To ease the scarcity of lithium(Li)resource and cut down on environmental pollution,an efficient,selective,inexpensive and sustainable Li recycling process from waste batteries is needed,which is yet to be achieved.He...To ease the scarcity of lithium(Li)resource and cut down on environmental pollution,an efficient,selective,inexpensive and sustainable Li recycling process from waste batteries is needed,which is yet to be achieved.Here,we report a low-potential photoelectrochemical(PEC)system that selectively and efficiently extracts Li metals from multi-cation electrolytes under 1 sun illumination.Based on the difference of redox potential,we can get rid of the disturbance of other cations(i.e.,Fe,Co and Ni ions)by a bias-free PEC device to realize the extraction of high-purity Li metals on a coplanar Si-based photocathode-TiO_(2) photoanode tandem device at 2 V of applied bias(far less than the redox potentials of Li^(+)/Li).In such system,the extraction rate of Li metals(purity>99.5%)exceeds 1.35 g h^(-1)m^(-2)with 90%of Faradaic efficiency.Long-term experiments,different electrode/electrolyte tests,and various price assessments further demonstrate the stability,compatibility and economy of PEC extraction system,enabling a solar-driven pathway for the recycling of critical metal resources.展开更多
Wide-temperature applications of sodium-ion batteries(SIBs)are severely limited by the sluggish ion insertion/diffusion kinetics of conversion-type anodes.Quantum-sized transition metal dichalcogenides possess unique ...Wide-temperature applications of sodium-ion batteries(SIBs)are severely limited by the sluggish ion insertion/diffusion kinetics of conversion-type anodes.Quantum-sized transition metal dichalcogenides possess unique advantages of charge delocalization and enrich uncoordinated electrons and short-range transfer kinetics,which are crucial to achieve rapid low-temperature charge transfer and high-temperature interface stability.Herein,a quantum-scale FeS_(2) loaded on three-dimensional Ti_(3)C_(2) MXene skeletons(FeS_(2) QD/MXene)fabricated as SIBs anode,demonstrating impressive performance under wide-temperature conditions(−35 to 65).The theoretical calculations combined with experimental characterization interprets that the unsaturated coordination edges of FeS_(2) QD can induce delocalized electronic regions,which reduces electrostatic potential and significantly facilitates efficient Na+diffusion across a broad temperature range.Moreover,the Ti_(3)C_(2) skeleton reinforces structural integrity via Fe-O-Ti bonding,while enabling excellent dispersion of FeS_(2) QD.As expected,FeS_(2) QD/MXene anode harvests capacities of 255.2 and 424.9 mAh g^(−1) at 0.1 A g^(−1) under−35 and 65,and the energy density of FeS_(2) QD/MXene//NVP full cell can reach to 162.4 Wh kg^(−1) at−35,highlighting its practical potential for wide-temperatures conditions.This work extends the uncoordinated regions induced by quantum-size effects for exceptional Na^(+)ion storage and diffusion performance at wide-temperatures environment.展开更多
Strategically coupling nanoparticle hybrids and internal thermosensitive molecular switches establishes an innovative paradigm for constructing micro/nanoscale-reconfigurable robots,facilitating energyefficient CO_(2)...Strategically coupling nanoparticle hybrids and internal thermosensitive molecular switches establishes an innovative paradigm for constructing micro/nanoscale-reconfigurable robots,facilitating energyefficient CO_(2) management in life-support systems of confined space.Here,a micro/nano-reconfigurable robot is constructed from the CO_(2) molecular hunters,temperature-sensitive molecular switch,solar photothermal conversion,and magnetically-driven function engines.The molecular hunters within the molecular extension state can capture 6.19 mmol g^(−1) of CO_(2) to form carbamic acid and ammonium bicarbonate.Interestingly,the molecular switch of the robot activates a molecular curling state that facilitates CO_(2) release through nano-reconfiguration,which is mediated by the temperature-sensitive curling of Pluronic F127 molecular chains during the photothermal desorption.Nano-reconfiguration of robot alters the amino microenvironment,including increasing surface electrostatic potential of the amino group and decreasing overall lowest unoccupied molecular orbital energy level.This weakened the nucleophilic attack ability of the amino group toward the adsorption product derivatives,thereby inhibiting the side reactions that generate hard-to-decompose urea structures,achieving the lowest regeneration temperature of 55℃ reported to date.The engine of the robot possesses non-contact magnetically-driven micro-reconfiguration capability to achieve efficient photothermal regeneration while avoiding local overheating.Notably,the robot successfully prolonged the survival time of mice in the sealed container by up to 54.61%,effectively addressing the issue of carbon suffocation in confined spaces.This work significantly enhances life-support systems for deep-space exploration,while stimulating innovations in sustainable carbon management technologies for terrestrial extreme environments.展开更多
The personalized fine-tuning of large languagemodels(LLMs)on edge devices is severely constrained by limited computation resources.Although split federated learning alleviates on-device burdens,its effectiveness dimin...The personalized fine-tuning of large languagemodels(LLMs)on edge devices is severely constrained by limited computation resources.Although split federated learning alleviates on-device burdens,its effectiveness diminishes in few-shot reasoning scenarios due to the low data efficiency of conventional supervised fine-tuning,which leads to excessive communication overhead.To address this,we propose Language-Empowered Split Fine-Tuning(LESFT),a framework that integrates split architectures with a contrastive-inspired fine-tuning paradigm.LESFT simultaneously learns frommultiple logically equivalent but linguistically diverse reasoning chains,providing richer supervisory signals and improving data efficiency.This process-oriented training allows more effective reasoning adaptation with fewer samples.Extensive experiments demonstrate that LESFT consistently outperforms strong baselines such as SplitLoRA in task accuracy.LESFT consistently outperforms strong baselines on GSM8K,CommonsenseQA,and AQUA_RAT,with the largest gains observed on Qwen2.5-3B.These results indicate that LESFT can effectively adapt large language models for reasoning tasks under the computational and communication constraints of edge environments.展开更多
In rock engineering,natural cracks in rock masses subjected to external loads tend to initiate and propagate,leading to potential safety hazards.To investigate the effect of cracking behavior on the mechanical propert...In rock engineering,natural cracks in rock masses subjected to external loads tend to initiate and propagate,leading to potential safety hazards.To investigate the effect of cracking behavior on the mechanical properties of rocks,the cracking processes of pre-cracked rocks have been extensively studied using numerical modeling methods.The peridynamics(PD)exhibits advantages over other numerical methods due to the absence of the requirements for remeshing and external crack growth criterion.However,for modeling pre-cracked rock cracking processes under impact,current PD implementations lack generally applicable rock constitutive models and impact contact models,which leads to difficulties in determining rock material parameters and efficiently calculating impact loads.This paper proposes a non-ordinary state-based peridynamics(NOSBPD)modeling method integrating the Drucker-Prager(DP)plasticity model and an efficient contact model to address the above problems.In the proposed method,the Drucker-Prager plasticity model is integrated into the NOSBPD,thereby equipping NOSBPD with the capability to accurately characterize the nonlinear stress-strain relationship inherent in rocks.An efficient contact model between particles and meshes is designed to calculate the impact loads,which is essentially a coupling method of PD with the finite element method(FEM).The effectiveness of the proposed NOSBPD modeling method is verified by comparison with other numerical methods and experiments.Experimental results indicate that the proposed method can effectively and accurately predict the 3D cracking processes of pre-cracked cracks under impact loading,and the maximum principal stress is the key driver behind wing crack formation in pre-cracked rocks.展开更多
Federated learning often experiences slow and unstable convergence due to edge-side data heterogeneity.This problem becomes more severe when edge participation rate is low,as the information collected from different e...Federated learning often experiences slow and unstable convergence due to edge-side data heterogeneity.This problem becomes more severe when edge participation rate is low,as the information collected from different edge devices varies significantly.As a result,communication overhead increases,which further slows down the convergence process.To address this challenge,we propose a simple yet effective federated learning framework that improves consistency among edge devices.The core idea is clusters the lookahead gradients collected from edge devices on the cloud server to obtain personalized momentum for steering local updates.In parallel,a global momentum is applied during model aggregation,enabling faster convergence while preserving personalization.This strategy enables efficient propagation of the estimated global update direction to all participating edge devices and maintains alignment in local training,without introducing extra memory or communication overhead.We conduct extensive experiments on benchmark datasets such as Cifar100 and Tiny-ImageNet.The results confirm the effectiveness of our framework.On CIFAR-100,our method reaches 55%accuracy with 37 fewer rounds and achieves a competitive final accuracy of 65.46%.Even under extreme non-IID scenarios,it delivers significant improvements in both accuracy and communication efficiency.The implementation is publicly available at https://github.com/sjmp525/CollaborativeComputing/tree/FedCCM(accessed on 20 October 2025).展开更多
Water scarcity,exacerbated by organic micropollutant contamination and climate change,necessitates energy-efficient,eco-friendly purification technologies.Membrane separation has emerged as a transformative solution,o...Water scarcity,exacerbated by organic micropollutant contamination and climate change,necessitates energy-efficient,eco-friendly purification technologies.Membrane separation has emerged as a transformative solution,outperforming energyintensive processes,such as distillation.As highlighted in Nature,traditional chemical separations—dominated by distillation—consume 10%-15%of global energy,whereas advanced membrane technologies can reduce energy use by up to 90%.展开更多
The Hainan Free Trade Port(FTP),with the launching of island-wide special customs operations,is allowing greater convenience for overseas goods to enter the island,trying to build it into an efficient hub connecting C...The Hainan Free Trade Port(FTP),with the launching of island-wide special customs operations,is allowing greater convenience for overseas goods to enter the island,trying to build it into an efficient hub connecting China with the rest of the world.Its core appeal no longer relies solely on policy incentives but has shifted toward transparent and predictable institutional guarantees which provide stable expectations for global investors.展开更多
车辆检测是智能交通系统和自动驾驶的重要组成部分。然而,实际交通场景中存在许多不确定因素,导致车辆检测模型的准确率低实时性差。为了解决这个问题,提出了一种快速准确的车辆检测算法——YOLOv8-DEL。使用DGCST(dynamic group convol...车辆检测是智能交通系统和自动驾驶的重要组成部分。然而,实际交通场景中存在许多不确定因素,导致车辆检测模型的准确率低实时性差。为了解决这个问题,提出了一种快速准确的车辆检测算法——YOLOv8-DEL。使用DGCST(dynamic group convolution shuffle transformer)模块代替C2f模块来重构主干网络,以增强特征提取能力并使网络更轻量;添加的P2检测层能使模型更敏锐地定位和检测小目标,同时采用Efficient RepGFPN进行多尺度特征融合,以丰富特征信息并提高模型的特征表达能力;通过结合GroupNorm和共享卷积的优点,设计了一种轻量型共享卷积检测头,在保持精度的前提下,有效减少参数量并提升检测速度。与YOLOv8相比,提出的YOLOv8-DEL在BDD100K数据集和KITTI数据集上,mAP@0.5分别提高了4.8个百分点和1.2个百分点,具有实时检测速度(208.6 FPS和216.4 FPS),在检测精度和速度方面实现了更有利的折中。展开更多
Traditional data-driven fault diagnosis methods depend on expert experience to manually extract effective fault features of signals,which has certain limitations.Conversely,deep learning techniques have gained promine...Traditional data-driven fault diagnosis methods depend on expert experience to manually extract effective fault features of signals,which has certain limitations.Conversely,deep learning techniques have gained prominence as a central focus of research in the field of fault diagnosis by strong fault feature extraction ability and end-to-end fault diagnosis efficiency.Recently,utilizing the respective advantages of convolution neural network(CNN)and Transformer in local and global feature extraction,research on cooperating the two have demonstrated promise in the field of fault diagnosis.However,the cross-channel convolution mechanism in CNN and the self-attention calculations in Transformer contribute to excessive complexity in the cooperative model.This complexity results in high computational costs and limited industrial applicability.To tackle the above challenges,this paper proposes a lightweight CNN-Transformer named as SEFormer for rotating machinery fault diagnosis.First,a separable multiscale depthwise convolution block is designed to extract and integrate multiscale feature information from different channel dimensions of vibration signals.Then,an efficient self-attention block is developed to capture critical fine-grained features of the signal from a global perspective.Finally,experimental results on the planetary gearbox dataset and themotor roller bearing dataset prove that the proposed framework can balance the advantages of robustness,generalization and lightweight compared to recent state-of-the-art fault diagnosis models based on CNN and Transformer.This study presents a feasible strategy for developing a lightweight rotating machinery fault diagnosis framework aimed at economical deployment.展开更多
现有基于红外图像的绝缘子发热缺陷识别方法,存在目标区域提取精度有限、温度提取受环境因素影响较大等问题。为此,本文提出一种复合绝缘子过热识别新方法:首先改进单阶段绝缘子实例分割算法You Only Look At CoefficienTs(YOLACT),引...现有基于红外图像的绝缘子发热缺陷识别方法,存在目标区域提取精度有限、温度提取受环境因素影响较大等问题。为此,本文提出一种复合绝缘子过热识别新方法:首先改进单阶段绝缘子实例分割算法You Only Look At CoefficienTs(YOLACT),引入嵌有Efficient Local Attention(ELA)机制的MobileNetV2作主干网络提升检测速度,融合特征金字塔网络(Feature Pyramid Network,FPN)各层特征图并加入聚焦纯卷积特征提取模块提高特征图质量;然后使用改进算法识别红外图像中复合绝缘子外轮廓,定位其棒芯位置;最后依据红外图像热矩阵获取棒芯温度矩阵,对比温度变化判断是否异常。实际生产环境中,本文方法整体准确率达到975,算法总耗时125ms;改进实例分割算法平均交互比(mIOU)为9297,平均像素准确率(mPA)为9615,每秒帧数(FPS)为19。结果显示,此方法分割定位效果好,能滤除多数环境因素导致的温度识别误差,为绝缘子温度异常识别提供新方案。展开更多
Generating dynamically feasible trajectory for fixed-wing Unmanned Aerial Vehicles(UAVs)in dense obstacle environments remains computationally intractable.This paper proposes a Safe Flight Corridor constrained Sequent...Generating dynamically feasible trajectory for fixed-wing Unmanned Aerial Vehicles(UAVs)in dense obstacle environments remains computationally intractable.This paper proposes a Safe Flight Corridor constrained Sequential Convex Programming(SFC-SCP)to improve the computation efficiency and reliability of trajectory generation.SFC-SCP combines the front-end convex polyhedron SFC construction and back-end SCP-based trajectory optimization.A Sparse A^(*)Search(SAS)driven SFC construction method is designed to efficiently generate polyhedron SFC according to the geometric relation among obstacles and collision-free waypoints.Via transforming the nonconvex obstacle-avoidance constraints to linear inequality constraints,SFC can mitigate infeasibility of trajectory planning and reduce computation complexity.Then,SCP casts the nonlinear trajectory optimization subject to SFC into convex programming subproblems to decrease the problem complexity.In addition,a convex optimizer based on interior point method is customized,where the search direction is calculated via successive elimination to further improve efficiency.Simulation experiments on dense obstacle scenarios show that SFC-SCP can generate dynamically feasible safe trajectory rapidly.Comparative studies with state-of-the-art SCP-based methods demonstrate the efficiency and reliability merits of SFC-SCP.Besides,the customized convex optimizer outperforms off-the-shelf optimizers in terms of computation time.展开更多
基金the National Natural Science Foundation of China(22479047,22409058)the Outstanding Youth Scientist Foundation of Hunan Province(2022JJ10023)the Provincial Natural Science Foundation of Guangdong(2023A1515011745)for financial support of this research。
文摘To ease the scarcity of lithium(Li)resource and cut down on environmental pollution,an efficient,selective,inexpensive and sustainable Li recycling process from waste batteries is needed,which is yet to be achieved.Here,we report a low-potential photoelectrochemical(PEC)system that selectively and efficiently extracts Li metals from multi-cation electrolytes under 1 sun illumination.Based on the difference of redox potential,we can get rid of the disturbance of other cations(i.e.,Fe,Co and Ni ions)by a bias-free PEC device to realize the extraction of high-purity Li metals on a coplanar Si-based photocathode-TiO_(2) photoanode tandem device at 2 V of applied bias(far less than the redox potentials of Li^(+)/Li).In such system,the extraction rate of Li metals(purity>99.5%)exceeds 1.35 g h^(-1)m^(-2)with 90%of Faradaic efficiency.Long-term experiments,different electrode/electrolyte tests,and various price assessments further demonstrate the stability,compatibility and economy of PEC extraction system,enabling a solar-driven pathway for the recycling of critical metal resources.
基金supported by the National Nature Science Foundation of China(Nos.52202335 and 52171227)Natural Science Foundation of Jiangsu Province(No.BK20221137)National Key R&D Program of China(2024YFE0108500).
文摘Wide-temperature applications of sodium-ion batteries(SIBs)are severely limited by the sluggish ion insertion/diffusion kinetics of conversion-type anodes.Quantum-sized transition metal dichalcogenides possess unique advantages of charge delocalization and enrich uncoordinated electrons and short-range transfer kinetics,which are crucial to achieve rapid low-temperature charge transfer and high-temperature interface stability.Herein,a quantum-scale FeS_(2) loaded on three-dimensional Ti_(3)C_(2) MXene skeletons(FeS_(2) QD/MXene)fabricated as SIBs anode,demonstrating impressive performance under wide-temperature conditions(−35 to 65).The theoretical calculations combined with experimental characterization interprets that the unsaturated coordination edges of FeS_(2) QD can induce delocalized electronic regions,which reduces electrostatic potential and significantly facilitates efficient Na+diffusion across a broad temperature range.Moreover,the Ti_(3)C_(2) skeleton reinforces structural integrity via Fe-O-Ti bonding,while enabling excellent dispersion of FeS_(2) QD.As expected,FeS_(2) QD/MXene anode harvests capacities of 255.2 and 424.9 mAh g^(−1) at 0.1 A g^(−1) under−35 and 65,and the energy density of FeS_(2) QD/MXene//NVP full cell can reach to 162.4 Wh kg^(−1) at−35,highlighting its practical potential for wide-temperatures conditions.This work extends the uncoordinated regions induced by quantum-size effects for exceptional Na^(+)ion storage and diffusion performance at wide-temperatures environment.
基金supported by the National Natural Science Foundation of China(22168008,22378085)the Guangxi Natural Science Foundation(2024GXNSFDA010053)+1 种基金the Technology Development Project of Guangxi Bossco Environmental Protection Technology Co.,Ltd(202100039)Innovation Project of Guangxi Graduate Education(YCBZ2024065).
文摘Strategically coupling nanoparticle hybrids and internal thermosensitive molecular switches establishes an innovative paradigm for constructing micro/nanoscale-reconfigurable robots,facilitating energyefficient CO_(2) management in life-support systems of confined space.Here,a micro/nano-reconfigurable robot is constructed from the CO_(2) molecular hunters,temperature-sensitive molecular switch,solar photothermal conversion,and magnetically-driven function engines.The molecular hunters within the molecular extension state can capture 6.19 mmol g^(−1) of CO_(2) to form carbamic acid and ammonium bicarbonate.Interestingly,the molecular switch of the robot activates a molecular curling state that facilitates CO_(2) release through nano-reconfiguration,which is mediated by the temperature-sensitive curling of Pluronic F127 molecular chains during the photothermal desorption.Nano-reconfiguration of robot alters the amino microenvironment,including increasing surface electrostatic potential of the amino group and decreasing overall lowest unoccupied molecular orbital energy level.This weakened the nucleophilic attack ability of the amino group toward the adsorption product derivatives,thereby inhibiting the side reactions that generate hard-to-decompose urea structures,achieving the lowest regeneration temperature of 55℃ reported to date.The engine of the robot possesses non-contact magnetically-driven micro-reconfiguration capability to achieve efficient photothermal regeneration while avoiding local overheating.Notably,the robot successfully prolonged the survival time of mice in the sealed container by up to 54.61%,effectively addressing the issue of carbon suffocation in confined spaces.This work significantly enhances life-support systems for deep-space exploration,while stimulating innovations in sustainable carbon management technologies for terrestrial extreme environments.
基金supported in part by the National Natural Science Foundation of China(NSFC)under Grant 62276109The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this work through the Research Group Project number(ORF-2025-585).
文摘The personalized fine-tuning of large languagemodels(LLMs)on edge devices is severely constrained by limited computation resources.Although split federated learning alleviates on-device burdens,its effectiveness diminishes in few-shot reasoning scenarios due to the low data efficiency of conventional supervised fine-tuning,which leads to excessive communication overhead.To address this,we propose Language-Empowered Split Fine-Tuning(LESFT),a framework that integrates split architectures with a contrastive-inspired fine-tuning paradigm.LESFT simultaneously learns frommultiple logically equivalent but linguistically diverse reasoning chains,providing richer supervisory signals and improving data efficiency.This process-oriented training allows more effective reasoning adaptation with fewer samples.Extensive experiments demonstrate that LESFT consistently outperforms strong baselines such as SplitLoRA in task accuracy.LESFT consistently outperforms strong baselines on GSM8K,CommonsenseQA,and AQUA_RAT,with the largest gains observed on Qwen2.5-3B.These results indicate that LESFT can effectively adapt large language models for reasoning tasks under the computational and communication constraints of edge environments.
基金support from the National Natural Science Foundation of China(Grant Nos.42277161 and 42230709).
文摘In rock engineering,natural cracks in rock masses subjected to external loads tend to initiate and propagate,leading to potential safety hazards.To investigate the effect of cracking behavior on the mechanical properties of rocks,the cracking processes of pre-cracked rocks have been extensively studied using numerical modeling methods.The peridynamics(PD)exhibits advantages over other numerical methods due to the absence of the requirements for remeshing and external crack growth criterion.However,for modeling pre-cracked rock cracking processes under impact,current PD implementations lack generally applicable rock constitutive models and impact contact models,which leads to difficulties in determining rock material parameters and efficiently calculating impact loads.This paper proposes a non-ordinary state-based peridynamics(NOSBPD)modeling method integrating the Drucker-Prager(DP)plasticity model and an efficient contact model to address the above problems.In the proposed method,the Drucker-Prager plasticity model is integrated into the NOSBPD,thereby equipping NOSBPD with the capability to accurately characterize the nonlinear stress-strain relationship inherent in rocks.An efficient contact model between particles and meshes is designed to calculate the impact loads,which is essentially a coupling method of PD with the finite element method(FEM).The effectiveness of the proposed NOSBPD modeling method is verified by comparison with other numerical methods and experiments.Experimental results indicate that the proposed method can effectively and accurately predict the 3D cracking processes of pre-cracked cracks under impact loading,and the maximum principal stress is the key driver behind wing crack formation in pre-cracked rocks.
基金supported by the National Natural Science Foundation of China(62462040)the Yunnan Fundamental Research Projects(202501AT070345)the Major Science and Technology Projects in Yunnan Province(202202AD080013).
文摘Federated learning often experiences slow and unstable convergence due to edge-side data heterogeneity.This problem becomes more severe when edge participation rate is low,as the information collected from different edge devices varies significantly.As a result,communication overhead increases,which further slows down the convergence process.To address this challenge,we propose a simple yet effective federated learning framework that improves consistency among edge devices.The core idea is clusters the lookahead gradients collected from edge devices on the cloud server to obtain personalized momentum for steering local updates.In parallel,a global momentum is applied during model aggregation,enabling faster convergence while preserving personalization.This strategy enables efficient propagation of the estimated global update direction to all participating edge devices and maintains alignment in local training,without introducing extra memory or communication overhead.We conduct extensive experiments on benchmark datasets such as Cifar100 and Tiny-ImageNet.The results confirm the effectiveness of our framework.On CIFAR-100,our method reaches 55%accuracy with 37 fewer rounds and achieves a competitive final accuracy of 65.46%.Even under extreme non-IID scenarios,it delivers significant improvements in both accuracy and communication efficiency.The implementation is publicly available at https://github.com/sjmp525/CollaborativeComputing/tree/FedCCM(accessed on 20 October 2025).
文摘Water scarcity,exacerbated by organic micropollutant contamination and climate change,necessitates energy-efficient,eco-friendly purification technologies.Membrane separation has emerged as a transformative solution,outperforming energyintensive processes,such as distillation.As highlighted in Nature,traditional chemical separations—dominated by distillation—consume 10%-15%of global energy,whereas advanced membrane technologies can reduce energy use by up to 90%.
文摘The Hainan Free Trade Port(FTP),with the launching of island-wide special customs operations,is allowing greater convenience for overseas goods to enter the island,trying to build it into an efficient hub connecting China with the rest of the world.Its core appeal no longer relies solely on policy incentives but has shifted toward transparent and predictable institutional guarantees which provide stable expectations for global investors.
基金supported by the National Natural Science Foundation of China(No.52277055).
文摘Traditional data-driven fault diagnosis methods depend on expert experience to manually extract effective fault features of signals,which has certain limitations.Conversely,deep learning techniques have gained prominence as a central focus of research in the field of fault diagnosis by strong fault feature extraction ability and end-to-end fault diagnosis efficiency.Recently,utilizing the respective advantages of convolution neural network(CNN)and Transformer in local and global feature extraction,research on cooperating the two have demonstrated promise in the field of fault diagnosis.However,the cross-channel convolution mechanism in CNN and the self-attention calculations in Transformer contribute to excessive complexity in the cooperative model.This complexity results in high computational costs and limited industrial applicability.To tackle the above challenges,this paper proposes a lightweight CNN-Transformer named as SEFormer for rotating machinery fault diagnosis.First,a separable multiscale depthwise convolution block is designed to extract and integrate multiscale feature information from different channel dimensions of vibration signals.Then,an efficient self-attention block is developed to capture critical fine-grained features of the signal from a global perspective.Finally,experimental results on the planetary gearbox dataset and themotor roller bearing dataset prove that the proposed framework can balance the advantages of robustness,generalization and lightweight compared to recent state-of-the-art fault diagnosis models based on CNN and Transformer.This study presents a feasible strategy for developing a lightweight rotating machinery fault diagnosis framework aimed at economical deployment.
文摘现有基于红外图像的绝缘子发热缺陷识别方法,存在目标区域提取精度有限、温度提取受环境因素影响较大等问题。为此,本文提出一种复合绝缘子过热识别新方法:首先改进单阶段绝缘子实例分割算法You Only Look At CoefficienTs(YOLACT),引入嵌有Efficient Local Attention(ELA)机制的MobileNetV2作主干网络提升检测速度,融合特征金字塔网络(Feature Pyramid Network,FPN)各层特征图并加入聚焦纯卷积特征提取模块提高特征图质量;然后使用改进算法识别红外图像中复合绝缘子外轮廓,定位其棒芯位置;最后依据红外图像热矩阵获取棒芯温度矩阵,对比温度变化判断是否异常。实际生产环境中,本文方法整体准确率达到975,算法总耗时125ms;改进实例分割算法平均交互比(mIOU)为9297,平均像素准确率(mPA)为9615,每秒帧数(FPS)为19。结果显示,此方法分割定位效果好,能滤除多数环境因素导致的温度识别误差,为绝缘子温度异常识别提供新方案。
基金supported by the National Natural Science Foundation of China(No.62203256)。
文摘Generating dynamically feasible trajectory for fixed-wing Unmanned Aerial Vehicles(UAVs)in dense obstacle environments remains computationally intractable.This paper proposes a Safe Flight Corridor constrained Sequential Convex Programming(SFC-SCP)to improve the computation efficiency and reliability of trajectory generation.SFC-SCP combines the front-end convex polyhedron SFC construction and back-end SCP-based trajectory optimization.A Sparse A^(*)Search(SAS)driven SFC construction method is designed to efficiently generate polyhedron SFC according to the geometric relation among obstacles and collision-free waypoints.Via transforming the nonconvex obstacle-avoidance constraints to linear inequality constraints,SFC can mitigate infeasibility of trajectory planning and reduce computation complexity.Then,SCP casts the nonlinear trajectory optimization subject to SFC into convex programming subproblems to decrease the problem complexity.In addition,a convex optimizer based on interior point method is customized,where the search direction is calculated via successive elimination to further improve efficiency.Simulation experiments on dense obstacle scenarios show that SFC-SCP can generate dynamically feasible safe trajectory rapidly.Comparative studies with state-of-the-art SCP-based methods demonstrate the efficiency and reliability merits of SFC-SCP.Besides,the customized convex optimizer outperforms off-the-shelf optimizers in terms of computation time.
