We designed and investigated a passive synchronized mode-locked fiber laser.The device utilizes a dual-cavity structure driven by the nonlinear polarization rotation(NPR)mechanism.Stable mode-locking is attained by sy...We designed and investigated a passive synchronized mode-locked fiber laser.The device utilizes a dual-cavity structure driven by the nonlinear polarization rotation(NPR)mechanism.Stable mode-locking is attained by synergistically controlling gain,polarization state,and optical path length in two symmetric sub-cavities.Experiments proved that repetition rate of the sub-cavities can be adjusted via the time delay line(TDL)to achieve synchronized mode-locking.The system stably generates multi-wavelength pulses at a single repetition frequency,evidenced by multiple spectral peaks and equidistant pulse sequences.These findings facilitate the development of high-performance multi-wavelength ultrashort pulse sources,crucial for optical communications,spectral analysis,and remote sensing.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
Quantum dot(QD)-based fluorescent inks offer high potential due to their tunable emission and high quantum yield,but their practical application suffers from poor environmental stability,aggregation,and challenges in ...Quantum dot(QD)-based fluorescent inks offer high potential due to their tunable emission and high quantum yield,but their practical application suffers from poor environmental stability,aggregation,and challenges in scalable flexible fabrication.In this study,a high-stability fluorescent ink was developed by incorporating QDs into a polydimethylsiloxane(PDMS)colloidal matrix.High-performance patterned films were then obtained via systematic optimization of screen-printing parameters,with film quality governed by substrate type(131μm PDMS),QD concentration(1.5 mg/mL),and screen mesh count(420 mesh).The optimized films exhibit outstanding environmental and photostability,retaining 75.6% of their fluorescence intensity after immersion in deionized water and 63.8% in 75%ethanol at 25℃ for 100 minutes.Under UV irradiation(365 nm,9 W,100 min),fluorescence intensity decreases by less than 20%.Utilizing their daylight transparency and UV-excitable luminescence,various patterns including QR codes and Code 93 standard barcodes were fabricated via screen printing with high pattern fidelity and machine readability.This study presents a scalable and reliable strategy for the fabrication of flexible,high-stability fluorescent films,supporting their integration into next-generation optoelectronic devices,advanced displays,and secure anti-counterfeiting.展开更多
This study investigates the reduction in polarization measurement accuracy caused by varying in-cident angles in a liquid crystal variable retarder(LCVR).The phase delay characteristics of the LCVR were examined,with ...This study investigates the reduction in polarization measurement accuracy caused by varying in-cident angles in a liquid crystal variable retarder(LCVR).The phase delay characteristics of the LCVR were examined,with particular emphasis on the influence of different two-dimensional incident angles on phase delay behavior.Building upon the calibration of phase delay under normal incidence,a phase delay calibra-tion model was developed to account for variations in incident angle and driving voltage.A mathematical re-lationship was established between phase delay and the azimuth angle(α)and pitch angle(β).Experimental validation was conducted under three conditions:α=20°,β=0°;α=0°,β=20°;and an arbitrary angle whereα=5°,β=15°.The results demonstrated that the maximum average deviation between theoretical pre-dictions and experimental measurements did not exceed 0.059 rad.The proposed calibration method proved to be both accurate and practical.This approach offers robust support for LCVR parameter calibration and performance optimization in optical systems,particularly in polarization imaging applications.展开更多
The Hainan Free Trade Port (FTP) is a key step for China to advance high-standard opening-up and seize a more active role in economic globalization. After early trials, a policy framework has taken shape, and islandwi...The Hainan Free Trade Port (FTP) is a key step for China to advance high-standard opening-up and seize a more active role in economic globalization. After early trials, a policy framework has taken shape, and islandwide special customs operations mark a shift from design to full implementation.As of January 18, island-wide special customs operations in Hainan FTP had been in place for a month. The first month ran smoothly, with more efficient flow of goods, people, and capital.This reflects China’s institutional opening-up and, amid slowing global growth, has injected fresh vitality into the regional and global economy.展开更多
Hainan FTP will boost alignment with international rules in China’s opening up The Hainan Free Trade Port(FTP)represents a significant step forward in China’s pursuit of high-standard opening up and greater engageme...Hainan FTP will boost alignment with international rules in China’s opening up The Hainan Free Trade Port(FTP)represents a significant step forward in China’s pursuit of high-standard opening up and greater engagement in economic globalisation.Following years of exploration,the policy framework for the Hainan FTP has come into being,and the recent launch of island-wide special customs operations marks an important transition-from institutional design to full implementation.展开更多
The Hainan Free Trade Port’s implementation of special customs operations marks a significant stride in China’s institutional opening-up,spurring global trade and investment flows.THE Hainan Free Trade Port(FTP)repr...The Hainan Free Trade Port’s implementation of special customs operations marks a significant stride in China’s institutional opening-up,spurring global trade and investment flows.THE Hainan Free Trade Port(FTP)represents a significant step forward in China’s pursuit of high-standard opening-up and greater engagement in economic globalization.展开更多
Founded in September 2020,the International SparkLink Alliance(iSLA)now has approximately 1,200 members in diverse sectors including terminals,homes,vehicles,manufacturing,transportation,finance and healthcare.The iSL...Founded in September 2020,the International SparkLink Alliance(iSLA)now has approximately 1,200 members in diverse sectors including terminals,homes,vehicles,manufacturing,transportation,finance and healthcare.The iSLA has established a technical standards system for wireless short-range communication covering full-stack standards such as the end-to-end protocol system.展开更多
As an ordinary Yunnan local,I never imagined becoming so closely connected to the exotic land of Laos.The luckiest event of my life was probably my choice to tick a box on a 2007 college entrance examination applicati...As an ordinary Yunnan local,I never imagined becoming so closely connected to the exotic land of Laos.The luckiest event of my life was probably my choice to tick a box on a 2007 college entrance examination application form,indicating my willingness to enrollin a major other than my preference,which led me into the world of the Lao language.展开更多
Vision Transformers(ViTs)have achieved remarkable success across various artificial intelligence-based computer vision applications.However,their demanding computational and memory requirements pose significant challe...Vision Transformers(ViTs)have achieved remarkable success across various artificial intelligence-based computer vision applications.However,their demanding computational and memory requirements pose significant challenges for de-ployment on resource-constrained edge devices.Although post-training quantization(PTQ)provides a promising solution by reducing model precision with minimal calibration data,aggressive low-bit quantization typically leads to substantial perfor-mance degradation.To address this challenge,we present the truncated uniform-log2 quantizer and progressive bit-decline reconstruction method for vision Transformer quantization(TP-ViT).It is an innovative PTQ framework specifically designed for ViTs,featuring two key technical contributions:(1)truncated uniform-log2 quantizer,a novel quantization approach which effectively handles outlier values in post-Softmax activations,significantly reducing quantization errors;(2)bit-decline optimiza-tion strategy,which employs transition weights to gradually reduce bit precision while maintaining model performance under extreme quantization conditions.Comprehensive experiments on image classification,object detection,and instance segmenta-tion tasks demonstrate TP-ViT’s superior performance compared to state-of-the-art PTQ methods,particularly in challenging 3-bit quantization scenarios.Our framework achieves a notable 6.18 percentage points improvement in top-1 accuracy for ViT-small under 3-bit quantization.These results validate TP-ViT’s robustness and general applicability,paving the way for more efficient deployment of ViT models in computer vision applications on edge hardware.展开更多
Recently,the zeroing neural network(ZNN)has demonstrated remarkable effectiveness in tackling time-varying problems,delivering robust performance across both noise-free and noisy environments.However,existing ZNN mode...Recently,the zeroing neural network(ZNN)has demonstrated remarkable effectiveness in tackling time-varying problems,delivering robust performance across both noise-free and noisy environments.However,existing ZNN models are limited in their ability to actively suppress noise,which constrains their robustness and precision in solving time-varying problems.This paper introduces a novel active noise rejection ZNN(ANR-ZNN)design that enhances noise suppression by integrating computational error dynamics and harmonic behaviour.Through rigorous theoretical analysis,we demonstrate that the proposed ANR-ZNN maintains robust convergence in computational error performance under environmental noise.As a case study,the ANR-ZNN model is specifically applied to time-varying matrix inversion.Comprehensive computer simulations and robotic experiments further validate the ANR-ZNN's effectiveness,emphasising the proposed design's superiority and potential for solving time-varying problems.展开更多
Heliostat field design for tower solar thermal plants must jointly address solar geometry,optical losses,and layout optimization under engineering constraints.We develop an end-to-end workflow that(i)adopts a consiste...Heliostat field design for tower solar thermal plants must jointly address solar geometry,optical losses,and layout optimization under engineering constraints.We develop an end-to-end workflow that(i)adopts a consistent East–North–Up(ENU)convention for all plant-and sun-related vectors;(ii)integrates cosine efficiency,projection-based shading and blocking(SB),atmospheric transmittance,and an HFLCAL(heliostat field local calculation)truncation model into a single optical chain;and(iii)couples an Eliminate-Blocking(EB)layout prior with an improved“Cheetah”metaheuristic to search ring topology,mirror sizes,and heights while enforcing spacing,kinematics,and rated-power requirements.Projection-based SB is calibrated against Monte-Carlo ray tracing at representative sun positions,and the HFLCAL truncation model is used to quantify sensitivities to sunshape and error-budget parameters.In a three-phase study(fixed-size baseline,uniform sizing,heterogeneous sizing),the EB-guided optimizer improves annual per-area output relative to a radial baseline and reliably attains a 60 MW target.Under equal evaluation budgets,the proposed optimizer converges faster and with lower variance than GA-and PSO-based baselines,while respecting panel-level peak-flux limits through a smooth penalization of flux violations.The resulting layouts exhibit outward-increasing azimuthal spacing and ring-wise size sharing that are consistent with recent heliostat-field deployment experience.The framework is modular,auditable,and readily adaptable to alternative receivers,sites,and cost-aware objectives.展开更多
For 5G millimeter wave(mm-Wave)user equipments(UEs),all test cases must be evaluated in Over-The-Air(OTA)manner.Test time increases dramatically compared to Sub-6 GHz.Therefore,test time reduction is of great signific...For 5G millimeter wave(mm-Wave)user equipments(UEs),all test cases must be evaluated in Over-The-Air(OTA)manner.Test time increases dramatically compared to Sub-6 GHz.Therefore,test time reduction is of great significance for 5G mm-Wave OTA testing.Among all test cases,beam peak search is the most time-consuming,taking up the majority of the overall test time.Therefore,the objective of this work is to determine a suitable beam peak search grid for 5G mm-Wave UEs with satisfactory accuracy and efficiency.Through radiation property investigation of 5G mm-Wave commercial UEs,more reasonable reference array configuration(4×2)and reference deployment scenario(composite beam)are proposed for beam peak search grid analysis.The effect of different grid configurations on beam peak search precision are characterized quantitatively.The determination of associated measurement uncertainty(MU)term along with quantitative analysis approach are proposed based on statistical analysis.Finally,the recommended minimum number of beam peak search grid points is 182 based on the proposed 4×2 array under composite beam scenario.Compared with currently-required 1106 points in 3GPP/CTIA specifications,over 80%reduction can be achieved without increasing the MU limit.The feasibility of the proposed MU analysis as well as the recommended grids is demonstrated through measurements.展开更多
It remains difficult to automate the creation and validation of Unified Modeling Language(UML)dia-grams due to unstructured requirements,limited automated pipelines,and the lack of reliable evaluation methods.This stu...It remains difficult to automate the creation and validation of Unified Modeling Language(UML)dia-grams due to unstructured requirements,limited automated pipelines,and the lack of reliable evaluation methods.This study introduces a cohesive architecture that amalgamates requirement development,UML synthesis,and multimodal validation.First,LLaMA-3.2-1B-Instruct was utilized to generate user-focused requirements.Then,DeepSeek-R1-Distill-Qwen-32B applies its reasoning skills to transform these requirements into PlantUML code.Using this dual-LLM pipeline,we constructed a synthetic dataset of 11,997 UML diagrams spanning six major diagram families.Rendering analysis showed that 89.5%of the generated diagrams compile correctly,while invalid cases were detected automatically.To assess quality,we employed a multimodal scoring method that combines Qwen2.5-VL-3B,LLaMA-3.2-11B-Vision-Instruct and Aya-Vision-8B,with weights based on MMMU performance.A study with 94 experts revealed strong alignment between automatic and manual evaluations,yielding a Pearson correlation of r=0.82 and a Fleiss’Kappa of 0.78.This indicates a high degree of concordance between automated metrics and human judgment.Overall,the results demonstrated that our scoring system is effective and that the proposed generation pipeline produces UML diagrams that are both syntactically correct and semantically coherent.More broadly,the system provides a scalable and reproducible foundation for future work in AI-driven software modeling and multimodal verification.展开更多
Social bots are automated programs designed to spread rumors and misinformation,posing significant threats to online security.Existing research shows that the structure of a social network significantly affects the be...Social bots are automated programs designed to spread rumors and misinformation,posing significant threats to online security.Existing research shows that the structure of a social network significantly affects the behavioral patterns of social bots:a higher number of connected components weakens their collaborative capabilities,thereby reducing their proportion within the overall network.However,current social bot detection methods still make limited use of topological features.Furthermore,both graph neural network(GNN)-based methods that rely on local features and those that leverage global features suffer from their own limitations,and existing studies lack an effective fusion of multi-scale information.To address these issues,this paper proposes a topology-aware multi-scale social bot detection method,which jointly learns local and global representations through a co-training mechanism.At the local level,topological features are effectively embedded into node representations,enhancing expressiveness while alleviating the over-smoothing problem in GNNs.At the global level,a clustering attention mechanism is introduced to learn global node representations,mitigating the over-globalization problem.Experimental results demonstrate that our method effectively overcomes the limitations of single-scale approaches.Our code is publicly available at https://anonymous.4open.science/r/TopoMSG-2C41/(accessed on 27 October 2025).展开更多
Given the growing importance of social media in digital rural development, this study systematically investigated the influence pathways of social media use among rural women in China, drawing on the Technology Accept...Given the growing importance of social media in digital rural development, this study systematically investigated the influence pathways of social media use among rural women in China, drawing on the Technology Acceptance Model(TAM). Employing quantitative research methods, the study conducted empirical tests based on 367 valid questionnaires using Partial Least Squares Structural Equation Modeling(PLS-SEM) via SmartPLS 4.0 software. Results indicate that significant associations exist between perceived ease of use, perceived usefulness, attitudes toward use, behavioral intention, and actual usage behavior. Specifically, the study finds that rural women's perceived ease of use of social media has a significant and positive influence on both perceived usefulness and attitudes toward use. Perceived usefulness further significantly promotes attitudes toward use and behavioral intention. Moreover, positive attitudes toward usage and strong behavioral intentions were effectively converted into actual social media usage behaviors. This study not only validates the applicability and explanatory power of the TAM model in understanding the digital behavior of Chinese rural women but also provides quantitative evidence for how social media enhances their “digital visibility.” These findings offer practical insights for governments and platform providers to optimize user experiences and strengthen digital skills training. Despite its limitations, including a cross-sectional design and a regional sample, this research holds significant theoretical and practical implications.展开更多
In the context of modern software development characterized by increasing complexity and compressed development cycles,traditional static vulnerability detection methods face prominent challenges including high false ...In the context of modern software development characterized by increasing complexity and compressed development cycles,traditional static vulnerability detection methods face prominent challenges including high false positive rates and missed detections of complex logic due to their over-reliance on rule templates.This paper proposes a Syntax-Aware Hierarchical Attention Network(SAHAN)model,which achieves high-precision vulnerability detection through grammar-rule-driven multi-granularity code slicing and hierarchical semantic fusion mechanisms.The SAHAN model first generates Syntax Independent Units(SIUs),which slices the code based on Abstract Syntax Tree(AST)and predefined grammar rules,retaining vulnerability-sensitive contexts.Following this,through a hierarchical attention mechanism,the local syntax-aware layer encodes fine-grained patterns within SIUs,while the global semantic correlation layer captures vulnerability chains across SIUs,achieving synergistic modeling of syntax and semantics.Experiments show that on benchmark datasets like QEMU,SAHAN significantly improves detection performance by 4.8%to 13.1%on average compared to baseline models such as Devign and VulDeePecker.展开更多
Owing to the development of communication technologies and control systems,the integration of numerous Internet of Things(IoT)nodes into the power grid has become increasingly prevalent.These nodes are deployed to gat...Owing to the development of communication technologies and control systems,the integration of numerous Internet of Things(IoT)nodes into the power grid has become increasingly prevalent.These nodes are deployed to gather operational data from various distributed energy sources and monitor real-time energy consumption,thereby transforming the traditional power grid into a smart grid(SG).However,the openness of wireless communication channels introduces vulnerabilities,as it allows potential eavesdroppers to intercept sensitive information.This poses threats to the secure and efficient operation of the IoT-driven smart grid.To address these challenges,we propose a novel scenario that incorporates an Unmanned Aerial Vehicle(UAV)as a relay gateway for multiple authorized smart meters.This scenario is further enhanced by the integration of Reconfigurable Intelligent Surface(RIS)technology,which dynamically adjusts the direction of information transmission.Our objective is to maximize the secure rate within this UAV-RIS-aided system with multiple authorized smart meters and an eavesdropper based on physical layer security(PLS)techniques.We formulate the problem of secure rate maximization by jointly optimizing the active beamforming of the UAV,the passive beamforming of the RIS,and the UAV’s trajectory.To solve this complex optimization problem,we introduce the Twin Soft Actor-Critic(TSAC)algorithm.This algorithm employs a dual-agent framework,where Agent 1 focuses on optimizing the beamforming for both the UAV and the RIS,while Agent 2 concurrently searches for the optimal trajectory of the UAV.Simulation results demonstrate the TSAC algorithm significantly enhances the secure rate of the system,achieving faster convergence and higher rewards under the worst communication conditions.The TSAC algorithm consistently outperforms the Twin Deep Deterministic Policy Gradient(TDDPG)and Twin Delayed Deep Deterministic Policy Gradient(TTD3)algorithms.Furthermore,the TSAC algorithm exhibits robust performance when the distribution of smart meters follows a Gaussian distribution,further validating its practical applicability and effectiveness in real-world scenarios.展开更多
UAV-mounted intelligent reflecting surface(IRS)helps address the line-of-sight(LoS)blockage between sensor nodes(SNs)and the fusion center(FC)in Internet of Things(IoT).This paper considers an IoT assisted by multiple...UAV-mounted intelligent reflecting surface(IRS)helps address the line-of-sight(LoS)blockage between sensor nodes(SNs)and the fusion center(FC)in Internet of Things(IoT).This paper considers an IoT assisted by multiple UAVs-mounted IRS(U-IRS),where the data from ground SNs are transmitted to the FC.In practice,energy efficiency(EE)and mission completion time are crucial metrics for evaluating system performance and operational costs.Recognizing their importance during data collection,we formulate a multi-objective optimization problem to maximize EE and minimize total mission completion time simultaneously.To characterize this tradeoff while considering optimization objective consistency,we construct an optimization problem that minimizes the weighted sum of the total mission completion time and the reciprocal of EE.Due to the non-convex nature of the formulated problem,obtaining optimal solutions is generally challenging.To tackle this issue,we decompose it into three subproblems:UAV-SN association,number of reflecting elements allocation,andUAVtrajectory optimization.An iterative algorithmcombining genetic algorithm,CS-BJ algorithm,and successive convex approximation technique is proposed to solve these sub-problems.Simulation results demonstrate that when the transmitted data amount is 10 and 30Mbits,compared to the static collection benchmark(the UAV hovers directly above each SN),the EE of the proposed method improves by more than 10.4% and 5.2%,while the total mission completion time is reduced by more than 5.4% and 3.3%,respectively.展开更多
Person recognition in photo collections is a critical yet challenging task in computer vision.Previous studies have used social relationships within photo collections to address this issue.However,these methods often ...Person recognition in photo collections is a critical yet challenging task in computer vision.Previous studies have used social relationships within photo collections to address this issue.However,these methods often fail when performing single-person-in-photos recognition in photo collections,as they cannot rely on social connections for recognition.In this work,we discard social relationships and instead measure the relationships between photos to solve this problem.We designed a new model that includes a multi-parameter attention network for adaptively fusing visual features and a unified formula for measuring photo intimacy.This model effectively recognizes individuals in single photo within the collection.Due to outdated annotations and missing photos in the existing PIPA(Person in Photo Album)dataset,wemanually re-annotated it and added approximately ten thousand photos of Asian individuals to address the underrepresentation issue.Our results on the re-annotated PIPA dataset are superior to previous studies in most cases,and experiments on the supplemented dataset further demonstrate the effectiveness of our method.We have made the PIPA dataset publicly available on Zenodo,with the DOI:10.5281/zenodo.12508096(accessed on 15 October 2025).展开更多
文摘We designed and investigated a passive synchronized mode-locked fiber laser.The device utilizes a dual-cavity structure driven by the nonlinear polarization rotation(NPR)mechanism.Stable mode-locking is attained by synergistically controlling gain,polarization state,and optical path length in two symmetric sub-cavities.Experiments proved that repetition rate of the sub-cavities can be adjusted via the time delay line(TDL)to achieve synchronized mode-locking.The system stably generates multi-wavelength pulses at a single repetition frequency,evidenced by multiple spectral peaks and equidistant pulse sequences.These findings facilitate the development of high-performance multi-wavelength ultrashort pulse sources,crucial for optical communications,spectral analysis,and remote sensing.
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Quantum dot(QD)-based fluorescent inks offer high potential due to their tunable emission and high quantum yield,but their practical application suffers from poor environmental stability,aggregation,and challenges in scalable flexible fabrication.In this study,a high-stability fluorescent ink was developed by incorporating QDs into a polydimethylsiloxane(PDMS)colloidal matrix.High-performance patterned films were then obtained via systematic optimization of screen-printing parameters,with film quality governed by substrate type(131μm PDMS),QD concentration(1.5 mg/mL),and screen mesh count(420 mesh).The optimized films exhibit outstanding environmental and photostability,retaining 75.6% of their fluorescence intensity after immersion in deionized water and 63.8% in 75%ethanol at 25℃ for 100 minutes.Under UV irradiation(365 nm,9 W,100 min),fluorescence intensity decreases by less than 20%.Utilizing their daylight transparency and UV-excitable luminescence,various patterns including QR codes and Code 93 standard barcodes were fabricated via screen printing with high pattern fidelity and machine readability.This study presents a scalable and reliable strategy for the fabrication of flexible,high-stability fluorescent films,supporting their integration into next-generation optoelectronic devices,advanced displays,and secure anti-counterfeiting.
文摘This study investigates the reduction in polarization measurement accuracy caused by varying in-cident angles in a liquid crystal variable retarder(LCVR).The phase delay characteristics of the LCVR were examined,with particular emphasis on the influence of different two-dimensional incident angles on phase delay behavior.Building upon the calibration of phase delay under normal incidence,a phase delay calibra-tion model was developed to account for variations in incident angle and driving voltage.A mathematical re-lationship was established between phase delay and the azimuth angle(α)and pitch angle(β).Experimental validation was conducted under three conditions:α=20°,β=0°;α=0°,β=20°;and an arbitrary angle whereα=5°,β=15°.The results demonstrated that the maximum average deviation between theoretical pre-dictions and experimental measurements did not exceed 0.059 rad.The proposed calibration method proved to be both accurate and practical.This approach offers robust support for LCVR parameter calibration and performance optimization in optical systems,particularly in polarization imaging applications.
文摘The Hainan Free Trade Port (FTP) is a key step for China to advance high-standard opening-up and seize a more active role in economic globalization. After early trials, a policy framework has taken shape, and islandwide special customs operations mark a shift from design to full implementation.As of January 18, island-wide special customs operations in Hainan FTP had been in place for a month. The first month ran smoothly, with more efficient flow of goods, people, and capital.This reflects China’s institutional opening-up and, amid slowing global growth, has injected fresh vitality into the regional and global economy.
文摘Hainan FTP will boost alignment with international rules in China’s opening up The Hainan Free Trade Port(FTP)represents a significant step forward in China’s pursuit of high-standard opening up and greater engagement in economic globalisation.Following years of exploration,the policy framework for the Hainan FTP has come into being,and the recent launch of island-wide special customs operations marks an important transition-from institutional design to full implementation.
文摘The Hainan Free Trade Port’s implementation of special customs operations marks a significant stride in China’s institutional opening-up,spurring global trade and investment flows.THE Hainan Free Trade Port(FTP)represents a significant step forward in China’s pursuit of high-standard opening-up and greater engagement in economic globalization.
文摘Founded in September 2020,the International SparkLink Alliance(iSLA)now has approximately 1,200 members in diverse sectors including terminals,homes,vehicles,manufacturing,transportation,finance and healthcare.The iSLA has established a technical standards system for wireless short-range communication covering full-stack standards such as the end-to-end protocol system.
文摘As an ordinary Yunnan local,I never imagined becoming so closely connected to the exotic land of Laos.The luckiest event of my life was probably my choice to tick a box on a 2007 college entrance examination application form,indicating my willingness to enrollin a major other than my preference,which led me into the world of the Lao language.
基金supported by the National Natural Science Foundation of China(Nos.62301092 and 62301093).
文摘Vision Transformers(ViTs)have achieved remarkable success across various artificial intelligence-based computer vision applications.However,their demanding computational and memory requirements pose significant challenges for de-ployment on resource-constrained edge devices.Although post-training quantization(PTQ)provides a promising solution by reducing model precision with minimal calibration data,aggressive low-bit quantization typically leads to substantial perfor-mance degradation.To address this challenge,we present the truncated uniform-log2 quantizer and progressive bit-decline reconstruction method for vision Transformer quantization(TP-ViT).It is an innovative PTQ framework specifically designed for ViTs,featuring two key technical contributions:(1)truncated uniform-log2 quantizer,a novel quantization approach which effectively handles outlier values in post-Softmax activations,significantly reducing quantization errors;(2)bit-decline optimiza-tion strategy,which employs transition weights to gradually reduce bit precision while maintaining model performance under extreme quantization conditions.Comprehensive experiments on image classification,object detection,and instance segmenta-tion tasks demonstrate TP-ViT’s superior performance compared to state-of-the-art PTQ methods,particularly in challenging 3-bit quantization scenarios.Our framework achieves a notable 6.18 percentage points improvement in top-1 accuracy for ViT-small under 3-bit quantization.These results validate TP-ViT’s robustness and general applicability,paving the way for more efficient deployment of ViT models in computer vision applications on edge hardware.
基金supported by the National Science and Technology Major Project(2022ZD0119901)the National Natural Science Foundation of China under Grant(U2141234,62463004 and U24A20260)+1 种基金the Hainan Province Science and Technology Special Fund(ZDYF2024GXJS003)the Scientific Research Fund of Hainan University(KYQD(ZR)23025).
文摘Recently,the zeroing neural network(ZNN)has demonstrated remarkable effectiveness in tackling time-varying problems,delivering robust performance across both noise-free and noisy environments.However,existing ZNN models are limited in their ability to actively suppress noise,which constrains their robustness and precision in solving time-varying problems.This paper introduces a novel active noise rejection ZNN(ANR-ZNN)design that enhances noise suppression by integrating computational error dynamics and harmonic behaviour.Through rigorous theoretical analysis,we demonstrate that the proposed ANR-ZNN maintains robust convergence in computational error performance under environmental noise.As a case study,the ANR-ZNN model is specifically applied to time-varying matrix inversion.Comprehensive computer simulations and robotic experiments further validate the ANR-ZNN's effectiveness,emphasising the proposed design's superiority and potential for solving time-varying problems.
文摘Heliostat field design for tower solar thermal plants must jointly address solar geometry,optical losses,and layout optimization under engineering constraints.We develop an end-to-end workflow that(i)adopts a consistent East–North–Up(ENU)convention for all plant-and sun-related vectors;(ii)integrates cosine efficiency,projection-based shading and blocking(SB),atmospheric transmittance,and an HFLCAL(heliostat field local calculation)truncation model into a single optical chain;and(iii)couples an Eliminate-Blocking(EB)layout prior with an improved“Cheetah”metaheuristic to search ring topology,mirror sizes,and heights while enforcing spacing,kinematics,and rated-power requirements.Projection-based SB is calibrated against Monte-Carlo ray tracing at representative sun positions,and the HFLCAL truncation model is used to quantify sensitivities to sunshape and error-budget parameters.In a three-phase study(fixed-size baseline,uniform sizing,heterogeneous sizing),the EB-guided optimizer improves annual per-area output relative to a radial baseline and reliably attains a 60 MW target.Under equal evaluation budgets,the proposed optimizer converges faster and with lower variance than GA-and PSO-based baselines,while respecting panel-level peak-flux limits through a smooth penalization of flux violations.The resulting layouts exhibit outward-increasing azimuthal spacing and ring-wise size sharing that are consistent with recent heliostat-field deployment experience.The framework is modular,auditable,and readily adaptable to alternative receivers,sites,and cost-aware objectives.
基金supported by the Beijing Natural Science Foundation under Grant L253002.
文摘For 5G millimeter wave(mm-Wave)user equipments(UEs),all test cases must be evaluated in Over-The-Air(OTA)manner.Test time increases dramatically compared to Sub-6 GHz.Therefore,test time reduction is of great significance for 5G mm-Wave OTA testing.Among all test cases,beam peak search is the most time-consuming,taking up the majority of the overall test time.Therefore,the objective of this work is to determine a suitable beam peak search grid for 5G mm-Wave UEs with satisfactory accuracy and efficiency.Through radiation property investigation of 5G mm-Wave commercial UEs,more reasonable reference array configuration(4×2)and reference deployment scenario(composite beam)are proposed for beam peak search grid analysis.The effect of different grid configurations on beam peak search precision are characterized quantitatively.The determination of associated measurement uncertainty(MU)term along with quantitative analysis approach are proposed based on statistical analysis.Finally,the recommended minimum number of beam peak search grid points is 182 based on the proposed 4×2 array under composite beam scenario.Compared with currently-required 1106 points in 3GPP/CTIA specifications,over 80%reduction can be achieved without increasing the MU limit.The feasibility of the proposed MU analysis as well as the recommended grids is demonstrated through measurements.
基金supported by the DH2025-TN07-07 project conducted at the Thai Nguyen University of Information and Communication Technology,Thai Nguyen,Vietnam,with additional support from the AI in Software Engineering Lab.
文摘It remains difficult to automate the creation and validation of Unified Modeling Language(UML)dia-grams due to unstructured requirements,limited automated pipelines,and the lack of reliable evaluation methods.This study introduces a cohesive architecture that amalgamates requirement development,UML synthesis,and multimodal validation.First,LLaMA-3.2-1B-Instruct was utilized to generate user-focused requirements.Then,DeepSeek-R1-Distill-Qwen-32B applies its reasoning skills to transform these requirements into PlantUML code.Using this dual-LLM pipeline,we constructed a synthetic dataset of 11,997 UML diagrams spanning six major diagram families.Rendering analysis showed that 89.5%of the generated diagrams compile correctly,while invalid cases were detected automatically.To assess quality,we employed a multimodal scoring method that combines Qwen2.5-VL-3B,LLaMA-3.2-11B-Vision-Instruct and Aya-Vision-8B,with weights based on MMMU performance.A study with 94 experts revealed strong alignment between automatic and manual evaluations,yielding a Pearson correlation of r=0.82 and a Fleiss’Kappa of 0.78.This indicates a high degree of concordance between automated metrics and human judgment.Overall,the results demonstrated that our scoring system is effective and that the proposed generation pipeline produces UML diagrams that are both syntactically correct and semantically coherent.More broadly,the system provides a scalable and reproducible foundation for future work in AI-driven software modeling and multimodal verification.
基金supported by“the Fundamental Research Funds for the Central Universities”(Grant No.CUCAI2511).
文摘Social bots are automated programs designed to spread rumors and misinformation,posing significant threats to online security.Existing research shows that the structure of a social network significantly affects the behavioral patterns of social bots:a higher number of connected components weakens their collaborative capabilities,thereby reducing their proportion within the overall network.However,current social bot detection methods still make limited use of topological features.Furthermore,both graph neural network(GNN)-based methods that rely on local features and those that leverage global features suffer from their own limitations,and existing studies lack an effective fusion of multi-scale information.To address these issues,this paper proposes a topology-aware multi-scale social bot detection method,which jointly learns local and global representations through a co-training mechanism.At the local level,topological features are effectively embedded into node representations,enhancing expressiveness while alleviating the over-smoothing problem in GNNs.At the global level,a clustering attention mechanism is introduced to learn global node representations,mitigating the over-globalization problem.Experimental results demonstrate that our method effectively overcomes the limitations of single-scale approaches.Our code is publicly available at https://anonymous.4open.science/r/TopoMSG-2C41/(accessed on 27 October 2025).
文摘Given the growing importance of social media in digital rural development, this study systematically investigated the influence pathways of social media use among rural women in China, drawing on the Technology Acceptance Model(TAM). Employing quantitative research methods, the study conducted empirical tests based on 367 valid questionnaires using Partial Least Squares Structural Equation Modeling(PLS-SEM) via SmartPLS 4.0 software. Results indicate that significant associations exist between perceived ease of use, perceived usefulness, attitudes toward use, behavioral intention, and actual usage behavior. Specifically, the study finds that rural women's perceived ease of use of social media has a significant and positive influence on both perceived usefulness and attitudes toward use. Perceived usefulness further significantly promotes attitudes toward use and behavioral intention. Moreover, positive attitudes toward usage and strong behavioral intentions were effectively converted into actual social media usage behaviors. This study not only validates the applicability and explanatory power of the TAM model in understanding the digital behavior of Chinese rural women but also provides quantitative evidence for how social media enhances their “digital visibility.” These findings offer practical insights for governments and platform providers to optimize user experiences and strengthen digital skills training. Despite its limitations, including a cross-sectional design and a regional sample, this research holds significant theoretical and practical implications.
基金supported by the research start-up funds for invited doctor of Lanzhou University of Technology under Grant 14/062402。
文摘In the context of modern software development characterized by increasing complexity and compressed development cycles,traditional static vulnerability detection methods face prominent challenges including high false positive rates and missed detections of complex logic due to their over-reliance on rule templates.This paper proposes a Syntax-Aware Hierarchical Attention Network(SAHAN)model,which achieves high-precision vulnerability detection through grammar-rule-driven multi-granularity code slicing and hierarchical semantic fusion mechanisms.The SAHAN model first generates Syntax Independent Units(SIUs),which slices the code based on Abstract Syntax Tree(AST)and predefined grammar rules,retaining vulnerability-sensitive contexts.Following this,through a hierarchical attention mechanism,the local syntax-aware layer encodes fine-grained patterns within SIUs,while the global semantic correlation layer captures vulnerability chains across SIUs,achieving synergistic modeling of syntax and semantics.Experiments show that on benchmark datasets like QEMU,SAHAN significantly improves detection performance by 4.8%to 13.1%on average compared to baseline models such as Devign and VulDeePecker.
基金supported by State Grid Shanxi Electric Power Company’s Science and Technology Projects(No.52051C230102).
文摘Owing to the development of communication technologies and control systems,the integration of numerous Internet of Things(IoT)nodes into the power grid has become increasingly prevalent.These nodes are deployed to gather operational data from various distributed energy sources and monitor real-time energy consumption,thereby transforming the traditional power grid into a smart grid(SG).However,the openness of wireless communication channels introduces vulnerabilities,as it allows potential eavesdroppers to intercept sensitive information.This poses threats to the secure and efficient operation of the IoT-driven smart grid.To address these challenges,we propose a novel scenario that incorporates an Unmanned Aerial Vehicle(UAV)as a relay gateway for multiple authorized smart meters.This scenario is further enhanced by the integration of Reconfigurable Intelligent Surface(RIS)technology,which dynamically adjusts the direction of information transmission.Our objective is to maximize the secure rate within this UAV-RIS-aided system with multiple authorized smart meters and an eavesdropper based on physical layer security(PLS)techniques.We formulate the problem of secure rate maximization by jointly optimizing the active beamforming of the UAV,the passive beamforming of the RIS,and the UAV’s trajectory.To solve this complex optimization problem,we introduce the Twin Soft Actor-Critic(TSAC)algorithm.This algorithm employs a dual-agent framework,where Agent 1 focuses on optimizing the beamforming for both the UAV and the RIS,while Agent 2 concurrently searches for the optimal trajectory of the UAV.Simulation results demonstrate the TSAC algorithm significantly enhances the secure rate of the system,achieving faster convergence and higher rewards under the worst communication conditions.The TSAC algorithm consistently outperforms the Twin Deep Deterministic Policy Gradient(TDDPG)and Twin Delayed Deep Deterministic Policy Gradient(TTD3)algorithms.Furthermore,the TSAC algorithm exhibits robust performance when the distribution of smart meters follows a Gaussian distribution,further validating its practical applicability and effectiveness in real-world scenarios.
基金supported in part by the Opening Project of Guangxi Wireless Broadband Communication and Signal Processing Key Laboratory under Grant AD25069102in part by the Basic Ability Improvement Project of Young and Middle Aged Teachers in Guangxi Universities,under Grant 2023KY0226+6 种基金in part by Key Laboratory of Cognitive Radio and Information Processing,Ministry of Education of China,underGrant CRKL220108in part by the Innovation Project of Guangxi Graduate Education,under Grant YCBZ2023131in part by the Doctoral Research Foundation of Guilin University of Electronic Technology,under Grant UF23038Yin part by the Bagui Youth Top Talent Projectin part by the Guangxi Key Research and Development Program under Grant AB25069510in part by Open Fund of IPOC(BUPT),No.IPOC2024B07in part by Guangxi Key Laboratory of Precision Navigation Technology and Application,under Grant DH202309.
文摘UAV-mounted intelligent reflecting surface(IRS)helps address the line-of-sight(LoS)blockage between sensor nodes(SNs)and the fusion center(FC)in Internet of Things(IoT).This paper considers an IoT assisted by multiple UAVs-mounted IRS(U-IRS),where the data from ground SNs are transmitted to the FC.In practice,energy efficiency(EE)and mission completion time are crucial metrics for evaluating system performance and operational costs.Recognizing their importance during data collection,we formulate a multi-objective optimization problem to maximize EE and minimize total mission completion time simultaneously.To characterize this tradeoff while considering optimization objective consistency,we construct an optimization problem that minimizes the weighted sum of the total mission completion time and the reciprocal of EE.Due to the non-convex nature of the formulated problem,obtaining optimal solutions is generally challenging.To tackle this issue,we decompose it into three subproblems:UAV-SN association,number of reflecting elements allocation,andUAVtrajectory optimization.An iterative algorithmcombining genetic algorithm,CS-BJ algorithm,and successive convex approximation technique is proposed to solve these sub-problems.Simulation results demonstrate that when the transmitted data amount is 10 and 30Mbits,compared to the static collection benchmark(the UAV hovers directly above each SN),the EE of the proposed method improves by more than 10.4% and 5.2%,while the total mission completion time is reduced by more than 5.4% and 3.3%,respectively.
基金supported by“the Fundamental Research Funds for the Central Universities”(GrantNos.:3282025045,3282024008)“Science and Technology Project of the State ArchivesAdministration ofChina”(Grant No.:2025-Z-009).
文摘Person recognition in photo collections is a critical yet challenging task in computer vision.Previous studies have used social relationships within photo collections to address this issue.However,these methods often fail when performing single-person-in-photos recognition in photo collections,as they cannot rely on social connections for recognition.In this work,we discard social relationships and instead measure the relationships between photos to solve this problem.We designed a new model that includes a multi-parameter attention network for adaptively fusing visual features and a unified formula for measuring photo intimacy.This model effectively recognizes individuals in single photo within the collection.Due to outdated annotations and missing photos in the existing PIPA(Person in Photo Album)dataset,wemanually re-annotated it and added approximately ten thousand photos of Asian individuals to address the underrepresentation issue.Our results on the re-annotated PIPA dataset are superior to previous studies in most cases,and experiments on the supplemented dataset further demonstrate the effectiveness of our method.We have made the PIPA dataset publicly available on Zenodo,with the DOI:10.5281/zenodo.12508096(accessed on 15 October 2025).
