To address the issues of low coverage, poor network connectivity, and energy imbalance in existing deployment strategies for elongated spaces, this study proposes a collaborative optimization deployment and autonomous multicriteria decision-making (MCDM) method based on a new improved multiobjective whale optimization algorithm (IMOWOA). First, a 3-D elongated spatial model (ESM) and a heterogeneous node probability perception model are constructed to characterize the coverage properties of nodes within the elongated space. Second, an elite-oriented multimode adaptive perturbation (EMAP) and random singledimensional update (RSDU) strategy are proposed, enabling the whale optimization algorithm (WOA) to focus on elite regions and strengthen local exploration. Then, a method for calculating crowding distance is proposed, which integrates multiscale neighborhoods and nonlinear weights, producing a high-quality, evenly distributed set of nondominated solutions. After obtaining the nondominated solution set, the entropy-based technique for order preference by similarity to an ideal solution (TOPSIS) method is employed to select the final deployment scheme. Finally, the performance of IMOWOA is tested using the CEC2020 multimodal multiobjective test functions. In the simulation model of the ESM, the proposed IMOWOA effectively balances multiple complex deployment objectives. The deployment optimization coverage of HWSN is improved by 18.48%, 2.05%, 17.54%, 20.03%, and 1.88% compared to multiobjective whale optimization algorithm (MOWOA), non-dominated sorting genetic algorithm II (NSGA-II), multiple objective particle swarm optimization (MOPSO), competitive multi-objective marine predators algorithm (CMOMPA), and multiobjective transboundary search (MOTS), respectively. This demonstrates that the method can effectively handle the complex constraints of elongated spaces and provides a practical HWSN node deployment scheme for facility and structural monitoring in elongated environments. The source code is available on https://github.com/Drleach/IMOWOA
{"title":"Multiobjective Deployment Optimization and Final Solution Decision for Heterogeneous WSN Nodes in Elongated Structure Spaces","authors":"Jiguang Yang;Jiuyuan Huo;Fang Cao;Cong Mu","doi":"10.1109/JSEN.2025.3649820","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3649820","url":null,"abstract":"The node deployment optimization of heterogeneous wireless sensor networks (HWSNs) in elongated structural spaces faces complex multiobjective tradeoffs. To address the issues of low coverage, poor network connectivity, and energy imbalance in existing deployment strategies for elongated spaces, this study proposes a collaborative optimization deployment and autonomous multicriteria decision-making (MCDM) method based on a new improved multiobjective whale optimization algorithm (IMOWOA). First, a 3-D elongated spatial model (ESM) and a heterogeneous node probability perception model are constructed to characterize the coverage properties of nodes within the elongated space. Second, an elite-oriented multimode adaptive perturbation (EMAP) and random singledimensional update (RSDU) strategy are proposed, enabling the whale optimization algorithm (WOA) to focus on elite regions and strengthen local exploration. Then, a method for calculating crowding distance is proposed, which integrates multiscale neighborhoods and nonlinear weights, producing a high-quality, evenly distributed set of nondominated solutions. After obtaining the nondominated solution set, the entropy-based technique for order preference by similarity to an ideal solution (TOPSIS) method is employed to select the final deployment scheme. Finally, the performance of IMOWOA is tested using the CEC2020 multimodal multiobjective test functions. In the simulation model of the ESM, the proposed IMOWOA effectively balances multiple complex deployment objectives. The deployment optimization coverage of HWSN is improved by 18.48%, 2.05%, 17.54%, 20.03%, and 1.88% compared to multiobjective whale optimization algorithm (MOWOA), non-dominated sorting genetic algorithm II (NSGA-II), multiple objective particle swarm optimization (MOPSO), competitive multi-objective marine predators algorithm (CMOMPA), and multiobjective transboundary search (MOTS), respectively. This demonstrates that the method can effectively handle the complex constraints of elongated spaces and provides a practical HWSN node deployment scheme for facility and structural monitoring in elongated environments. The source code is available on <uri>https://github.com/Drleach/IMOWOA</uri>","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6418-6437"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162199","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}
Deep reinforcement learning (DRL)-based random access (RA) schemes break through the limitation of conventional RA protocols due to a lack of coordination among terminals, but they still face performance degradation in environmental instability, hindering their adaptability to wireless sensor networks (WSNs). To overcome this issue, a two-phase RA protocol is proposed in this article to realize coordination among terminals. In the scheme, the time frame is divided into a coordination phase and a transmission phase. During the coordination phase, nodes request resource units (RUs) in a distributed manner according to the optimal resource quotas calculated by the access point (AP). To minimize the time overhead caused by the coordination phase, we propose a lightweight learning algorithm that dynamically adjusts nodes’ request policies based on previous request outcomes. This mechanism enables the rapid convergence of the proposed scheme toward the optimal quota, and thus, the time overhead is substantially reduced. Featuring low computational complexity and inherent adaptability to environmental dynamics, the proposed algorithm is very suitable for WSNs. The simulation results validate that the time overhead of the proposed scheme is significantly lower than that of the existing state-of-the-art contention resolution (CR) algorithm. With the cost of higher energy consumption when the number of nodes is large, the proposed RA scheme achieves about 41.3% lower age of information (AoI) and 77.7% higher normalized throughput compared to the existing AoI-oriented nonorthogonal multiple access (NOMA)-RA scheme under common dynamic traffic models.
{"title":"A Novel Two-Phase NOMA-ALOHA Protocol Enhanced by User Coordination for Wireless Sensor Networks","authors":"Zhengyu Zhang;Guangliang Ren;Shuang Liang;Dandan Guan","doi":"10.1109/JSEN.2025.3644038","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3644038","url":null,"abstract":"Deep reinforcement learning (DRL)-based random access (RA) schemes break through the limitation of conventional RA protocols due to a lack of coordination among terminals, but they still face performance degradation in environmental instability, hindering their adaptability to wireless sensor networks (WSNs). To overcome this issue, a two-phase RA protocol is proposed in this article to realize coordination among terminals. In the scheme, the time frame is divided into a coordination phase and a transmission phase. During the coordination phase, nodes request resource units (RUs) in a distributed manner according to the optimal resource quotas calculated by the access point (AP). To minimize the time overhead caused by the coordination phase, we propose a lightweight learning algorithm that dynamically adjusts nodes’ request policies based on previous request outcomes. This mechanism enables the rapid convergence of the proposed scheme toward the optimal quota, and thus, the time overhead is substantially reduced. Featuring low computational complexity and inherent adaptability to environmental dynamics, the proposed algorithm is very suitable for WSNs. The simulation results validate that the time overhead of the proposed scheme is significantly lower than that of the existing state-of-the-art contention resolution (CR) algorithm. With the cost of higher energy consumption when the number of nodes is large, the proposed RA scheme achieves about 41.3% lower age of information (AoI) and 77.7% higher normalized throughput compared to the existing AoI-oriented nonorthogonal multiple access (NOMA)-RA scheme under common dynamic traffic models.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6372-6387"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162186","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}
Pub Date : 2026-01-12DOI: 10.1109/JSEN.2026.3651572
A. S. Gautam;P. P. Sahu
Metallic nanoparticles have garnered significant attention due to their unique physicochemical properties and its applicability especially in the detection of proteins present in biological fluids causing critical diseases. Numerous synthesis techniques have been explored to tailor these nanoparticles for selective chemical interaction with particular proteins. In this work, we present an ecofriendly synthesis of silver nanoparticles (AgNPs) by the reduction of silver salts, with employing carambola (Averrhoa carambola) fruit extract as a natural capping and reducing agent for the colorimetric detection of homocysteine. We have used colorimetric red, green, blue (RGB) analysis for the determination of homocysteine (Hcys) concentration ranging from 5 to 100 μM with very small sample volume of 2 mL. The proposed method also demonstrates selective detection of Hcys over wide range protein present in blood serum opening an avenue for early diagnosis of Parkinson's and Alzheimer's diseases.
{"title":"Detection of Homocysteine With Colorimetric Approach Using Carambola Fruit Extract Capped Silver Nanoparticles","authors":"A. S. Gautam;P. P. Sahu","doi":"10.1109/JSEN.2026.3651572","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3651572","url":null,"abstract":"Metallic nanoparticles have garnered significant attention due to their unique physicochemical properties and its applicability especially in the detection of proteins present in biological fluids causing critical diseases. Numerous synthesis techniques have been explored to tailor these nanoparticles for selective chemical interaction with particular proteins. In this work, we present an ecofriendly synthesis of silver nanoparticles (AgNPs) by the reduction of silver salts, with employing carambola (Averrhoa carambola) fruit extract as a natural capping and reducing agent for the colorimetric detection of homocysteine. We have used colorimetric red, green, blue (RGB) analysis for the determination of homocysteine (Hcys) concentration ranging from 5 to 100 μM with very small sample volume of 2 mL. The proposed method also demonstrates selective detection of Hcys over wide range protein present in blood serum opening an avenue for early diagnosis of Parkinson's and Alzheimer's diseases.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6498-6505"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162191","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}
Pub Date : 2026-01-12DOI: 10.1109/JSEN.2026.3651557
Yuanhong Meng;Zhenyu Liang;Zhiwei Guo;Xiaoqiang Su;Yanhong Liu;Fusheng Deng;Caixia Feng;Lijuan Dong;Weidong Hu
In the process of comprehensive mechanized caving coal mining, the monitoring technology for coal gangue mixing ratios at coal discharge outlets has long relied on manual experience judgment, lacking automated monitoring methods. This has led to widespread overdischarge or underdischarge phenomena during coal release, severely affecting mining efficiency and quality. To address this technical challenge, this article innovatively designs and develops an integrated sensing monitoring system based on the parity-time (PT) symmetry principle. By constructing a three-coil LC resonant circuit system, we utilize the high-sensitivity characteristics of third-order PT symmetry at exceptional points (EPs) to achieve stable monitoring of resonant frequency variations. Experimental results show that the sensitivity enhancement factor of the synthetic third-order PT system reaches up to 1.8 times that of second-order systems, effectively detecting frequency differences caused by changes in coal gangue mixture dielectric constants. Based on this, we establish a quantitative relationship model between coal gangue ratios and resonant frequencies, enabling precise determination of mixing ratios. Additionally, combining synthetic dimension theory, we design a sensing system with PT symmetric circuit architecture, achieving high-sensitivity monitoring of minute gangue ratio variations. This sensing monitoring system not only significantly reduces equipment size but also demonstrates excellent detection accuracy and stability. It provides reliable technical support for improving coal quality and achieving automated mining in coal mine working faces, playing a significant role in advancing intelligent development in the coal industry.
{"title":"Novel Coal Gangue Mixing Ratio Sensing Technique Based on Synthetic Parity-Time Symmetry","authors":"Yuanhong Meng;Zhenyu Liang;Zhiwei Guo;Xiaoqiang Su;Yanhong Liu;Fusheng Deng;Caixia Feng;Lijuan Dong;Weidong Hu","doi":"10.1109/JSEN.2026.3651557","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3651557","url":null,"abstract":"In the process of comprehensive mechanized caving coal mining, the monitoring technology for coal gangue mixing ratios at coal discharge outlets has long relied on manual experience judgment, lacking automated monitoring methods. This has led to widespread overdischarge or underdischarge phenomena during coal release, severely affecting mining efficiency and quality. To address this technical challenge, this article innovatively designs and develops an integrated sensing monitoring system based on the parity-time (PT) symmetry principle. By constructing a three-coil LC resonant circuit system, we utilize the high-sensitivity characteristics of third-order PT symmetry at exceptional points (EPs) to achieve stable monitoring of resonant frequency variations. Experimental results show that the sensitivity enhancement factor of the synthetic third-order PT system reaches up to 1.8 times that of second-order systems, effectively detecting frequency differences caused by changes in coal gangue mixture dielectric constants. Based on this, we establish a quantitative relationship model between coal gangue ratios and resonant frequencies, enabling precise determination of mixing ratios. Additionally, combining synthetic dimension theory, we design a sensing system with PT symmetric circuit architecture, achieving high-sensitivity monitoring of minute gangue ratio variations. This sensing monitoring system not only significantly reduces equipment size but also demonstrates excellent detection accuracy and stability. It provides reliable technical support for improving coal quality and achieving automated mining in coal mine working faces, playing a significant role in advancing intelligent development in the coal industry.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 5","pages":"7670-7679"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299608","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}
Pub Date : 2026-01-12DOI: 10.1109/JSEN.2026.3651611
Gewei Lou;Wenkai Lu;Yonghuang Zheng;Tingzheng Shen;Jun Chen;Xiangbo Suo;Xuliang Liu
Regression-based LiDAR localization has achieved impressive accuracy but remains challenging due to viewpoint variations and computational inefficiency. This article introduces SE(2)-MambaLoc, an end-to-end regression-based framework combining SE(2)-equivariant feature learning and Mamba diffusion models for robust and efficient bird’s-eye-view (BEV) LiDAR localization. Our approach first constructs a BEV height-weighted density map (BEV-HWDM) to preserve elevation-aware structural features while reducing storage demands. An SE(2)-equivariant feature extractor (SEFE), built on a modified ResNet18, interleaves lightweight decoupled SE(2) group convolution blocks, which are decomposed into kernel generators and positional encoders, with standard residual stages. It replaces the classification head with parallel inception bottleneck and context anchor attention (CAA) modules to produce a sequence of rotation-robust spatial BEV descriptors without data augmentation. For pose regression, we reformulate localization as an iterative denoising process using a Mamba diffusion model. By treating the spatial feature map as a sequence of tokens, our Mamba backbone captures the geometric dependencies between the pose and the environmental context with linear-time complexity, avoiding the quadratic bottleneck of Transformers. Extensive experiments on Oxford Radar RobotCar and NCLT datasets demonstrate that our proposed SE(2)-MambaLoc achieves superior orientation (yaw) accuracy over state-of-the-art (SOTA) methods with comparable position accuracy and a 49% reduction in training time. Ablation studies validate the effectiveness of BEV-HWDM, SEFE, and Mamba diffusion components, underscoring their roles in enhancing robustness and efficiency.
{"title":"SE(2)-MambaLoc: Regression-Based LiDAR Localization via SE(2)-Equivariant Feature Learning and Mamba Diffusion","authors":"Gewei Lou;Wenkai Lu;Yonghuang Zheng;Tingzheng Shen;Jun Chen;Xiangbo Suo;Xuliang Liu","doi":"10.1109/JSEN.2026.3651611","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3651611","url":null,"abstract":"Regression-based LiDAR localization has achieved impressive accuracy but remains challenging due to viewpoint variations and computational inefficiency. This article introduces SE(2)-MambaLoc, an end-to-end regression-based framework combining SE(2)-equivariant feature learning and Mamba diffusion models for robust and efficient bird’s-eye-view (BEV) LiDAR localization. Our approach first constructs a BEV height-weighted density map (BEV-HWDM) to preserve elevation-aware structural features while reducing storage demands. An SE(2)-equivariant feature extractor (SEFE), built on a modified ResNet18, interleaves lightweight decoupled SE(2) group convolution blocks, which are decomposed into kernel generators and positional encoders, with standard residual stages. It replaces the classification head with parallel inception bottleneck and context anchor attention (CAA) modules to produce a sequence of rotation-robust spatial BEV descriptors without data augmentation. For pose regression, we reformulate localization as an iterative denoising process using a Mamba diffusion model. By treating the spatial feature map as a sequence of tokens, our Mamba backbone captures the geometric dependencies between the pose and the environmental context with linear-time complexity, avoiding the quadratic bottleneck of Transformers. Extensive experiments on Oxford Radar RobotCar and NCLT datasets demonstrate that our proposed SE(2)-MambaLoc achieves superior orientation (yaw) accuracy over state-of-the-art (SOTA) methods with comparable position accuracy and a 49% reduction in training time. Ablation studies validate the effectiveness of BEV-HWDM, SEFE, and Mamba diffusion components, underscoring their roles in enhancing robustness and efficiency.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6464-6477"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162202","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}
Pub Date : 2026-01-12DOI: 10.1109/TPS.2025.3647270
Wenying Yang;Fansong Meng;Daoyi Wu
The bypass switch is an important component for protecting the modular multilevel converter submodule in the flexible DC transmission system. Due to the low impedance of the flexible DC system, the fault current rises rapidly, imposing strict requirements on the closing speed of the bypass switch. As a linear actuator, the electromagnetic repulsion mechanism (ERM) can be a solution for the mechanical part of the bypass switch. At present, the ERM is primarily designed and optimized by the finite element method (FEM) and the equivalent circuit method (ECM). However, the FEM suffers from limitations such as low computational efficiency, large resource consumption, and model modification, while the inductance and inductance gradient required by the ECM are difficult to obtain. Moreover, the ECM divides the armature into multiple regions, resulting in highly ill-conditioned inductance matrix and inductance gradient matrix, which increases the risk of numerical errors. In this regard, a series armature equivalent method (SAEM) is proposed in this article to calculate the dynamic characteristics of ERM. First, the principle of SAEM is introduced, and the formulas of the electrical parameters are given. The calculation accuracy of the electrical parameters is verified by the FEM. Then, the dynamic characteristics of the ERM calculated by the ECM, the SAEM, and the FEM, are compared with the experimental results. The results indicate that SAEM has similar accuracy to the ECM, and it has more advantages in numerical stability. In addition, the SAEM has more advantages in calculation efficiency than the ECM and the FEM.
{"title":"An Improved Calculation Method for Dynamic Characteristics of Electromagnetic Repulsion Mechanism Based on Series Armature Equivalence","authors":"Wenying Yang;Fansong Meng;Daoyi Wu","doi":"10.1109/TPS.2025.3647270","DOIUrl":"https://doi.org/10.1109/TPS.2025.3647270","url":null,"abstract":"The bypass switch is an important component for protecting the modular multilevel converter submodule in the flexible DC transmission system. Due to the low impedance of the flexible DC system, the fault current rises rapidly, imposing strict requirements on the closing speed of the bypass switch. As a linear actuator, the electromagnetic repulsion mechanism (ERM) can be a solution for the mechanical part of the bypass switch. At present, the ERM is primarily designed and optimized by the finite element method (FEM) and the equivalent circuit method (ECM). However, the FEM suffers from limitations such as low computational efficiency, large resource consumption, and model modification, while the inductance and inductance gradient required by the ECM are difficult to obtain. Moreover, the ECM divides the armature into multiple regions, resulting in highly ill-conditioned inductance matrix and inductance gradient matrix, which increases the risk of numerical errors. In this regard, a series armature equivalent method (SAEM) is proposed in this article to calculate the dynamic characteristics of ERM. First, the principle of SAEM is introduced, and the formulas of the electrical parameters are given. The calculation accuracy of the electrical parameters is verified by the FEM. Then, the dynamic characteristics of the ERM calculated by the ECM, the SAEM, and the FEM, are compared with the experimental results. The results indicate that SAEM has similar accuracy to the ECM, and it has more advantages in numerical stability. In addition, the SAEM has more advantages in calculation efficiency than the ECM and the FEM.","PeriodicalId":450,"journal":{"name":"IEEE Transactions on Plasma Science","volume":"54 2","pages":"774-783"},"PeriodicalIF":1.5,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146199234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1109/JSEN.2026.3650983
Xianlong Ma;Tao Liu;Dmitry Kiesewetter;Hong Li;Victor Malyugin;Changsen Sun
A low-coherent optical sensor combined with a hydrostatic leveling configuration has been developed to monitor ground settlement (GS) in engineering. However, if the monitoring sites extend over a large span, the inhomogeneous spatial temperature field can deteriorate the performance of the sensor determined by the temperaturedependent liquid density distribution in the hydrostatic system. Therefore, the fluctuations of the environmental temperature cause a hydrostatic leveling error in a timevarying way. Based on the measured results, a compensation method using an improved back-propagation (BP) neural network is proposed to suppress the effects of temperature. Besides the network structure itself, a genetic algorithm (GA) incorporated with a specially designed fitness function and blend crossover with $alpha$ (BLX-$alpha$ ) operator is employed to optimize the weights and biases of the neural network. These treatments have improved the global searching capability, training speed, and convergence efficiency. Based on training with 9000 samples, an improved GA combined with a BP neural network reduces temperature-induced error by approximately 50%. The system achieves 0.3 mm accuracy across a 20 cm measurement range. It is proved by a practical testing experiment configured in a binary temperature field lasting for one month. This method could be a good reference for further practical applications.
{"title":"Temperature Compensation for the Low-Coherent Optical Settlement Sensor by Using an IGA-BP Neural Network","authors":"Xianlong Ma;Tao Liu;Dmitry Kiesewetter;Hong Li;Victor Malyugin;Changsen Sun","doi":"10.1109/JSEN.2026.3650983","DOIUrl":"https://doi.org/10.1109/JSEN.2026.3650983","url":null,"abstract":"A low-coherent optical sensor combined with a hydrostatic leveling configuration has been developed to monitor ground settlement (GS) in engineering. However, if the monitoring sites extend over a large span, the inhomogeneous spatial temperature field can deteriorate the performance of the sensor determined by the temperaturedependent liquid density distribution in the hydrostatic system. Therefore, the fluctuations of the environmental temperature cause a hydrostatic leveling error in a timevarying way. Based on the measured results, a compensation method using an improved back-propagation (BP) neural network is proposed to suppress the effects of temperature. Besides the network structure itself, a genetic algorithm (GA) incorporated with a specially designed fitness function and blend crossover with <inline-formula> <tex-math>$alpha$ </tex-math></inline-formula> (BLX-<inline-formula> <tex-math>$alpha$ </tex-math></inline-formula>) operator is employed to optimize the weights and biases of the neural network. These treatments have improved the global searching capability, training speed, and convergence efficiency. Based on training with 9000 samples, an improved GA combined with a BP neural network reduces temperature-induced error by approximately 50%. The system achieves 0.3 mm accuracy across a 20 cm measurement range. It is proved by a practical testing experiment configured in a binary temperature field lasting for one month. This method could be a good reference for further practical applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 4","pages":"6353-6359"},"PeriodicalIF":4.3,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146162198","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}
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：