期刊：
Materials Science in Semiconductor Processing,2026年201:110038 ISSN：1369-8001
通讯作者：
Dachang Chen
作者机构：
[Jingzhi Zhao; Yongqing Qian; Dachang Chen] School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan, 430023, China;[Zihan Xu] School of Automation, Northwestern Polytechnical University, Xi'an, 710129, China;[Xiaoxing Zhang] Key Laboratory for High-Efficiency Utilization of Solar Energy and Operation Control of Energy Storage System, Hubei University of Technology, Wuhan, 430068, China;[Beibei Xiao] School of Energy and Power Engineering, Jiangsu University of Science and Technology, Zhenjiang, 212003, China
通讯机构：
[Dachang Chen] S;School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan, 430023, China
摘要：
The adsorption and sensing properties of pristine α-In 2 Se 3 and Pt doped In 2 Se 3 monolayers for five thermal runaway gases (H 2 , CO, CO 2 , CH 4 , C 2 H 4 ) have been investigated using density functional theory (DFT). The adsorption energy, charge transfer ( Q T ), energy density of states, work function, recovery time and sensing response were compared to elucidate the gas adsorption behavior and electronic properties. The results indicate that, when Pt replaces doping the In 2 Se 3 (↑) surface, the phase state of α-In 2 Se 3 changes, turning into a more stable β phase. However, in the Pt-In 2 Se 3 (↓) surface, no phase transition was observed in α-In 2 Se 3 . And for the adsorption of H 2 , CO and C 2 H 4 , the introduction of Pt atoms significantly enhances the adsorption energies. Additionally, the absolute value of the integral crystal orbital Hamiltonian population (ICOHP) is the highest for CO adsorbed on Pt doped In 2 Se 3 monolayer systems, with values of −2.73 and −2.77 for the upward and downward polarization directions, respectively. This indicates a stronger interaction between CO and Pt atoms, suggesting an enhanced potential for chemical bond formation at Pt sites. The adsorption of gas molecules has shown pronounced differences in its impact on the work function. The Pt-In 2 Se 3 (↑) surface exhibits considerable sensitivity to CO and CH 4 , resulting in the most significant work function shift (work function shift of CO is 0.14 eV; work function shift of C 2 H 4 is 0.3 eV), while its response to H 2 , CO 2 and CH 4 is nearly negligible. Meanwhile, the adsorption of C 2 H 4 on the Pt-In 2 Se 3 (↓) yields a significant work function shift of 0.49 eV, further indicating the material's potential for selective detection of specific gases. A comprehensive analysis of adsorption energy and work function further reveals that the Pt-In 2 Se 3 (↑) surface exhibits the most pronounced gas-sensing properties toward CO (adsorption energy of 1.26 eV and work function shift to 6.01 eV), while the Pt-In 2 Se 3 (↓) surface shows significant sensitivity to both CO (adsorption energy of 1.34 eV and work function shift to 5.31 eV) and C 2 H 4 (adsorption energy of 1.05 eV and work function shift to 5.25 eV). When both sensitivity and recovery characteristics are taken into account, Pt-In 2 Se 3 monolayers demonstrate a remarkable negative sensing response towards C 2 H 4 , with values of −99.61 % for the In 2 Se 3 (↑) surface and −96.87 % for the In 2 Se 3 (↓) surface. Furthermore, the recovery time for C 2 H 4 on the Pt-In 2 Se 3 (↑) surface is as short as 4.19 × 10 −3 s at 498 K, while on the Pt-In 2 Se 3 (↓) surface it is 4.56 × 10 −2 s. These findings indicate that Pt functionalization not only enables ultrahigh sensitivity but also allows for rapid recovery of the sensor after gas exposure, suggesting that Pt-doped In 2 Se 3 monolayers are highly promising candidates for efficient and real-time detection of ethylene. These findings provide a theoretical basis for understanding the microscopic mechanisms of gas detection in In 2 Se 3 -based sensors and contribute to the design of advanced sensing materials for detecting thermal runaway gases in LIBs.
The adsorption and sensing properties of pristine α-In 2 Se 3 and Pt doped In 2 Se 3 monolayers for five thermal runaway gases (H 2 , CO, CO 2 , CH 4 , C 2 H 4 ) have been investigated using density functional theory (DFT). The adsorption energy, charge transfer ( Q T ), energy density of states, work function, recovery time and sensing response were compared to elucidate the gas adsorption behavior and electronic properties. The results indicate that, when Pt replaces doping the In 2 Se 3 (↑) surface, the phase state of α-In 2 Se 3 changes, turning into a more stable β phase. However, in the Pt-In 2 Se 3 (↓) surface, no phase transition was observed in α-In 2 Se 3 . And for the adsorption of H 2 , CO and C 2 H 4 , the introduction of Pt atoms significantly enhances the adsorption energies. Additionally, the absolute value of the integral crystal orbital Hamiltonian population (ICOHP) is the highest for CO adsorbed on Pt doped In 2 Se 3 monolayer systems, with values of −2.73 and −2.77 for the upward and downward polarization directions, respectively. This indicates a stronger interaction between CO and Pt atoms, suggesting an enhanced potential for chemical bond formation at Pt sites. The adsorption of gas molecules has shown pronounced differences in its impact on the work function. The Pt-In 2 Se 3 (↑) surface exhibits considerable sensitivity to CO and CH 4 , resulting in the most significant work function shift (work function shift of CO is 0.14 eV; work function shift of C 2 H 4 is 0.3 eV), while its response to H 2 , CO 2 and CH 4 is nearly negligible. Meanwhile, the adsorption of C 2 H 4 on the Pt-In 2 Se 3 (↓) yields a significant work function shift of 0.49 eV, further indicating the material's potential for selective detection of specific gases. A comprehensive analysis of adsorption energy and work function further reveals that the Pt-In 2 Se 3 (↑) surface exhibits the most pronounced gas-sensing properties toward CO (adsorption energy of 1.26 eV and work function shift to 6.01 eV), while the Pt-In 2 Se 3 (↓) surface shows significant sensitivity to both CO (adsorption energy of 1.34 eV and work function shift to 5.31 eV) and C 2 H 4 (adsorption energy of 1.05 eV and work function shift to 5.25 eV). When both sensitivity and recovery characteristics are taken into account, Pt-In 2 Se 3 monolayers demonstrate a remarkable negative sensing response towards C 2 H 4 , with values of −99.61 % for the In 2 Se 3 (↑) surface and −96.87 % for the In 2 Se 3 (↓) surface. Furthermore, the recovery time for C 2 H 4 on the Pt-In 2 Se 3 (↑) surface is as short as 4.19 × 10 −3 s at 498 K, while on the Pt-In 2 Se 3 (↓) surface it is 4.56 × 10 −2 s. These findings indicate that Pt functionalization not only enables ultrahigh sensitivity but also allows for rapid recovery of the sensor after gas exposure, suggesting that Pt-doped In 2 Se 3 monolayers are highly promising candidates for efficient and real-time detection of ethylene. These findings provide a theoretical basis for understanding the microscopic mechanisms of gas detection in In 2 Se 3 -based sensors and contribute to the design of advanced sensing materials for detecting thermal runaway gases in LIBs.
摘要：
In the industrial routine inspection process, accurately tracking pipeline defects is essential for ensuring safe operation. However, this accuracy is adversely affected by the complexity of pipeline internal environment, similarity in the appearance of different defects, and difference in the appearance of the same defect with the movement of the camera used. Herein, we present a novel model called global prototype tracking network (GP-Net), which integrates prototype learning, global attention, and global tracking to accurately track the defects. The proposed model mitigates the feature bias caused by prototype singularity via iterative clustering. A dilated attention method is applied at multiple scales to resolve the redundant computations in global attention. Our experiments demonstrated that GP-Net achieved multi-object tracking accuracy, identity F1 score, and multi-object tracking precision of 79.6%, 85.5%, and 93.3%, respectively. Therefore, we demonstrated the effectiveness of GP-Net in tracking pipeline defects, thereby exploiting the potential application of deep learning in industrial maintenance. The source code and partial dataset are publicly available at https://github.com/cui19981127/GP-Net .
In the industrial routine inspection process, accurately tracking pipeline defects is essential for ensuring safe operation. However, this accuracy is adversely affected by the complexity of pipeline internal environment, similarity in the appearance of different defects, and difference in the appearance of the same defect with the movement of the camera used. Herein, we present a novel model called global prototype tracking network (GP-Net), which integrates prototype learning, global attention, and global tracking to accurately track the defects. The proposed model mitigates the feature bias caused by prototype singularity via iterative clustering. A dilated attention method is applied at multiple scales to resolve the redundant computations in global attention. Our experiments demonstrated that GP-Net achieved multi-object tracking accuracy, identity F1 score, and multi-object tracking precision of 79.6%, 85.5%, and 93.3%, respectively. Therefore, we demonstrated the effectiveness of GP-Net in tracking pipeline defects, thereby exploiting the potential application of deep learning in industrial maintenance. The source code and partial dataset are publicly available at https://github.com/cui19981127/GP-Net .
摘要：
To enhance the efficiency of banana quality management, an automated freshness grading system based on real-time image processing was developed. The system integrates a camera for image acquisition and employs portable edge devices (Nvidia Jetson Orin NX or Jetson TX2) to classify banana freshness into four levels, displaying results along with confidence scores in real-time. A lightweight YOLO-based banana freshness detection method (YOLO-BFD), built on the YOLOv7-tiny framework, was designed to enable high-accuracy, real-time processing on edge devices. Key optimizations—including reduced model width factors, the LightSP module, VoVSSCSP, and Ghost convolutions—resulted in a 118.4 % increase in detection speed and a 1.4 % improvement in mean average precision (mAP) over the baseline, along with a 5 % increase in precision, 1.5 % increase in recall, and a 3 % increase in F1 score. Additionally, the system achieved a 93.5 % reduction in computational costs and a 41.9 % decrease in model parameters. The system demonstrated an average accuracy of 97.8 % in freshness grading, with real-time detection speeds of 61 frames per second on the Jetson Orin NX and 29 frames per second on the Jetson TX2. Overall, the proposed system provides a practical and efficient solution for portable banana quality assessment in agricultural applications.
To enhance the efficiency of banana quality management, an automated freshness grading system based on real-time image processing was developed. The system integrates a camera for image acquisition and employs portable edge devices (Nvidia Jetson Orin NX or Jetson TX2) to classify banana freshness into four levels, displaying results along with confidence scores in real-time. A lightweight YOLO-based banana freshness detection method (YOLO-BFD), built on the YOLOv7-tiny framework, was designed to enable high-accuracy, real-time processing on edge devices. Key optimizations—including reduced model width factors, the LightSP module, VoVSSCSP, and Ghost convolutions—resulted in a 118.4 % increase in detection speed and a 1.4 % improvement in mean average precision (mAP) over the baseline, along with a 5 % increase in precision, 1.5 % increase in recall, and a 3 % increase in F1 score. Additionally, the system achieved a 93.5 % reduction in computational costs and a 41.9 % decrease in model parameters. The system demonstrated an average accuracy of 97.8 % in freshness grading, with real-time detection speeds of 61 frames per second on the Jetson Orin NX and 29 frames per second on the Jetson TX2. Overall, the proposed system provides a practical and efficient solution for portable banana quality assessment in agricultural applications.
作者机构：
[Xiyuan Chen] Key Laboratory of Micro-Inertial Instrument and Advanced Navigation Technology, Ministry of Education, Southeast University, Nanjing 210096, China;[Junwei Wang] School of Artificial Intelligence and Automation, Hohai University, Nanjing 210098, China;[Jianguo Liu; Yi Mou; Teng Li; Lili Wan] School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan 430048, China
通讯机构：
[Jianguo Liu] S;School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan 430048, China
摘要：
This article investigates a fast azimuth gyro drift calibration method based on low-cost single-axis rotation strapdown inertial navigation system (SINS) under mooring conditions, utilizing an adaptive robust Kalman filter (ARKF) framework. Firstly, a rotation modulation KF model based on the angular rate extended measurement equation is designed, which effectively enhances the observability of heading misalignment angle and azimuth gyro drift. The validity of the proposed method is verified through singular value decomposition. Secondly, an online estimation method for measurement noise variance based on variational Bayesian (VB) is employed to address the issue of inaccurate acquisition of the extended measurement equation under dynamic conditions. To tackle the degradation of filter performance caused by VB’s iterative computations, a novel robust method employing multi-correction factors is formulated. Subsequently, a multivariate nonlinear optimization model utilizing an improved particle swarm optimization algorithm is established, enabling global optimization of correction factors while preventing local optima entrapment. This framework achieves multi-channel measurement error decoupling and selective error correction, thereby substantially improving filter stability. Experimental validation via Stewart platform testing demonstrates the method’s efficacy, with azimuth gyro drift estimation accuracy reaching 0.45°/h over one-hour trials.
This article investigates a fast azimuth gyro drift calibration method based on low-cost single-axis rotation strapdown inertial navigation system (SINS) under mooring conditions, utilizing an adaptive robust Kalman filter (ARKF) framework. Firstly, a rotation modulation KF model based on the angular rate extended measurement equation is designed, which effectively enhances the observability of heading misalignment angle and azimuth gyro drift. The validity of the proposed method is verified through singular value decomposition. Secondly, an online estimation method for measurement noise variance based on variational Bayesian (VB) is employed to address the issue of inaccurate acquisition of the extended measurement equation under dynamic conditions. To tackle the degradation of filter performance caused by VB’s iterative computations, a novel robust method employing multi-correction factors is formulated. Subsequently, a multivariate nonlinear optimization model utilizing an improved particle swarm optimization algorithm is established, enabling global optimization of correction factors while preventing local optima entrapment. This framework achieves multi-channel measurement error decoupling and selective error correction, thereby substantially improving filter stability. Experimental validation via Stewart platform testing demonstrates the method’s efficacy, with azimuth gyro drift estimation accuracy reaching 0.45°/h over one-hour trials.
摘要：
The development of materials to replace precious metal catalysts is key to achieving energy conversion at low cost. In this paper, a series of bifunctional catalysts based on 3d transition metal (TM) atoms and transition metal sulfide (TMD) heterojunctions with single-atom catalysts (SACs) based on density-functional theory (DFT) are reported. Oxygen electrocatalyst activity and stability were evaluated for 20 structures. The results showed that Ni-based SAC had significant OER/ORR catalytic activity with the lowest overpotential(0.41/0.78V). The mechanism of catalytic activity enhancement was investigated by descriptors of intermediate adsorption energy . A dual-atom catalyst (DAC) NiCo@Site Ⅰ was designed using the synergistic effect of active sites to reduce the reaction overpotential further. The overpotential of the active sites of the designed DACs was reduced compared to that of the SACs. This work contributes to solving key challenges in energy conversion and provides theoretical guidance for the design of bifunctional catalysts.
The development of materials to replace precious metal catalysts is key to achieving energy conversion at low cost. In this paper, a series of bifunctional catalysts based on 3d transition metal (TM) atoms and transition metal sulfide (TMD) heterojunctions with single-atom catalysts (SACs) based on density-functional theory (DFT) are reported. Oxygen electrocatalyst activity and stability were evaluated for 20 structures. The results showed that Ni-based SAC had significant OER/ORR catalytic activity with the lowest overpotential(0.41/0.78V). The mechanism of catalytic activity enhancement was investigated by descriptors of intermediate adsorption energy . A dual-atom catalyst (DAC) NiCo@Site Ⅰ was designed using the synergistic effect of active sites to reduce the reaction overpotential further. The overpotential of the active sites of the designed DACs was reduced compared to that of the SACs. This work contributes to solving key challenges in energy conversion and provides theoretical guidance for the design of bifunctional catalysts.
摘要：
We investigated the doping effects of Dy 3+ ions at Er 3+ sites on the anisotropy and magnetic phase transition in ErFeO 3 single crystal under static magnetic fields and pulsed high magnetic fields of up to 58 T. Experimental results show that the Dy 3+ -dopant has weak influence on the interaction between rare earth ions but weakens the interaction energy between Fe-3 d and Er-4 f electrons, shifting the spin reorientation and spin switching transitions to lower temperature region. Interestingly, in the doped system, it can be inferred that a Γ 3 ( C x , F y , A z ) phase emerges additionally to the Γ 4 ( G x , A y , F z ) to Γ 2 ( F x , C y , G z ) spin reorientation transition , and these transitions share the same critical temperature (∼88.9 K). Upon further cooling, a spin switching transition, which is similar with the one that shows in the Γ 2 phase, appears within the Γ 3 phase. That is, the rare-earth Dy 3+ ions possess antiparallel magnetic moments against the weak ferromagnetic vector F y of Γ 3 ( C x , F y , A z ) phase, and the direction of the net magnetic moments reverses at ∼5.9 K and ∼ 31 K in the FC and ZFC processes, respectively. These results indicate that, in the rare-earth orthoferrite R FeO 3 , doping is an effective way to modulate the interactions of Fe 3+ -Fe 3+ , R 3+ - R 3+ and R 3+ -Fe 3+ and to induce magnetic phases.
We investigated the doping effects of Dy 3+ ions at Er 3+ sites on the anisotropy and magnetic phase transition in ErFeO 3 single crystal under static magnetic fields and pulsed high magnetic fields of up to 58 T. Experimental results show that the Dy 3+ -dopant has weak influence on the interaction between rare earth ions but weakens the interaction energy between Fe-3 d and Er-4 f electrons, shifting the spin reorientation and spin switching transitions to lower temperature region. Interestingly, in the doped system, it can be inferred that a Γ 3 ( C x , F y , A z ) phase emerges additionally to the Γ 4 ( G x , A y , F z ) to Γ 2 ( F x , C y , G z ) spin reorientation transition , and these transitions share the same critical temperature (∼88.9 K). Upon further cooling, a spin switching transition, which is similar with the one that shows in the Γ 2 phase, appears within the Γ 3 phase. That is, the rare-earth Dy 3+ ions possess antiparallel magnetic moments against the weak ferromagnetic vector F y of Γ 3 ( C x , F y , A z ) phase, and the direction of the net magnetic moments reverses at ∼5.9 K and ∼ 31 K in the FC and ZFC processes, respectively. These results indicate that, in the rare-earth orthoferrite R FeO 3 , doping is an effective way to modulate the interactions of Fe 3+ -Fe 3+ , R 3+ - R 3+ and R 3+ -Fe 3+ and to induce magnetic phases.
摘要：
Elemental diffusion and reactions at interfaces significantly increase resistivity and reduce stability, particularly in thermoelectric (TE) systems containing highly diffusive elements like Te. Using thermodynamically stable phases as contact layers have eliminated interfacial reactions but have not been able to completely stop elemental diffusion. This work introduces an innovative approach to effectively increase the activation energy of cross-interface atom diffusion by creating a continuously interfacial symmetric strain field via a high density of interfacial edge dislocations. Specifically, a dense interfacial strain barrier layer is constructed using Ni 0.5 Te as contact layer, resulting in an atomically continuous Te 0.985 Sb 0.015 /Ni 0.5 Te interface. This design achieves a notable reduction in contact resistivity to 9 μΩ cm 2 while maintaining more than 75 % of the theoretical device efficiency at a hot-end temperature, ( T h ) of 523 K even after 21,600 min of aging. This method of optimizing both the interfacial microstructure and chemical composition provides a new avenue for constructing stably high-performance heterostructure devices.
Elemental diffusion and reactions at interfaces significantly increase resistivity and reduce stability, particularly in thermoelectric (TE) systems containing highly diffusive elements like Te. Using thermodynamically stable phases as contact layers have eliminated interfacial reactions but have not been able to completely stop elemental diffusion. This work introduces an innovative approach to effectively increase the activation energy of cross-interface atom diffusion by creating a continuously interfacial symmetric strain field via a high density of interfacial edge dislocations. Specifically, a dense interfacial strain barrier layer is constructed using Ni 0.5 Te as contact layer, resulting in an atomically continuous Te 0.985 Sb 0.015 /Ni 0.5 Te interface. This design achieves a notable reduction in contact resistivity to 9 μΩ cm 2 while maintaining more than 75 % of the theoretical device efficiency at a hot-end temperature, ( T h ) of 523 K even after 21,600 min of aging. This method of optimizing both the interfacial microstructure and chemical composition provides a new avenue for constructing stably high-performance heterostructure devices.
作者机构：
[Qiyao Luo; Yongqing Qian] School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan, China
会议名称：
2025 6th International Conference on Artificial Intelligence and Electromechanical Automation (AIEA)
会议时间：
01 August 2025
会议地点：
Hefei, China
会议论文集名称：
2025 6th International Conference on Artificial Intelligence and Electromechanical Automation (AIEA)
关键词：
Foggy weather;BIFPN;MANet;Dysample
摘要：
Due to reduced visibility and other reasons, the probability of accidents involving vehicles and pedestrians increases in foggy weather. In this paper, we present a real-time and efficient vehicle-pedestrian detection model for foggy images. In the backbone network, we added Mixed Aggregation Network (MANet) to replace C2f module to achieve stronger feature extraction. In the neck network, we introduced the Bidirectional Feature Pyramid Network (BIPFN) structure and used dysample for upsampling to achieve fast and efficient feature fusion in the model. In the loss function, we introduced Wise-IoU, which enables the model to focus more on samples with general quality, thereby improving overall performance. Our experiments were performed on the Real-world Task-Driven Testing Set (RTTS). Compared with the YOLOv10 model, our model has improved the detection accuracy by 3.5 %.
关键词：
2D;first-principles calculations;quantum anomalous Hall effect;spin-orbit coupling
摘要：
The pursuit of ferromagnetic semiconductors capable of realizing the quantum anomalous hall effect (QAHE) at room temperature holds significant importance for the development and application of spintronic devices. However, current experimental realizations of QAHE in 2D materials are often limited by extremely low TC and minute nontrivial bandgaps. Herein, based on first-principles calculations, a stable QAHE system that can exist at room temperature is successfully achieved by adsorbing N and O atoms on opposite sides of arsenene. According to the computational results, this novel 2D O & horbar;As & horbar;N system exhibits ferromagnetic semiconducting behavior with a TC of 350 K and a bandgap of approximate to 131 meV. Further calculations and analysis of the system's gapless chiral edge states, Chern number (C = 1), and quantized quantum Hall conductivity confirm the topological nontriviality of the bandgap. This work sheds light on the physical mechanisms for developing spintronic devices utilizing room-temperature ferromagnetic semiconductors and realizing lossless devices through the application of room-temperature QAHE.
摘要：
The exploration of multiphases and 0D/2D heterojunction in transition metal phosphides (TMPs) and transition metal sulfides (TMDs) is of major interest for hydrogen evolution reaction (HER). Herein, a novel combination route where 0D mixed-phased 1T/2H molybdenum sulfide quantum dots (MoS 2 QDs) are uniformly anchored on the 2D CoP x nanosheets is developed. MoS 2 QDs and CoP x were prepared via hydrothermal method and mixed with different ratios (Mo:Co ratios of 2:1, 1:1, and 1:2) and annealed under different temperatures to modulate their application in acidic HER processes. Specifically, 2Mo/1Co exhibited advanced performance for HER in 0.5 M H 2 SO 4 solution and required 14 mV to deliver 10 mA cm −2 and revealed a descended Tafel slope of 75.42 mV dec −1 with 240 h stability except obvious deactivation. The successful design and construction of 0D/2D mixed-dimensional materials would broaden the application of MoS 2 and CoP x for electrocatalytic hydrogen evolution.
The exploration of multiphases and 0D/2D heterojunction in transition metal phosphides (TMPs) and transition metal sulfides (TMDs) is of major interest for hydrogen evolution reaction (HER). Herein, a novel combination route where 0D mixed-phased 1T/2H molybdenum sulfide quantum dots (MoS 2 QDs) are uniformly anchored on the 2D CoP x nanosheets is developed. MoS 2 QDs and CoP x were prepared via hydrothermal method and mixed with different ratios (Mo:Co ratios of 2:1, 1:1, and 1:2) and annealed under different temperatures to modulate their application in acidic HER processes. Specifically, 2Mo/1Co exhibited advanced performance for HER in 0.5 M H 2 SO 4 solution and required 14 mV to deliver 10 mA cm −2 and revealed a descended Tafel slope of 75.42 mV dec −1 with 240 h stability except obvious deactivation. The successful design and construction of 0D/2D mixed-dimensional materials would broaden the application of MoS 2 and CoP x for electrocatalytic hydrogen evolution.
关键词：
GC-LSTM;trajectory reconstruction;wireless pipeline robots;multi-sensor fusion;inertial measurement unit
摘要：
Wireless pipeline robots often suffer from localization drift and position loss due to electromagnetic attenuation and shielding in complex pipeline configurations, which hinders accurate pipeline reconstruction. This paper proposes a trajectory reconstruction method based on Geometric Constraint–Long Short-Term Memory (GC-LSTM). First, a motor control system based on Field-Oriented Control (FOC) was developed for the proposed pipeline robot; second, trajectory errors are mitigated by exploiting pipeline geometric characteristics; third, a Long Short-Term Memory (LSTM) network is used to predict and compensate the robot’s velocity when odometer slip occurs; finally, multi-sensor fusion is employed to obtain the reconstructed trajectory. In straight-pipe tests, the GC-LSTM method reduced the maximum deviation and mean absolute deviation by 69.79% and 72.53%, respectively, compared with the Back Propagation (BP) method, resulting in a maximum deviation of 0.0933 m and a mean absolute deviation of 0.0351 m. In bend-pipe tests, GC-LSTM reduced the maximum deviation and the mean absolute deviation by 60.48% and 69.91%, respectively, compared with BP, yielding a maximum deviation of 0.2519 m and a mean absolute deviation of 0.0850 m. The proposed method significantly improves localization accuracy for wireless pipeline robots and enables more precise reconstruction of pipeline environments, providing a practical reference for accurate localization in pipeline inspection applications.
作者机构：
[Jiao, Qian] Wuhan Polytech Univ, Sch Elect & Elect Engn, Wuhan 430023, Peoples R China.;[Zhu, Ming; Yu, Haiyang; Ai, Mei] Chinese Acad Sci, Natl Astron Observ, Beijing 100101, Peoples R China.;[Zhu, Ming; Yu, Haiyang] Univ Chinese Acad Sci, Beijing 100049, Peoples R China.;[Zhu, Ming; Ai, Mei] Guizhou Univ, Guizhou Radio Astron Observ, Guiyang 550000, Peoples R China.;[Zhu, Ming; Ai, Mei] Chinese Acad Sci, CAS Key Lab FAST, Natl Astron Observ, Beijing 100101, Peoples R China.
通讯机构：
[Zhu, M ] C;[Jiao, Q ] W;Wuhan Polytech Univ, Sch Elect & Elect Engn, Wuhan 430023, Peoples R China.;Chinese Acad Sci, Natl Astron Observ, Beijing 100101, Peoples R China.;Univ Chinese Acad Sci, Beijing 100049, Peoples R China.
摘要：
We present the results of our recent H i observations conducted on the edge-on galaxy NGC 2683 using the Five-hundred-meter Aperture Spherical radio Telescope (FAST). In comparison to previous observations made by the Very Large Array (VLA), FAST has detected a more extensive distribution of H i. Particularly noteworthy is that the detections made by FAST extend approximately 4 times farther than those of the VLA in the vertical direction from the galactic plane. The total H i flux measured for NGC 2683 amounts to FH i = 112.1 Jy km s-1 (equivalent to a total H i mass of MH i = 2.32 x 109 M circle dot), which is slightly higher than that detected by VLA. FAST has also identified three dwarf galaxies in close proximity to NGC 2683, namely KK 69, NGC 2683dw1, and NGC 2683dw3? (hereafter dw3?). dw3? is situated within the extended H i distribution of NGC 2683 in projection and lies near the tail of KK 69 extending toward NGC 2683. These observations suggest that dw3? is likely a result of the accretion process from NGC 2683 to KK 69. Furthermore, FAST has detected three high-velocity clouds, with complex B potentially undergoing accretion with NGC 2683. Based on the model from B. Vollmer et al. and incorporating the H i halo component, we found that the model with the added H i halo aligns more closely with our FAST observations in NGC 2683. The estimated mass of this H i halo is 3 x 108 M circle dot, constituting approximately 13% of the total H i mass of the galaxy. We suggest that the origination of this H i halo is more likely attributed to external gas accretion.
摘要：
In order to plan suitable navigation operation paths for the characteristics of rice fields in the middle and lower reaches of the Yangtze River and the operational requirements of straw rotary burying, this paper proposes a combination of the Hough matrix and RANSAC algorithms to extract the starting routes of straw boundaries; the algorithm adopts the Hough matrix to extract the characteristic points of the straw boundaries and remove the redundancies, and then reduces the influence of noise points caused by different straw shapes using the RANSAC algorithm to improve the accuracy of the starting route extraction. The algorithm extracts the starting routes of straw boundaries and the characteristic points of the straw boundaries and removes the redundancies, so as to improve the accuracy of the starting route extraction. The extraction test shows that under different scenes, the recognition accuracy of the path extraction method combining the Hough matrix and RANSAC algorithm is above 90%, and the algorithm takes no more than 0.51 s. Finally, the road test shows that the method meets the characteristics of tractor operation with a large turning radius and without reversing and satisfies the unmanned operation requirements of straw rotary burying in the field.
摘要：
Detecting lung nodules through computed tomography (CT) scans plays an important role in early prevention, clinical diagnosis, and monitoring of lung cancer. However, the small, irregular shape and low resolution of lung nodules, together with multiscale challenges, often hinder accurate detection, especially for small nodules. To overcome these obstacles, we propose an improved detection algorithm, a real-time detection transformer (RT-DETR) nodule, specifically for lung nodules in CT images. The proposed RT-DETR framework includes several improvements aimed at increasing the detection accuracy. First, a RFAFasterNet is designed by replacing the backbone network with FasterNet network and incorporating the receptive field attention convolution (RFAConv) module. This integration improves the network’s capacity to capture both local and distant features, guided by the receptive field attention mechanism while effectively minimizing redundant computations. In addition, a small object detection layer (P2) is added to solve the problem of low-resolution small nodule detection. Finally, a hierarchical scale–based path aggregation network is proposed to further refine the detection of different nodule sizes by integrating deep-level and shallow-level features. Comprehensive experiments on the LUNA16 dataset demonstrate that the competition performance metric reaches 88.3%, which is 2.7% higher than the baseline model. In addition, the experimental results also show that the model has strong generalization in different datasets.
通讯机构：
[Xiang, ST ] W;Wuhan Polytech Univ, Sch Elect & Elect Engn, Wuhan 430048, Peoples R China.
关键词：
photovoltaic power generation;maximum power point tracking;Harris hawks optimization algorithm;partial shading;tent chaotic mapping;perturb and observe method
摘要：
This paper proposes an improved Harris Hawks Optimization (IHHO) algorithm to enhance maximum power point tracking (MPPT) performance in photovoltaic (PV) systems operating under various conditions. The IHHO introduces Tent chaotic mapping to improve population diversity and avoid premature convergence, a nonlinear decreasing inertia weight to dynamically balance exploration and exploitation, and a hybrid perturbation mechanism based on differential evolution to enhance local refinement. Additionally, a dynamic step-size adjustment and an escape energy mechanism responsive to irradiance changes improve real-time tracking adaptability. Parallel computing is employed to accelerate fitness evaluations and improve computational efficiency. Simulation results under multiple static and dynamic shading scenarios demonstrate that the proposed IHHO algorithm consistently achieves faster convergence, higher tracking accuracy, and stronger robustness than conventional methods such as particle swarm optimization (PSO) and Jaya. These results confirm the effectiveness of IHHO for reliable MPPT control in practical PV applications under diverse and challenging environmental conditions.
摘要：
The monolayer Janus FeClS, whose mirror inversion symmetry is broken due to its different non-metallic atomic compositions, exhibits spontaneous polarity and obvious spin polarization and has unique half-metal properties different from the corresponding pure monolayer FeCl 2 and FeS 2 , that is, the spin-up band appears as a semiconductor and the spin-down band appears as a metal. As the most common and effective means of structure and performance regulation, biaxial strain is applied to monolayer FeClS for study. Importantly, the strain-induced crystal structure changes cause different inter-hybridization of Fe-3 d electron orbitals in Janus FeClS, which leads to changes in electronic structures. In particular, the half-metal properties, dipole moments, and magnetic axis orientation of monolayer Janus FeClS remain basically unchanged under a certain degree of tensile and compressive strain, and the critical strain degree is basically the same when these properties are changed, indicating that Janus FeClS has a robust electronic structure and polarity. In addition, the magnetic moment and the absorption of visible light in monolayer FeClS are also regulated by biaxial strain. Therefore, this study expands the prospect and direction for the effective use and regulation of the excellent spin polarization characteristics of polar two-dimensional materials.
作者机构：
[Shuaishuai Li] School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan 430048, China;Author to whom correspondence should be addressed.;[Weizhen Chen] School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan 430048, China<&wdkj&>Author to whom correspondence should be addressed.
通讯机构：
[Weizhen Chen] S;School of Electrical and Electronic Engineering, Wuhan Polytechnic University, Wuhan 430048, China<&wdkj&>Author to whom correspondence should be addressed.
关键词：
power load forecasting;spatiotemporal feature fusion;attention mechanism;interpretability
摘要：
Driven by the global “double carbon” goal, the volatility of renewable energy poses a challenge to the stability of power systems. Traditional methods have difficulty dealing with high-dimensional nonlinear data, and the single deep learning model has the limitations of spatiotemporal feature decoupling and being a “black box”. Aiming at the problem of insufficient accuracy and interpretability of power load forecasting in a renewable energy grid connected scenario, this study proposes an interpretable spatiotemporal feature fusion model based on an attention mechanism. Through CNN layered extraction of multi-dimensional space–time features such as meteorology and electricity price, BiLSTM bi-directional modeling time series rely on capturing the evolution rules of load series before and after, and the improved self-attention mechanism dynamically focuses on key features. Combined with the SHAP quantitative feature contribution and feature deletion experiment, a complete chain of “feature extraction time series modeling weight allocation interpretation and verification” is constructed. The experimental results show that the determination coefficient R2 of the model on the Australian electricity market data set reaches 0.9935, which is 84.6% and 59.8% higher than that of the LSTM and GRU models, respectively. The prediction error (RMSE = 105.5079) is 9.7% lower than that of TCN-LSTM model and 52.1% compared to the GNN (220.6049). Cross scenario validation shows that the generalization performance is excellent (R2 ≥ 0.9849). The interpretability analysis reveals that electricity price (average absolute value of SHAP 716.7761) is the core influencing factor, and its lack leads to a 0.76% decline in R2. The research breaks through the limitation of time–space decoupling and the unexplainable bottleneck of traditional models, provides a transparent basis for power dispatching, and has an important reference value for the construction of new power systems.
期刊：
Review of Scientific Instruments,2025年96(6):064504 ISSN：0034-6748
通讯作者：
Blair, C
作者机构：
[Wang, Tailong] Wuhan Polytech Univ, Sch Elect & Elect Engn, Wuhan 430023, Peoples R China.;[Zhao, Chunnong; Blair, Carl; Al-Jodah, Ammar; Blair, C; Adam, Alexandra; Ju, Li; Liu, Jian; Goodwin-Jones, Aaron; Wang, Tailong; Winterflood, John] Univ Western Australia, OzGrav, Crawley, WA 6009, Australia.;[Goodwin-Jones, Aaron] CALTECH, Pasadena, CA 91125 USA.
通讯机构：
[Blair, C ] U;Univ Western Australia, OzGrav, Crawley, WA 6009, Australia.
摘要：
Seismic isolation is crucial for gravitational wave detectors as it minimizes ground vibrations, enabling the detection of faint gravitational wave signals. An active seismic isolation platform for precision measurement experiments is described. The table features piezo actuation along five degrees of freedom: three translation and two tip-tilt degrees of freedom. It is stiff in rotation about the vertical axes. A seismometer is used to sense table motion. Piezo actuators are used to suppress seismic noise with feedback control bandwidth of 0.3-3 Hz. Suppression levels ranging from 21 to 36 dB of seismic noise within the frequency range of 0.5-1.3 Hz are demonstrated, as measured by a witness seismometer on the table, with the suppression direction along the optic axis of the suspended mirror on the table. The suppression results in 1 nm/root Hz residual horizontal motion at 1 Hz. Limitations such as tilt-to-translation coupling that prevent actuation over the desired frequency range are discussed.
关键词：
Image deblurring;self-attention mechanism;lightweight model;hinge loss function
摘要：
The acquisition of clear images is a critical aspect in various fields including computer vision, aerial detection, and medical imaging. The issue of image blur caused by object motion poses a challenge in obtaining clear images. To address this, an improved AT-DGAN network model is proposed in this paper. This model integrates the pyramid generator module of the DeblurGAN-v2 network with a self-attention mechanism. The feature pyramid is employed for image feature extraction and representation, while the self-attention mechanism dynamically adjusts the weight of important features in each pyramid layer and performs weighted fusion, thereby compensating for the information loss during feature extraction in the feature pyramid network. Additionally, a hinge loss function is designed for the proposed model to balance the discriminator and the generator, enhancing the stability and training efficiency of the generative adversarial network. The experimental results show that compared to other algorithms of the same type, this improved algorithm has increased the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) of restored images by 0.58 dB and 1.5%, respectively.
