检索结果-南通市图书馆

Heterogeneous decentralised machine unlearning with seed model distillation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

CAAI Transactions on Intelligence Technology 2024年第3期9卷 608-619页

作者： Guanhua ye Tong chen Quoc Viet Hung Nguyen Hongzhi Yin School of Information Technology and Electrical Engineering The University of QueenslandSt LuciaQueenslandAustralia School of Information and Communication Technology Griffith UniversityBrisbaneQueenslandAustralia

As some recent information security legislation endowed users with unconditional rights to be forgotten by any trained machine learning model,personalised IoT service pro-viders have to put unlearning functionality into their *** most straight-forward method to unlearn users'contribution is to retrain the model from the initial state,which is not realistic in high throughput applications with frequent unlearning *** some machine unlearning frameworks have been proposed to speed up the retraining process,they fail to match decentralised learning scenarios.A decentralised unlearning framework called heterogeneous decentralised unlearning framework with seed(HDUS)is designed,which uses distilled seed models to construct erasable en-sembles for all ***,the framework is compatible with heterogeneous on-device models,representing stronger scalability in real-world *** experiments on three real-world datasets show that our HDUS achieves state-of-the-art performance.

关键词： data mining data privacy machine learning

High-frequency Resonance Analysis and Impedance Reshaping Control of MMC-HVDC System Based on Frequency Coupling Impedance Model

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Modern Power Systems and Clean Energy 2024年第2期12卷 646-657页

作者： Juanjuan Wang Wei chen Yuekun Liu Chuang Fu Yunming ye Junjie Feng School of Electric Power South China University of TechnologyGuangzhouChina Guangdong Power Grid Co. Ltd.GuangzhouChina Electric Power Research Institute China Southern Power GridGuangzhouChina

In recent years, high-frequency resonance (HFR) events occurred in several modular multilevel converter based high-voltage direct current (MMC-HVDC) projects. The time delay of an MMC-HVDC system is the critical factor that induces HFR. The frequency coupling affects the impedance characteristics of an MMC and further deteriorates system stability. Therefore, in this paper, a multi-input multi-output admittance model of an MMC-HVDC system is developed to analyze its frequency characteristics. The effects of current loop, power loop, phase-locked loop, and operating point on the MMC frequency coupling degree are analyzed in detail. Meanwhile, to further suppress HFR in the MMC-HVDC system, an enhanced impedance reshaping control strategy based on the equivalent single-input single-output impedance model is proposed. Finally, the accuracy of the enhanced impedance model and the effectiveness of the impedance reshaping control are verified by electromagnetic transient simulations in PSCAD.

关键词： Modular multilevel converter based high-voltage direct current(MMC-HVDC) high-frequency resonance frequency coupling impedance model impedance reshaping control

Accelerating H^(*)desorption of hollow Mo_(2)C nanoreactor via in-situ grown carbon dots for electrocatalytic hydrogen evolution

在线全文

学校读者我要写书评

暂无评论

Journal of Energy Chemistry 2024年第9期96卷 464-471页

作者： Mengmeng Liu Yuanyuan Jiang Zhuwei Cao Lulu Liu Hong chen Sheng ye School of Materials and Chemistry Anhui Agricultural UniversityHefei 230036AnhuiChina

Molybdenum carbide(Mo_(2)C)is a promising non-noble metal electrocatalyst with electronic structures similar to Pt for hydrogen evolution reaction(HER).However,strong H^(*)adsorption at the Mo sites hinders the improvement of HER ***,we synthesized monodisperse hollow Mo_(2)C nanoreactors,in which the carbon dots(CD)were in situ formed onto the surface of Mo_(2)C through carburization *** to finite element simulation and analysis,the CD@Mo_(2)C possesses better mesoscale diffusion properties than Mo_(2)C *** optimized CD@Mo_(2)C nanoreactor demonstrates superior HER performance in alkaline electrolyte with a low overpotential of 57 mV at 10 mA cm^(−2),which is better than most Mo_(2)C-based ***,CD@Mo_(2)C exhibits excellent electrochemical stability during 240 h,confirmed by operando Raman and X-ray diffraction(XRD).Density functional theory(DFT)calculations show that carbon dots cause the d-band center of CD@Mo_(2)C to shift away from Fermi level,promoting water dissociation and the desorption of H^(*).This study provides a reasonable strategy towards high-activity Mo-based HER eletrocatalysts by modulating the strength of Mo–H bonds.

关键词： Mo_(2)C nanoreactor Carbon dots H^(*)desorption Electrocatalytic hydrogen evolution

Unified deep learning model for predicting fundus fluorescein angiography image from fundus structure image

在线全文

学校读者我要写书评

暂无评论

Journal of Innovative Optical Health Sciences 2024年第3期17卷 105-113页

作者： Yiwei chen Yi He Hong ye Lina Xing Xin Zhang Guohua Shi Jiangsu Key Laboratory of Medical Optics Suzhou Institute of Biomedical Engineering and TechnologyChinese Academy of SciencesSuzhou 215163P.R.China School of Biomedical Engineering(Suzhou) Division of Life Sciences and MedicineUniversity of Science and Technology of China Hefei 230026P.R.China Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of SciencesShanghai 200031P.R.China

The prediction of fundus fluorescein angiography(FFA)images from fundus structural images is a cutting-edge research topic in ophthalmological image *** comprises estimating FFA from fundus camera imaging,single-phase FFA from scanning laser ophthalmoscopy(SLO),and three-phase FFA also from *** many deep learning models are available,a single model can only perform one or two of these prediction *** accomplish three prediction tasks using a unified method,we propose a unified deep learning model for predicting FFA images from fundus structure images using a supervised generative adversarial *** three prediction tasks are processed as follows:data preparation,network training under FFA supervision,and FFA image prediction from fundus structure images on a test *** comparing the FFA images predicted by our model,pix2pix,and CycleGAN,we demonstrate the remarkable progress achieved by our *** high performance of our model is validated in terms of the peak signal-to-noise ratio,structural similarity index,and mean squared error.

关键词： Fundus fluorescein angiography image fundus structure image image translation unified deep learning model generative adversarial networks

In-situ measurement via the flow-through method and numerical simulations for radon exhalation during measurements of the radon exhalation rate

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Nuclear Science and Techniques 2024年第7期35卷 192-207页

作者： Ming Xia Yong-Jun ye Shan-Wei Shang Ting Yu Dai-Jia chen School of Resources Environment and Safety Engineering University of South ChinaHengyang421001China Key Discipline Laboratory for National Defense for Biotechnology in Uranium Mining and Hydrometallurgy University of South ChinaHengyang421001China

Small-scale measurements of the radon exhalation rate using the flow-through and closed-loop methods were conducted on the surface of a uranium tailing pond to better understand the differences between the two *** abnormal radon exhalation behavior was observed,leading to computational fluid dynamics(CFD)-based simulations in which dynamic radon migration in a porous medium and accumulation chamber was *** on the in-situ experimental and numerical simulation results,variations in the radon exhalation rate subject to permeability,flow rate,and insertion depth were quantified and *** in-situ radon exhalation rates measured using the flow-through method were higher than those measured using the closed-loop method,which could be explained by the negative pressure difference between the inside and outside of the chamber during the *** consistency of the variations in the radon exhalation rate between the experiments and simulations suggests the reliability of CFD-based techniques in obtaining the dynamic evolution of transient radon exhalation rates for diffusion and convection at the porous medium-air *** synergistic effects of the three factors(insertion depth,flow rate,and permeability)on the negative pressure difference and measured exhalation rate were quantified,and multivariate regression models were established,with positive correlations in most cases;the exhalation rate decreased with increasing insertion depth at a permeability of 1×10^(−11) m^(2).CFD-based simulations can provide theoretical guidance for improving the flow-through method and thus achieve accurate measurements.

关键词： Radon exhalation Flow-through Numerical simulation Accumulation chamber Multivariate regression

Enhancing hydrolytic stability of dynamic imine bonds and polymers in acidic media with internal protecting groups

在线全文

学校读者我要写书评

暂无评论

Chinese Chemical Letters 2024年第5期35卷 161-165页

作者： Hang chen chengzhi Cui Hebo ye Hanxun Zou Lei You College of Environmental and Biological Engineering Fujian Provincial Key Laboratory of Ecology-Toxicological Effects&Control for Emerging ContaminantsPutian UniversityPutian 351100China State Key Laboratory of Structural Chemistry Fujian Institute of Research on the Structure of MatterChinese Academy of SciencesFuzhou 350002China College of Chemistry and Material Science Fujian Normal UniversityFuzhou 350007China Fujian Science&Technology Innovation Laboratory for Optoelectronic Information of China Fuzhou 350108China

Imine bonds are among the most explored building motifs in dynamic chemistry,polymers,and materials,and yet,their acid-resistance remains a longstanding *** we demonstrate a concept of internal protecting groups for improving the kinetic stability of dynamic imine bonds and *** examination of structure-reactivity relationship of a series of aldehydes/imines bearing a neighboring carboxyl allowed uncovering of required structural features for dynamically masking imine bonds with cyclic *** studies indicated that noncovalent interactions along with sterics control the ringchain equilibrium and the stability of imine *** incorporation of internal protecting groups into imine polymers further enabled their controlled stability in acidic ***,a combination of dynamic covalent network and coordination supramolecular network provided a facile means for the modulation of luminescent and mechanical properties of *** strategies and results reported should be beneficial to molecular assemblies,dynamic polymers,biological delivery,and intelligent materials.

关键词： Dynamic covalent chemistry Internal protecting group Kinetic stability Imine bond Dynamic polymer

Polyethylene interfacial dielectric layer for organic semiconductor single crystal based field-effect transistors

在线全文

学校读者我要写书评

暂无评论

Chinese Chemical Letters 2024年第4期35卷 431-435页

作者： Min chen Boyu Peng Xuyun Guo ye Zhu Hanying Li Department of Polymer Science and Engineering MOE Key Laboratory of Macromolecular Synthesis and FunctionalizationInternational Research Center for X PolymersShanxi-Zheda Institute of Advanced Materials and Chemical EngineeringZhejiang UniversityHangzhou 310027China Department of Applied Physics Research Institute for Smart EnergyThe Hong Kong Polytechnic UniversityHong Kong 999077China

Organic semiconductor single crystals(OSSCs) have shown their promising potential in high-performance organic field-effect transistors(OFETs). The interfacial dielectric layers are critical in these OFETs as they not only govern the key semiconductor/dielectric interface quality but also determine the growth of OSSCs by their wetting properties. However, reported interfacial dielectric layers either need rigorous preparation processes, rely on certain surface chemistry reactions, or exhibit poor solvent resistance, which limits their applications in low-cost, large-area, monolithic fabrication of OSSC-based OFETs. In this work, polyethylene(PE) thin films and lamellar single crystals are utilized as the interfacial dielectric layers, providing solvent resistive but wettable surfaces that facilitate the crystallization of 6,13-bis(tri-isopropylsilylethynyl)pentacene(TIPS-PEN) and 6,13-bis(triisopropylsilylethynyl)-5,7,12,14-tetraazapentacene(TIPS-TAP). As evidenced by the presence of ambipolar behavior in TIPS-PEN single crystals and the high electron mobility(2.3 ± 0.34 cm^(2)V^(-1)s^(-1)) in TIPS-TAP single crystals, a general improvement on electron transport with PE interfacial dielectric layers is revealed, which likely associates with the chemically inertness of the saturated C-H bonds. With the advantages in both processing and device operation, the PE interfacial dielectric layer potentially offers a monolithic way for the enhancement of electron transport in solution-processed OSSC-based OFETs.

关键词： Polyethylene Interfacial dielectric layer Organic semiconductor Single crystal Organic field-effect transistors

Deep learning for joint channel estimation and feedback in massive MIMO systems

在线全文

学校读者我要写书评

暂无评论

Digital Communications and Networks 2024年第1期10卷 83-93页

作者： Jiajia Guo Tong chen Shi Jin Geoffrey ye Li Xin Wang Xiaolin Hou National Mobile Communications Research Laboratory Southeast UniversityNanjing210096China Department of Electrical and Electronic Engineering Imperial College LondonLondonSW72AZUK DOCOMO Beijing Communications Laboratories Co. LtdBeijingChina

The great potentials of massive Multiple-Input Multiple-Output(MIMO)in Frequency Division Duplex(FDD)mode can be fully exploited when the downlink Channel State Information(CSI)is available at base ***,the accurate CsI is difficult to obtain due to the large amount of feedback overhead caused by massive *** this paper,we propose a deep learning based joint channel estimation and feedback framework,which comprehensively realizes the estimation,compression,and reconstruction of downlink channels in FDD massive MIMO *** networks are constructed to perform estimation and feedback explicitly and *** explicit network adopts a multi-Signal-to-Noise-Ratios(SNRs)technique to obtain a single trained channel estimation subnet that works well with different SNRs and employs a deep residual network to reconstruct the channels,while the implicit network directly compresses pilots and sends them back to reduce network *** module is also designed to generate data-bearing *** results show that the two proposed networks exhibit excellent performance of reconstruction and are robust to different environments and quantization errors.

关键词： Channel estimation CSI feedback Deep learning Massive MIMO FDD

Effect of Sintering Temperature on the Microstructure and Mechanical Properties of Nanocrystalline Cemented Carbide

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Wuhan University of Technology(Materials Science) 2024年第3期39卷 664-672页

作者：陈先富刘颖 ye Jinwen WANG Lu School of Materials Science and Engineering Sichuan UniversityChengdu 610065China Key Laboratory of Advanced Special Ma-terial&Technology Ministry of EducationChengdu 610065China

WC-Co nanocrystalline nitrogen-containing cemented carbides were prepared by vacuum sintering and low pressure *** sintering processes of Cr_(2)(C,N)doped nano WC-Co powders were studied by using thermogravimetric analysis(TGA)and differential scanning calorimetry(DSC).The effect of sintering temperature on the microstructure and mechanical properties of nanocrystalline cemented carbide was studied by scanning electron microscope(SEM),high resolution transmission electron microscope(HRTEM)and mechanical property *** results showed that the nano WC grains began to grow in the solid phase sintering stage.A high-performance nano-nitrogen-containing cemented carbide with uniform microstructure and good interfacial bonding can be obtained by increasing the sintering temperature to 1380℃.It has a transverse rupture strength(TRS)of 5057 MPa and a hardness of 1956 HV30.

关键词： nano nitrogen cemented carbide sintering temperature microstructure mechanical properties