This review paper examines the various types of electrical generators used to convert wave energy into electrical energy.The focus is on both linear and rotary generators,including their design principles,operational ...This review paper examines the various types of electrical generators used to convert wave energy into electrical energy.The focus is on both linear and rotary generators,including their design principles,operational efficiencies,and technological advancements.Linear generators,such as Induction,permanent magnet synchronous,and switched reluctance types,are highlighted for their direct conversion capability,eliminating the need for mechanical gearboxes.Rotary Induction generators,permanent magnet synchronous generators,and doubly-fed Induction generators are evaluated for their established engineering principles and integration with existing grid infrastructure.The paper discusses the historical development,environmental benefits,and ongoing advancements in wave energy technologies,emphasizing the increasing feasibility and scalability of wave energy as a renewable source.Through a comprehensive analysis,this review provides insights into the current state and future prospects of electrical generators in wave energy conversion,underscoring their potential to significantly reduce reliance on fossil fuels and mitigate environmental impacts.展开更多
Converter transformers are the core components of ultra-high voltage(UHV)transmission systems.The main cause of faults in converter transformers is irreversible deterioration of oil-pressboard insulation under combine...Converter transformers are the core components of ultra-high voltage(UHV)transmission systems.The main cause of faults in converter transformers is irreversible deterioration of oil-pressboard insulation under combined electrical-thermal-mechanical stress over long operating times.In this paper,the chemical characteristics of oil-pressboard insulation samples subjected to electrical-thermal-mechanical ageing for different times are studied.An image processing algorithm is used to analyse the discharge propagation characteristics of the samples under combined alternating current(AC)-direct current(DC)voltage,and the current pulse curves and phase resolved partial discharge spectrogram corresponding to the discharge images are analysed.An improved wavelet packet algorithm is used to denoise the discharge current pulse.Finally,the influence of electrical-thermal-mechanical ageing on discharge characteristics is analysed using radar charts.The condition of oil-pressboard insulation is one of the main factors determining the life expectancy of converter transformers.The results obtained here therefore have practical significance for understanding the process of insulation failure caused by accelerated ageing of oil-pressboard insulation.展开更多
Reconfigurable surface acoustic wave(SAW)phase shifters have garnered significant attention owing to their potential applications in emerging fields such as secure wireless communication,adaptable signal processing,an...Reconfigurable surface acoustic wave(SAW)phase shifters have garnered significant attention owing to their potential applications in emerging fields such as secure wireless communication,adaptable signal processing,and intelligent sensing systems.Among various modulation methods,employing gate voltage-controlled tuning methodologies that leverage acoustoelectric interactions has proven to be an efficient modulation approach that requires a low bias voltage.However,current acoustoelectric devices suffer from limited tunability,intricate heterogeneous structures,and complex manufacturing processes,all of which impede their practical applications.In this study,we present a novel material system for voltage-tunable SAW phase shifters.This system incorporates an atomic layer deposition ZnO thin-film transistors on LiNbO_(3)structure.This structure combines the benefits of LiNbO_(3)'s high electromechanical coupling coefficient(K^(2))and ZnO's superior conductivity adjustability.Besides,the device possesses a simplified structural configuration,which is easy to fabricate.Devices with different mesa lengths were fabricated and measured,and two of the different modes were compared.The results indicate that both the maximum phase shift and attenuation of the Rayleigh mode and longitudinal leaky SAW(LLSAW)increase proportionally with mesa length.Furthermore,LLSAW with larger effective electromechanical coupling coefficients(K_(eff)^(2))values exhibits greater phase velocity shifts and attenuation coefficients,with a maximum phase velocity tuning of 1.22%achieved.It is anticipated that the proposed devices will find utility in a variety of applications necessitating tunable acoustic components.展开更多
Noise is inevitable in electrical capacitance tomography(ECT)measurements.This paper describes the influence of noise on ECT performance for measuring gas-solids fluidized bed characteristics.The noise distribution is...Noise is inevitable in electrical capacitance tomography(ECT)measurements.This paper describes the influence of noise on ECT performance for measuring gas-solids fluidized bed characteristics.The noise distribution is approximated by the Gaussian distribution and added to experimental capacitance data with various intensities.The equivalent signal strength(Ф)that equals the signal-to-noise ratio of packed beds is used to evaluate noise levels.Results show that the Pearson correlation coefficient,which indicates the similarity of solids fraction distributions over pixels,increases with Ф,and reconstructed images are more deteriorated at lower Ф.Nevertheless,relative errors for average solids fraction and bubble size in each frame are less sensitive to noise,attributed to noise compromise caused by the process of pixel values.These findings provide useful guidance for assessing the accuracy of ECT measurements of multiphase flows.展开更多
Optimizing routing and resource allocation in decentralized unmanned aerial vehicle(UAV)networks remains challenging due to interference and rapidly changing topologies.The authors introduce a novel framework combinin...Optimizing routing and resource allocation in decentralized unmanned aerial vehicle(UAV)networks remains challenging due to interference and rapidly changing topologies.The authors introduce a novel framework combining double deep Q-networks(DDQNs)and graph neural networks(GNNs)for joint routing and resource allocation.The framework uses GNNs to model the network topology and DDQNs to adaptively control routing and resource allocation,addressing interference and improving network performance.Simulation results show that the proposed approach outperforms traditional methods such as Closest-to-Destination(c2Dst),Max-SINR(mSINR),and Multi-Layer Perceptron(MLP)-based models,achieving approximately 23.5% improvement in throughput,50% increase in connection probability,and 17.6% reduction in number of hops,demonstrating its effectiveness in dynamic UAV networks.展开更多
Microelectromechanical systems(MEMS)technology has gained significant attention over the past decade for measuring inertial angular velocity.However,due to inherent complexity,MEMS gyroscopes typically feature up to t...Microelectromechanical systems(MEMS)technology has gained significant attention over the past decade for measuring inertial angular velocity.However,due to inherent complexity,MEMS gyroscopes typically feature up to ten times more parameters than traditional sensors,making selection a challenging task even for experts.This study addresses this challenge,focusing on defensive guidance,navigation,and control(GNC)systems where precise and reliable angular velocity measurement is critical to overall performance.A comprehensive mathematical model is introduced to encapsulate all key MEMS parameters,accompanied by discussions on calibration and Allan variance interpretation.For six leading MEMS gyroscope applications,namely inertial navigation,integrated navigation,autopilot systems,rotating projectiles,homing guidance,and north finding,the most critical parameters are identified,distinguishing suitable and unsuitable sensor choices.Special emphasis is placed on inertial navigation systems,where practical rules of thumb for error evaluation are derived using six degrees of freedom motion equations.Rigorous simulations demonstrate the influence of various sensor parameters through real-world case studies,including static navigation,multi-rotor attitude estimation,gimbal stabilization,and north finding via a turntable.This work aims to be a beacon for practitioners across diverse fields,empowering them to make more informed design decisions.展开更多
Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challeng...Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challenging.This study presents the Wavelet-Guided Transformer U-Net(WGT-UNet)model,a new hybrid net-work that combines Convolutional Neural Networks(CNNs),Discrete Wavelet Transform(DWT),and Transformer architectures.The model’s primary contribution is based on spatial and channel attention mechanisms derived from wavelet subbands to guide the Transformer’s self-attention structure.Thus,low and high frequency components are separated at each stage using DWT,suppressing structural noise and making linear objects more prominent.The developed design is supported by multi-component hybrid cost functions that simultaneously solve class imbalance,edge sharpness,structural integrity,and spatial regularity issues.Furthermore,high segmentation success has been achieved in producing sharp boundaries and continuous line structures with the DWT-guided attention mechanism.Experiments conducted on the TTPLA dataset reveal that the version using the ConvNeXt backbone outperforms the current state-of-the-art approaches with an F1-Score of 79.33%and an Intersection over Union(IoU)value of 68.38%.The models and visual outputs of the developed method and all compared models can be accessed at https://github.com/burhanbarakli/WGT-UNET.展开更多
We present a computer-modeling framework for photovoltaic(PV)source emulation that preserves the exact single-diode physics while enabling iteration-free,real-time evaluation.We derive two closed-form explicit solvers...We present a computer-modeling framework for photovoltaic(PV)source emulation that preserves the exact single-diode physics while enabling iteration-free,real-time evaluation.We derive two closed-form explicit solvers based on the Lambert W function:a voltage-driven V-Lambert solver for high-fidelity I–V computation and a resistance-driven R-Lambert solver designed for seamless integration in a closed-loop PV emulator.Unlike Taylor-linearized explicit models,our proposed formulation retains the exponential nonlinearity of the PV equations.It employs a numerically stable analytical evaluation that eliminates the need for lookup tables and root-finding,all while maintaining limited computational costs and a small memory footprint.The R-Lambert model is integrated into a buck-converter emulator equipped with a discrete PI regulator,which generates current references directly from sensed operating points,thus supporting hardware-constrained implementation.Comprehensive numerical experiments conducted on six commercial modules from various technologies(mono,poly,and multicrystalline)demonstrate significant accuracy improvements under the IEC EN 50530 near-MPP criterion:the V-Lambert solver reduces the±10%Vmpp band error by up to 61 times compared to an explicit-model baseline.Dynamic simulations under varying irradiance,temperature,and load conditions achieve millisecond-scale settling with accurate trajectory tracking.Additionally,processor-in-the-loop experimental validation on an embedded microcontroller supports the simulation results.By unifying exact analytical modeling with embedded realization,this work advances computer modeling for PV emulation,MPPT benchmarking,and controller verification in integrated renewable energy systems.展开更多
Rapid evolutions of the Internet of Electric Vehicles(IoEVs)are reshaping and modernizing transport systems,yet challenges remain in energy efficiency,better battery aging,and grid stability.Typical charging methods a...Rapid evolutions of the Internet of Electric Vehicles(IoEVs)are reshaping and modernizing transport systems,yet challenges remain in energy efficiency,better battery aging,and grid stability.Typical charging methods allow for EVs to be charged without thought being given to the condition of the battery or the grid demand,thus increasing energy costs and battery aging.This study proposes a smart charging station with an AI-powered Battery Management System(BMS),developed and simulated in MATLAB/Simulink,to increase optimality in energy flow,battery health,and impractical scheduling within the IoEV environment.The system operates through real-time communication,load scheduling based on priorities,and adaptive charging based on batterymathematically computed State of Charge(SOC),State of Health(SOH),and thermal state,with bidirectional power flow(V2G),thus allowing EVs’participation towards grid stabilization.Simulation results revealed that the proposed model can reduce peak grid load by 37.8%;charging efficiency is enhanced by 92.6%;battery temperature lessened by 4.4℃;SOH extended over 100 cycles by 6.5%,if compared against the conventional technique.By this way,charging time was decreased by 12.4% and energy costs dropped by more than 20%.These results showed that smart charging with intelligent BMS can boost greatly the operational efficiency and sustainability of the IoEV ecosystem.展开更多
With the recent increase in data volume and diversity,traditional text representation techniques are struggling to capture context,particularly in environments with sparse data.To address these challenges,this study p...With the recent increase in data volume and diversity,traditional text representation techniques are struggling to capture context,particularly in environments with sparse data.To address these challenges,this study proposes a new model,the Masked Joint Representation Model(MJRM).MJRM approximates the original hypothesis by leveraging multiple elements in a limited context.It dynamically adapts to changes in characteristics based on data distribution through three main components.First,masking-based representation learning,termed selective dynamic masking,integrates topic modeling and sentiment clustering to generate and train multiple instances across different data subsets,whose predictions are then aggregated with optimized weights.This design alleviates sparsity,suppresses noise,and preserves contextual structures.Second,regularization-based improvements are applied.Third,techniques for addressing sparse data are used to perform final inference.As a result,MJRM improves performance by up to 4%compared to existing AI techniques.In our experiments,we analyzed the contribution of each factor,demonstrating that masking,dynamic learning,and aggregating multiple instances complement each other to improve performance.This demonstrates that a masking-based multi-learning strategy is effective for context-aware sparse text classification,and can be useful even in challenging situations such as data shortage or data distribution variations.We expect that the approach can be extended to diverse fields such as sentiment analysis,spam filtering,and domain-specific document classification.展开更多
Grid-Forming(GFM)converters are prone to fault-induced overcurrent and power angle instability during grid fault-induced voltage sags.To address this,this paper develops a multi-loop coordinated fault ridethrough(FRT)...Grid-Forming(GFM)converters are prone to fault-induced overcurrent and power angle instability during grid fault-induced voltage sags.To address this,this paper develops a multi-loop coordinated fault ridethrough(FRT)control strategy based on a power outer loop and voltage-current inner loops,aiming to enhance the stability and current-limiting capability of GFM converters during grid fault conditions.During voltage sags,the GFM converter’s voltage source behavior is maintained by dynamically adjusting the reactive power reference to provide voltage support,thereby effectively suppressing the steady-state component of the fault current.To address the active power imbalance induced by voltage sags,a dynamic active power reference correction method based on apparent power is designed to mitigate power angle oscillations and limit transient current.Moreover,an adaptive virtual impedance loop is implemented to enhance dynamic transient current-limiting performance during the fault initiation phase.This approach improves the responsiveness of the inner loop and ensures safe system operation under various fault severities.Under asymmetric fault conditions,a negative-sequence reactive current compensation strategy is incorporated to further suppress negative-sequence voltage and improve voltage symmetry.The proposed control scheme enables coordinated operation of multiple control objectives,including voltage support,current suppression,and power angle stability,across different fault scenarios.Finally,MATLAB/Simulink simulation results validate the effectiveness of the proposed strategy,showcasing its superior performance in current limiting and power angle stability,thereby significantly enhancing the system’s fault ride-through capability.展开更多
Decarbonising the building sector,particularly residential heating,represents a critical challenge for achieving carbon-neutral energy systems.Efficient solutions must integrate both technological performance and rene...Decarbonising the building sector,particularly residential heating,represents a critical challenge for achieving carbon-neutral energy systems.Efficient solutions must integrate both technological performance and renewable energy sources while considering operational constraints of existing systems.This study investigates a hybrid heating system combining a natural gas boiler(NGB)with an air-to-water heat pump(AWHP),evaluated through a combination of laboratory experiments and dynamic modelling.A prototype developed in the Electrical and Energy Engineering Laboratory enabled the characterization of both heat generators,the collection of experimental data,and the calibration of a MATLAB/Simulink model,including emissions and exhaust analyses.Sensitivity analyses were performed to identify optimal configurations for energy efficiency and system control,accounting for interactions between subsystems.Results highlight that hybridisation significantly improves primary energy efficiency and reduces fuel consumption compared to conventional NGB-only systems.Environmental performance,assessed through CO_(2) and NOx emissions and renewable energy integration,demonstrates the benefits of partial electrification in the residential sector.Economic assessment further quantifies decarbonization costs and fuel savings,illustrating tradeoffs between low-capital,moderate-performance systems and high-efficiency,high-renewable solutions requiring larger investments.The analysis shows that strategic decisions for residential decarbonisation cannot be separated from system-wide considerations,including control strategies,component integration,and economic feasibility.The study underlines the importance of hybrid and renewable-based solutions as pivotal pathways for energy transition in the residential building sector.展开更多
In order to address environmental pollution and resource depletion caused by traditional power generation,this paper proposes an adaptive iterative dynamic-balance optimization algorithm that integrates the Improved D...In order to address environmental pollution and resource depletion caused by traditional power generation,this paper proposes an adaptive iterative dynamic-balance optimization algorithm that integrates the Improved Dung Beetle Optimizer(IDBO)with VariationalMode Decomposition(VMD).The IDBO-VMD method is designed to enhance the accuracy and efficiency of wind-speed time-series decomposition and to effectively smooth photovoltaic power fluctuations.This study innovatively improves the traditional variational mode decomposition(VMD)algorithm,and significantly improves the accuracy and adaptive ability of signal decomposition by IDBO selfoptimization of key parameters K and a.On this basis,Fourier transform technology is used to define the boundary point between high frequency and low frequency signals,and a targeted energy distribution strategy is proposed:high frequency fluctuations are allocated to supercapacitors to quickly respond to transient power fluctuations;Lowfrequency components are distributed to lead-carbon batteries,optimizing long-term energy storage and scheduling efficiency.This strategy effectively improves the response speed and stability of the energy storage system.The experimental results demonstrate that the IDBO-VMD algorithm markedly outperforms traditional methods in both decomposition accuracy and computational efficiency.Specifically,it effectively reduces the charge–discharge frequency of the battery,prolongs battery life,and optimizes the operating ranges of the state-of-charge(SOC)for both leadcarbon batteries and supercapacitors.In addition,the energy management strategy based on the algorithm not only improves the overall energy utilization efficiency of the system,but also shows excellent performance in the dynamic management and intelligent scheduling of renewable energy generation.展开更多
In the quest to enhance energy efficiency and reduce environmental impact in the transportation sector,the recovery of waste heat from diesel engines has become a critical area of focus.This study provided an exhausti...In the quest to enhance energy efficiency and reduce environmental impact in the transportation sector,the recovery of waste heat from diesel engines has become a critical area of focus.This study provided an exhaustive thermodynamic analysis optimizing Organic Rankine Cycle(ORC)systems forwaste heat recovery fromdiesel engines.Thestudy assessed the performance of five candidateworking fluids—R11,R123,R113,R245fa,and R141b—under a range of operating conditions,specifically varying overheat temperatures and evaporation pressures.The results indicated that the choice of working fluid substantially influences the system’s exergetic efficiency,net output power,and thermal efficiency.R245fa showed an outstanding net output power of 30.39 kW at high overheat conditions,outperforming R11,which is significant for high-temperature waste heat recovery.At lower temperatures,R11 and R113 demonstrated higher exergetic efficiencies,with R11 reaching a peak exergetic efficiency of 7.4%at an evaporation pressure of 10 bar and an overheat of 10℃.The study also revealed that controlling the overheat and optimizing the evaporation pressure are crucial for enhancing the net output power of the ORC system.Specifically,at an evaporation pressure of 30 bar and an overheat of 0℃,R113 exhibited the lowest exergetic destruction of 544.5 kJ/kg,making it a suitable choice for minimizing irreversible losses.These findings are instrumental for understanding the performance of ORC systems in waste heat recovery applications and offer valuable insights for the design and operation of more efficient and environmentally friendly diesel engine systems.展开更多
Automated detection of Motor Imagery(MI)tasks is extremely useful for prosthetic arms and legs of stroke patients for their rehabilitation.Prediction of MI tasks can be performed with the help of Electroencephalogram(...Automated detection of Motor Imagery(MI)tasks is extremely useful for prosthetic arms and legs of stroke patients for their rehabilitation.Prediction of MI tasks can be performed with the help of Electroencephalogram(EEG)signals recorded by placing electrodes on the scalp of subjects;however,accurate prediction of MI tasks remains a challenge due to noise that is incurred during the EEG signal recording process,the extraction of a feature vector with high interclass variance,and accurate classification.The proposed method consists of preprocessing,feature extraction,and classification.First,EEG signals are denoised using a bandpass filter followed by Independent Component Analysis(ICA).Multiple channels are combined to form a single surrogate channel.Short Time Fourier Transform(STFT)is then applied to convert time domain EEG signals into the frequency domain.Handcrafted and automated features are extracted from EEG signals and then concatenated to form a single feature vector.We propose a customized two-dimensional Convolutional Neural Network(CNN)for automated feature extraction with high interclass variance.Feature selection is performed using Particle Swarm Optimization(PSO)to obtain optimal features.The final feature vector is passed to three different classifiers:Support Vector Machine(SVM),Random Forest(RF),and Long Short-Term Memory(LSTM).The final decision is made using the Model-Agnostic Meta Learning(MAML).The Proposed method has been tested on two datasets,including PhysioNet and BCI Competition IV-2a,and it achieved better results in terms of accuracy and F1 score than existing state-of-the-art methods.The proposed framework achieved an accuracy and F1 score of 96%on the PhysioNet dataset and 95.5%on the BCI Competition IV,dataset 2a.We also present SHapley Additive exPlanations(SHAP)and Gradient-weighted Class Activation Mapping(Grad-CAM)explainable techniques to enhance model interpretability in a clinical setting.展开更多
Modern power systems increasingly depend on interconnected microgrids to enhance reliability and renewable energy utilization.However,the high penetration of intermittent renewable sources often causes frequency devia...Modern power systems increasingly depend on interconnected microgrids to enhance reliability and renewable energy utilization.However,the high penetration of intermittent renewable sources often causes frequency deviations,voltage fluctuations,and poor reactive power coordination,posing serious challenges to grid stability.Conventional Interconnection FlowControllers(IFCs)primarily regulate active power flowand fail to effectively handle dynamic frequency variations or reactive power sharing in multi-microgrid networks.To overcome these limitations,this study proposes an enhanced Interconnection Flow Controller(e-IFC)that integrates frequency response balancing and an Interconnection Reactive Power Flow Controller(IRFC)within a unified adaptive control structure.The proposed e-IFC is implemented and analyzed in DIgSILENT PowerFactory to evaluate its performance under various grid disturbances,including frequency drops,load changes,and reactive power fluctuations.Simulation results reveal that the e-IFC achieves 27.4% higher active power sharing accuracy,19.6% lower reactive power deviation,and 18.2% improved frequency stability compared to the conventional IFC.The adaptive controller ensures seamless transitions between grid-connected and islanded modes and maintains stable operation even under communication delays and data noise.Overall,the proposed e-IFCsignificantly enhances active-reactive power coordination and dynamic stability in renewable-integrated multi-microgrid systems.Future research will focus on coupling the e-IFC with tertiary-level optimization frameworks and conducting hardware-in-the-loop validation to enable its application in large-scale smart microgrid environments.展开更多
This paper presents an intelligent patrol and security robot integrating 2D LiDAR and RGB-D vision sensors to achieve semantic simultaneous localization and mapping(SLAM),real-time object recognition,and dynamic obsta...This paper presents an intelligent patrol and security robot integrating 2D LiDAR and RGB-D vision sensors to achieve semantic simultaneous localization and mapping(SLAM),real-time object recognition,and dynamic obstacle avoidance.The system employs the YOLOv7 deep-learning framework for semantic detection and SLAM for localization and mapping,fusing geometric and visual data to build a high-fidelity 2D semantic map.This map enables the robot to identify and project object information for improved situational awareness.Experimental results show that object recognition reached 95.4%mAP@0.5.Semantic completeness increased from 68.7%(single view)to 94.1%(multi-view)with an average position error of 3.1 cm.During navigation,the robot achieved 98.0%reliability,avoided moving obstacles in 90.0%of encounters,and replanned paths in 0.42 s on average.The integration of LiDAR-based SLAMwith deep-learning–driven semantic perception establishes a robust foundation for intelligent,adaptive,and safe robotic navigation in dynamic environments.展开更多
Federated Learning(FL)has become a leading decentralized solution that enables multiple clients to train a model in a collaborative environment without directly sharing raw data,making it suitable for privacy-sensitiv...Federated Learning(FL)has become a leading decentralized solution that enables multiple clients to train a model in a collaborative environment without directly sharing raw data,making it suitable for privacy-sensitive applications such as healthcare,finance,and smart systems.As the field continues to evolve,the research field has become more complex and scattered,covering different system designs,training methods,and privacy techniques.This survey is organized around the three core challenges:how the data is distributed,how models are synchronized,and how to defend against attacks.It provides a structured and up-to-date review of FL research from 2023 to 2025,offering a unified taxonomy that categorizes works by data distribution(Horizontal FL,Vertical FL,Federated Transfer Learning,and Personalized FL),training synchronization(synchronous and asynchronous FL),optimization strategies,and threat models(data leakage and poisoning attacks).In particular,we summarize the latest contributions in Vertical FL frameworks for secure multi-party learning,communication-efficient Horizontal FL,and domain-adaptive Federated Transfer Learning.Furthermore,we examine synchronization techniques addressing system heterogeneity,including straggler mitigation in synchronous FL and staleness management in asynchronous FL.The survey covers security threats in FL,such as gradient inversion,membership inference,and poisoning attacks,as well as their defense strategies that include privacy-preserving aggregation and anomaly detection.The paper concludes by outlining unresolved issues and highlighting challenges in handling personalized models,scalability,and real-world adoption.展开更多
In the competitive retail industry of the digital era,data-driven insights into gender-specific customer behavior are essential.They support the optimization of store performance,layout design,product placement,and ta...In the competitive retail industry of the digital era,data-driven insights into gender-specific customer behavior are essential.They support the optimization of store performance,layout design,product placement,and targeted marketing.However,existing computer vision solutions often rely on facial recognition to gather such insights,raising significant privacy and ethical concerns.To address these issues,this paper presents a privacypreserving customer analytics system through two key strategies.First,we deploy a deep learning framework using YOLOv9s,trained on the RCA-TVGender dataset.Cameras are positioned perpendicular to observation areas to reduce facial visibility while maintaining accurate gender classification.Second,we apply AES-128 encryption to customer position data,ensuring secure access and regulatory compliance.Our system achieved overall performance,with 81.5%mAP@50,77.7%precision,and 75.7%recall.Moreover,a 90-min observational study confirmed the system’s ability to generate privacy-protected heatmaps revealing distinct behavioral patterns between male and female customers.For instance,women spent more time in certain areas and showed interest in different products.These results confirm the system’s effectiveness in enabling personalized layout and marketing strategies without compromising privacy.展开更多
In recent years,three-dimensional reconstruction technologies that employ multiple cameras have continued to evolve significantly,enabling remote collaboration among users in extended Reality(XR)environments.In additi...In recent years,three-dimensional reconstruction technologies that employ multiple cameras have continued to evolve significantly,enabling remote collaboration among users in extended Reality(XR)environments.In addition,methods for deploying multiple cameras for motion capture of users(e.g.,performers)are widely used in computer graphics.As the need to minimize and optimize the number of cameras grows to reduce costs,various technologies and research approaches focused on Optimal Camera Placement(OCP)are continually being proposed.However,as most existing studies assume homogeneous camera setups,there is a growing demand for studies on heterogeneous camera setups.For instance,technical demands keep emerging in scenarios with minimal camera configurations,especially regarding cost factors,the physical placement of cameras given the spatial structure,and image capture strategies for heterogeneous cameras,such as high-resolution RGB cameras and depth cameras.In this study,we propose a pre-visualization and simulation method for the optimal placement of heterogeneous cameras in XR environments,accounting for both the specifications of heterogeneous cameras(e.g.,field of view)and the physical configuration(e.g.,wall configuration)in real-world spaces.The proposed method performs a visibility analysis of cameras by considering each camera’s field-of-view volume,resolution,and unique characteristics,along with physicalspace constraints.This approach enables the optimal position and rotation of each camera to be recommended,along with the minimum number of cameras required.In the results of our study conducted in heterogeneous camera combinations,the proposed method achieved 81.7%~82.7%coverage of the target visual information using only 2~3 cameras.In contrast,single(or homogeneous)-typed cameras were required to use 11 cameras for 81.6%coverage.Accordingly,we found that camera deployment resources can be reduced with the proposed approaches.展开更多
文摘This review paper examines the various types of electrical generators used to convert wave energy into electrical energy.The focus is on both linear and rotary generators,including their design principles,operational efficiencies,and technological advancements.Linear generators,such as Induction,permanent magnet synchronous,and switched reluctance types,are highlighted for their direct conversion capability,eliminating the need for mechanical gearboxes.Rotary Induction generators,permanent magnet synchronous generators,and doubly-fed Induction generators are evaluated for their established engineering principles and integration with existing grid infrastructure.The paper discusses the historical development,environmental benefits,and ongoing advancements in wave energy technologies,emphasizing the increasing feasibility and scalability of wave energy as a renewable source.Through a comprehensive analysis,this review provides insights into the current state and future prospects of electrical generators in wave energy conversion,underscoring their potential to significantly reduce reliance on fossil fuels and mitigate environmental impacts.
基金supported by the Central Guided Local Science and Technology Development Project(Grant YDZX2022001).
文摘Converter transformers are the core components of ultra-high voltage(UHV)transmission systems.The main cause of faults in converter transformers is irreversible deterioration of oil-pressboard insulation under combined electrical-thermal-mechanical stress over long operating times.In this paper,the chemical characteristics of oil-pressboard insulation samples subjected to electrical-thermal-mechanical ageing for different times are studied.An image processing algorithm is used to analyse the discharge propagation characteristics of the samples under combined alternating current(AC)-direct current(DC)voltage,and the current pulse curves and phase resolved partial discharge spectrogram corresponding to the discharge images are analysed.An improved wavelet packet algorithm is used to denoise the discharge current pulse.Finally,the influence of electrical-thermal-mechanical ageing on discharge characteristics is analysed using radar charts.The condition of oil-pressboard insulation is one of the main factors determining the life expectancy of converter transformers.The results obtained here therefore have practical significance for understanding the process of insulation failure caused by accelerated ageing of oil-pressboard insulation.
基金supported by National Natural Science Foundation of China(Grant Nos:62122004 and 62274082)Beijing Natural Science Foundation(Grant No.Z210006)+5 种基金Hong Kong Research Grant Council(Grant Nos.27206321,17205922,17212923,C1009-22G and T45-701/22-R)Shenzhen Science and Technology Innovation Commission(SGDX20220530111405040,JCYJ20220530115411025 and JCYJ20210324120409025)Research on mechanism of source/drain ohmic contact and the related Ga N p-FET(Grant No:2023A1515030034)Research on high-reliable Ga N power device and the related industrial power system(Grant No:HZQB-KCZYZ-2021052)supported by ACCESS-AI Chip Center for Emerging Smart Systems,sponsored by Innovation and Technology Fund(ITF),Hong Kong SARthe assistance of SUSTech Core Research Facilities。
文摘Reconfigurable surface acoustic wave(SAW)phase shifters have garnered significant attention owing to their potential applications in emerging fields such as secure wireless communication,adaptable signal processing,and intelligent sensing systems.Among various modulation methods,employing gate voltage-controlled tuning methodologies that leverage acoustoelectric interactions has proven to be an efficient modulation approach that requires a low bias voltage.However,current acoustoelectric devices suffer from limited tunability,intricate heterogeneous structures,and complex manufacturing processes,all of which impede their practical applications.In this study,we present a novel material system for voltage-tunable SAW phase shifters.This system incorporates an atomic layer deposition ZnO thin-film transistors on LiNbO_(3)structure.This structure combines the benefits of LiNbO_(3)'s high electromechanical coupling coefficient(K^(2))and ZnO's superior conductivity adjustability.Besides,the device possesses a simplified structural configuration,which is easy to fabricate.Devices with different mesa lengths were fabricated and measured,and two of the different modes were compared.The results indicate that both the maximum phase shift and attenuation of the Rayleigh mode and longitudinal leaky SAW(LLSAW)increase proportionally with mesa length.Furthermore,LLSAW with larger effective electromechanical coupling coefficients(K_(eff)^(2))values exhibits greater phase velocity shifts and attenuation coefficients,with a maximum phase velocity tuning of 1.22%achieved.It is anticipated that the proposed devices will find utility in a variety of applications necessitating tunable acoustic components.
基金National Key Research and Development Program of China(2021YFA1501302)the National Natural Science Foundation of China(22121004,22122808)+1 种基金the Haihe Laboratory of Sustainable Chemical Transformations and the Program of Introducing Talents of Discipline to Universities(BP0618007)for financial supportsupported by the XPLORER PRIZE.
文摘Noise is inevitable in electrical capacitance tomography(ECT)measurements.This paper describes the influence of noise on ECT performance for measuring gas-solids fluidized bed characteristics.The noise distribution is approximated by the Gaussian distribution and added to experimental capacitance data with various intensities.The equivalent signal strength(Ф)that equals the signal-to-noise ratio of packed beds is used to evaluate noise levels.Results show that the Pearson correlation coefficient,which indicates the similarity of solids fraction distributions over pixels,increases with Ф,and reconstructed images are more deteriorated at lower Ф.Nevertheless,relative errors for average solids fraction and bubble size in each frame are less sensitive to noise,attributed to noise compromise caused by the process of pixel values.These findings provide useful guidance for assessing the accuracy of ECT measurements of multiphase flows.
文摘Optimizing routing and resource allocation in decentralized unmanned aerial vehicle(UAV)networks remains challenging due to interference and rapidly changing topologies.The authors introduce a novel framework combining double deep Q-networks(DDQNs)and graph neural networks(GNNs)for joint routing and resource allocation.The framework uses GNNs to model the network topology and DDQNs to adaptively control routing and resource allocation,addressing interference and improving network performance.Simulation results show that the proposed approach outperforms traditional methods such as Closest-to-Destination(c2Dst),Max-SINR(mSINR),and Multi-Layer Perceptron(MLP)-based models,achieving approximately 23.5% improvement in throughput,50% increase in connection probability,and 17.6% reduction in number of hops,demonstrating its effectiveness in dynamic UAV networks.
文摘Microelectromechanical systems(MEMS)technology has gained significant attention over the past decade for measuring inertial angular velocity.However,due to inherent complexity,MEMS gyroscopes typically feature up to ten times more parameters than traditional sensors,making selection a challenging task even for experts.This study addresses this challenge,focusing on defensive guidance,navigation,and control(GNC)systems where precise and reliable angular velocity measurement is critical to overall performance.A comprehensive mathematical model is introduced to encapsulate all key MEMS parameters,accompanied by discussions on calibration and Allan variance interpretation.For six leading MEMS gyroscope applications,namely inertial navigation,integrated navigation,autopilot systems,rotating projectiles,homing guidance,and north finding,the most critical parameters are identified,distinguishing suitable and unsuitable sensor choices.Special emphasis is placed on inertial navigation systems,where practical rules of thumb for error evaluation are derived using six degrees of freedom motion equations.Rigorous simulations demonstrate the influence of various sensor parameters through real-world case studies,including static navigation,multi-rotor attitude estimation,gimbal stabilization,and north finding via a turntable.This work aims to be a beacon for practitioners across diverse fields,empowering them to make more informed design decisions.
文摘Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challenging.This study presents the Wavelet-Guided Transformer U-Net(WGT-UNet)model,a new hybrid net-work that combines Convolutional Neural Networks(CNNs),Discrete Wavelet Transform(DWT),and Transformer architectures.The model’s primary contribution is based on spatial and channel attention mechanisms derived from wavelet subbands to guide the Transformer’s self-attention structure.Thus,low and high frequency components are separated at each stage using DWT,suppressing structural noise and making linear objects more prominent.The developed design is supported by multi-component hybrid cost functions that simultaneously solve class imbalance,edge sharpness,structural integrity,and spatial regularity issues.Furthermore,high segmentation success has been achieved in producing sharp boundaries and continuous line structures with the DWT-guided attention mechanism.Experiments conducted on the TTPLA dataset reveal that the version using the ConvNeXt backbone outperforms the current state-of-the-art approaches with an F1-Score of 79.33%and an Intersection over Union(IoU)value of 68.38%.The models and visual outputs of the developed method and all compared models can be accessed at https://github.com/burhanbarakli/WGT-UNET.
基金funded by Scientific Research Deanship at University of Ha’il-Saudi Arabia through project number(RG-24014).
文摘We present a computer-modeling framework for photovoltaic(PV)source emulation that preserves the exact single-diode physics while enabling iteration-free,real-time evaluation.We derive two closed-form explicit solvers based on the Lambert W function:a voltage-driven V-Lambert solver for high-fidelity I–V computation and a resistance-driven R-Lambert solver designed for seamless integration in a closed-loop PV emulator.Unlike Taylor-linearized explicit models,our proposed formulation retains the exponential nonlinearity of the PV equations.It employs a numerically stable analytical evaluation that eliminates the need for lookup tables and root-finding,all while maintaining limited computational costs and a small memory footprint.The R-Lambert model is integrated into a buck-converter emulator equipped with a discrete PI regulator,which generates current references directly from sensed operating points,thus supporting hardware-constrained implementation.Comprehensive numerical experiments conducted on six commercial modules from various technologies(mono,poly,and multicrystalline)demonstrate significant accuracy improvements under the IEC EN 50530 near-MPP criterion:the V-Lambert solver reduces the±10%Vmpp band error by up to 61 times compared to an explicit-model baseline.Dynamic simulations under varying irradiance,temperature,and load conditions achieve millisecond-scale settling with accurate trajectory tracking.Additionally,processor-in-the-loop experimental validation on an embedded microcontroller supports the simulation results.By unifying exact analytical modeling with embedded realization,this work advances computer modeling for PV emulation,MPPT benchmarking,and controller verification in integrated renewable energy systems.
文摘Rapid evolutions of the Internet of Electric Vehicles(IoEVs)are reshaping and modernizing transport systems,yet challenges remain in energy efficiency,better battery aging,and grid stability.Typical charging methods allow for EVs to be charged without thought being given to the condition of the battery or the grid demand,thus increasing energy costs and battery aging.This study proposes a smart charging station with an AI-powered Battery Management System(BMS),developed and simulated in MATLAB/Simulink,to increase optimality in energy flow,battery health,and impractical scheduling within the IoEV environment.The system operates through real-time communication,load scheduling based on priorities,and adaptive charging based on batterymathematically computed State of Charge(SOC),State of Health(SOH),and thermal state,with bidirectional power flow(V2G),thus allowing EVs’participation towards grid stabilization.Simulation results revealed that the proposed model can reduce peak grid load by 37.8%;charging efficiency is enhanced by 92.6%;battery temperature lessened by 4.4℃;SOH extended over 100 cycles by 6.5%,if compared against the conventional technique.By this way,charging time was decreased by 12.4% and energy costs dropped by more than 20%.These results showed that smart charging with intelligent BMS can boost greatly the operational efficiency and sustainability of the IoEV ecosystem.
基金supported by the SungKyunKwan University and the BK21 FOUR(Graduate School Innovation)funded by the Ministry of Education(MOE,Korea)and National Research Foundation of Korea(NRF).
文摘With the recent increase in data volume and diversity,traditional text representation techniques are struggling to capture context,particularly in environments with sparse data.To address these challenges,this study proposes a new model,the Masked Joint Representation Model(MJRM).MJRM approximates the original hypothesis by leveraging multiple elements in a limited context.It dynamically adapts to changes in characteristics based on data distribution through three main components.First,masking-based representation learning,termed selective dynamic masking,integrates topic modeling and sentiment clustering to generate and train multiple instances across different data subsets,whose predictions are then aggregated with optimized weights.This design alleviates sparsity,suppresses noise,and preserves contextual structures.Second,regularization-based improvements are applied.Third,techniques for addressing sparse data are used to perform final inference.As a result,MJRM improves performance by up to 4%compared to existing AI techniques.In our experiments,we analyzed the contribution of each factor,demonstrating that masking,dynamic learning,and aggregating multiple instances complement each other to improve performance.This demonstrates that a masking-based multi-learning strategy is effective for context-aware sparse text classification,and can be useful even in challenging situations such as data shortage or data distribution variations.We expect that the approach can be extended to diverse fields such as sentiment analysis,spam filtering,and domain-specific document classification.
文摘Grid-Forming(GFM)converters are prone to fault-induced overcurrent and power angle instability during grid fault-induced voltage sags.To address this,this paper develops a multi-loop coordinated fault ridethrough(FRT)control strategy based on a power outer loop and voltage-current inner loops,aiming to enhance the stability and current-limiting capability of GFM converters during grid fault conditions.During voltage sags,the GFM converter’s voltage source behavior is maintained by dynamically adjusting the reactive power reference to provide voltage support,thereby effectively suppressing the steady-state component of the fault current.To address the active power imbalance induced by voltage sags,a dynamic active power reference correction method based on apparent power is designed to mitigate power angle oscillations and limit transient current.Moreover,an adaptive virtual impedance loop is implemented to enhance dynamic transient current-limiting performance during the fault initiation phase.This approach improves the responsiveness of the inner loop and ensures safe system operation under various fault severities.Under asymmetric fault conditions,a negative-sequence reactive current compensation strategy is incorporated to further suppress negative-sequence voltage and improve voltage symmetry.The proposed control scheme enables coordinated operation of multiple control objectives,including voltage support,current suppression,and power angle stability,across different fault scenarios.Finally,MATLAB/Simulink simulation results validate the effectiveness of the proposed strategy,showcasing its superior performance in current limiting and power angle stability,thereby significantly enhancing the system’s fault ride-through capability.
基金supported by European Commission and is a part of the HORIZON2020 project RES Heatfunding from the European Union’s Horizon 2020 program in the field of research and innovation on the basis of grant agreement No.956255.
文摘Decarbonising the building sector,particularly residential heating,represents a critical challenge for achieving carbon-neutral energy systems.Efficient solutions must integrate both technological performance and renewable energy sources while considering operational constraints of existing systems.This study investigates a hybrid heating system combining a natural gas boiler(NGB)with an air-to-water heat pump(AWHP),evaluated through a combination of laboratory experiments and dynamic modelling.A prototype developed in the Electrical and Energy Engineering Laboratory enabled the characterization of both heat generators,the collection of experimental data,and the calibration of a MATLAB/Simulink model,including emissions and exhaust analyses.Sensitivity analyses were performed to identify optimal configurations for energy efficiency and system control,accounting for interactions between subsystems.Results highlight that hybridisation significantly improves primary energy efficiency and reduces fuel consumption compared to conventional NGB-only systems.Environmental performance,assessed through CO_(2) and NOx emissions and renewable energy integration,demonstrates the benefits of partial electrification in the residential sector.Economic assessment further quantifies decarbonization costs and fuel savings,illustrating tradeoffs between low-capital,moderate-performance systems and high-efficiency,high-renewable solutions requiring larger investments.The analysis shows that strategic decisions for residential decarbonisation cannot be separated from system-wide considerations,including control strategies,component integration,and economic feasibility.The study underlines the importance of hybrid and renewable-based solutions as pivotal pathways for energy transition in the residential building sector.
基金funded by the Institute of Smart Energy,Huaiyin Institute of Technology,under Grant No.HIT-ISE-2024-07.
文摘In order to address environmental pollution and resource depletion caused by traditional power generation,this paper proposes an adaptive iterative dynamic-balance optimization algorithm that integrates the Improved Dung Beetle Optimizer(IDBO)with VariationalMode Decomposition(VMD).The IDBO-VMD method is designed to enhance the accuracy and efficiency of wind-speed time-series decomposition and to effectively smooth photovoltaic power fluctuations.This study innovatively improves the traditional variational mode decomposition(VMD)algorithm,and significantly improves the accuracy and adaptive ability of signal decomposition by IDBO selfoptimization of key parameters K and a.On this basis,Fourier transform technology is used to define the boundary point between high frequency and low frequency signals,and a targeted energy distribution strategy is proposed:high frequency fluctuations are allocated to supercapacitors to quickly respond to transient power fluctuations;Lowfrequency components are distributed to lead-carbon batteries,optimizing long-term energy storage and scheduling efficiency.This strategy effectively improves the response speed and stability of the energy storage system.The experimental results demonstrate that the IDBO-VMD algorithm markedly outperforms traditional methods in both decomposition accuracy and computational efficiency.Specifically,it effectively reduces the charge–discharge frequency of the battery,prolongs battery life,and optimizes the operating ranges of the state-of-charge(SOC)for both leadcarbon batteries and supercapacitors.In addition,the energy management strategy based on the algorithm not only improves the overall energy utilization efficiency of the system,but also shows excellent performance in the dynamic management and intelligent scheduling of renewable energy generation.
基金funded by the Huaiyin Institute of Technology—Institute of Smart Energy.
文摘In the quest to enhance energy efficiency and reduce environmental impact in the transportation sector,the recovery of waste heat from diesel engines has become a critical area of focus.This study provided an exhaustive thermodynamic analysis optimizing Organic Rankine Cycle(ORC)systems forwaste heat recovery fromdiesel engines.Thestudy assessed the performance of five candidateworking fluids—R11,R123,R113,R245fa,and R141b—under a range of operating conditions,specifically varying overheat temperatures and evaporation pressures.The results indicated that the choice of working fluid substantially influences the system’s exergetic efficiency,net output power,and thermal efficiency.R245fa showed an outstanding net output power of 30.39 kW at high overheat conditions,outperforming R11,which is significant for high-temperature waste heat recovery.At lower temperatures,R11 and R113 demonstrated higher exergetic efficiencies,with R11 reaching a peak exergetic efficiency of 7.4%at an evaporation pressure of 10 bar and an overheat of 10℃.The study also revealed that controlling the overheat and optimizing the evaporation pressure are crucial for enhancing the net output power of the ORC system.Specifically,at an evaporation pressure of 30 bar and an overheat of 0℃,R113 exhibited the lowest exergetic destruction of 544.5 kJ/kg,making it a suitable choice for minimizing irreversible losses.These findings are instrumental for understanding the performance of ORC systems in waste heat recovery applications and offer valuable insights for the design and operation of more efficient and environmentally friendly diesel engine systems.
基金supported and funded by the Deanship of Scientific Research at Imam Mohammad Ibn Saud Islamic University(IMSIU)(grant number IMSIU-DDRSP2601).
文摘Automated detection of Motor Imagery(MI)tasks is extremely useful for prosthetic arms and legs of stroke patients for their rehabilitation.Prediction of MI tasks can be performed with the help of Electroencephalogram(EEG)signals recorded by placing electrodes on the scalp of subjects;however,accurate prediction of MI tasks remains a challenge due to noise that is incurred during the EEG signal recording process,the extraction of a feature vector with high interclass variance,and accurate classification.The proposed method consists of preprocessing,feature extraction,and classification.First,EEG signals are denoised using a bandpass filter followed by Independent Component Analysis(ICA).Multiple channels are combined to form a single surrogate channel.Short Time Fourier Transform(STFT)is then applied to convert time domain EEG signals into the frequency domain.Handcrafted and automated features are extracted from EEG signals and then concatenated to form a single feature vector.We propose a customized two-dimensional Convolutional Neural Network(CNN)for automated feature extraction with high interclass variance.Feature selection is performed using Particle Swarm Optimization(PSO)to obtain optimal features.The final feature vector is passed to three different classifiers:Support Vector Machine(SVM),Random Forest(RF),and Long Short-Term Memory(LSTM).The final decision is made using the Model-Agnostic Meta Learning(MAML).The Proposed method has been tested on two datasets,including PhysioNet and BCI Competition IV-2a,and it achieved better results in terms of accuracy and F1 score than existing state-of-the-art methods.The proposed framework achieved an accuracy and F1 score of 96%on the PhysioNet dataset and 95.5%on the BCI Competition IV,dataset 2a.We also present SHapley Additive exPlanations(SHAP)and Gradient-weighted Class Activation Mapping(Grad-CAM)explainable techniques to enhance model interpretability in a clinical setting.
基金the Deanship of Scientific Research at Northern Border University,Arar,Saudi Arabia,for funding this research work through the project number“NBU-FFR-2025-3623-11”.
文摘Modern power systems increasingly depend on interconnected microgrids to enhance reliability and renewable energy utilization.However,the high penetration of intermittent renewable sources often causes frequency deviations,voltage fluctuations,and poor reactive power coordination,posing serious challenges to grid stability.Conventional Interconnection FlowControllers(IFCs)primarily regulate active power flowand fail to effectively handle dynamic frequency variations or reactive power sharing in multi-microgrid networks.To overcome these limitations,this study proposes an enhanced Interconnection Flow Controller(e-IFC)that integrates frequency response balancing and an Interconnection Reactive Power Flow Controller(IRFC)within a unified adaptive control structure.The proposed e-IFC is implemented and analyzed in DIgSILENT PowerFactory to evaluate its performance under various grid disturbances,including frequency drops,load changes,and reactive power fluctuations.Simulation results reveal that the e-IFC achieves 27.4% higher active power sharing accuracy,19.6% lower reactive power deviation,and 18.2% improved frequency stability compared to the conventional IFC.The adaptive controller ensures seamless transitions between grid-connected and islanded modes and maintains stable operation even under communication delays and data noise.Overall,the proposed e-IFCsignificantly enhances active-reactive power coordination and dynamic stability in renewable-integrated multi-microgrid systems.Future research will focus on coupling the e-IFC with tertiary-level optimization frameworks and conducting hardware-in-the-loop validation to enable its application in large-scale smart microgrid environments.
基金supported by the National Science and Technology Council of under Grant NSTC 114-2221-E-130-007.
文摘This paper presents an intelligent patrol and security robot integrating 2D LiDAR and RGB-D vision sensors to achieve semantic simultaneous localization and mapping(SLAM),real-time object recognition,and dynamic obstacle avoidance.The system employs the YOLOv7 deep-learning framework for semantic detection and SLAM for localization and mapping,fusing geometric and visual data to build a high-fidelity 2D semantic map.This map enables the robot to identify and project object information for improved situational awareness.Experimental results show that object recognition reached 95.4%mAP@0.5.Semantic completeness increased from 68.7%(single view)to 94.1%(multi-view)with an average position error of 3.1 cm.During navigation,the robot achieved 98.0%reliability,avoided moving obstacles in 90.0%of encounters,and replanned paths in 0.42 s on average.The integration of LiDAR-based SLAMwith deep-learning–driven semantic perception establishes a robust foundation for intelligent,adaptive,and safe robotic navigation in dynamic environments.
文摘Federated Learning(FL)has become a leading decentralized solution that enables multiple clients to train a model in a collaborative environment without directly sharing raw data,making it suitable for privacy-sensitive applications such as healthcare,finance,and smart systems.As the field continues to evolve,the research field has become more complex and scattered,covering different system designs,training methods,and privacy techniques.This survey is organized around the three core challenges:how the data is distributed,how models are synchronized,and how to defend against attacks.It provides a structured and up-to-date review of FL research from 2023 to 2025,offering a unified taxonomy that categorizes works by data distribution(Horizontal FL,Vertical FL,Federated Transfer Learning,and Personalized FL),training synchronization(synchronous and asynchronous FL),optimization strategies,and threat models(data leakage and poisoning attacks).In particular,we summarize the latest contributions in Vertical FL frameworks for secure multi-party learning,communication-efficient Horizontal FL,and domain-adaptive Federated Transfer Learning.Furthermore,we examine synchronization techniques addressing system heterogeneity,including straggler mitigation in synchronous FL and staleness management in asynchronous FL.The survey covers security threats in FL,such as gradient inversion,membership inference,and poisoning attacks,as well as their defense strategies that include privacy-preserving aggregation and anomaly detection.The paper concludes by outlining unresolved issues and highlighting challenges in handling personalized models,scalability,and real-world adoption.
文摘In the competitive retail industry of the digital era,data-driven insights into gender-specific customer behavior are essential.They support the optimization of store performance,layout design,product placement,and targeted marketing.However,existing computer vision solutions often rely on facial recognition to gather such insights,raising significant privacy and ethical concerns.To address these issues,this paper presents a privacypreserving customer analytics system through two key strategies.First,we deploy a deep learning framework using YOLOv9s,trained on the RCA-TVGender dataset.Cameras are positioned perpendicular to observation areas to reduce facial visibility while maintaining accurate gender classification.Second,we apply AES-128 encryption to customer position data,ensuring secure access and regulatory compliance.Our system achieved overall performance,with 81.5%mAP@50,77.7%precision,and 75.7%recall.Moreover,a 90-min observational study confirmed the system’s ability to generate privacy-protected heatmaps revealing distinct behavioral patterns between male and female customers.For instance,women spent more time in certain areas and showed interest in different products.These results confirm the system’s effectiveness in enabling personalized layout and marketing strategies without compromising privacy.
基金supported by the 2024 Research Fund of University of Ulsan.
文摘In recent years,three-dimensional reconstruction technologies that employ multiple cameras have continued to evolve significantly,enabling remote collaboration among users in extended Reality(XR)environments.In addition,methods for deploying multiple cameras for motion capture of users(e.g.,performers)are widely used in computer graphics.As the need to minimize and optimize the number of cameras grows to reduce costs,various technologies and research approaches focused on Optimal Camera Placement(OCP)are continually being proposed.However,as most existing studies assume homogeneous camera setups,there is a growing demand for studies on heterogeneous camera setups.For instance,technical demands keep emerging in scenarios with minimal camera configurations,especially regarding cost factors,the physical placement of cameras given the spatial structure,and image capture strategies for heterogeneous cameras,such as high-resolution RGB cameras and depth cameras.In this study,we propose a pre-visualization and simulation method for the optimal placement of heterogeneous cameras in XR environments,accounting for both the specifications of heterogeneous cameras(e.g.,field of view)and the physical configuration(e.g.,wall configuration)in real-world spaces.The proposed method performs a visibility analysis of cameras by considering each camera’s field-of-view volume,resolution,and unique characteristics,along with physicalspace constraints.This approach enables the optimal position and rotation of each camera to be recommended,along with the minimum number of cameras required.In the results of our study conducted in heterogeneous camera combinations,the proposed method achieved 81.7%~82.7%coverage of the target visual information using only 2~3 cameras.In contrast,single(or homogeneous)-typed cameras were required to use 11 cameras for 81.6%coverage.Accordingly,we found that camera deployment resources can be reduced with the proposed approaches.
