检索结果-南通市图书馆

Towards human-leveled vision systems

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第8期67卷 2331-2349页

作者： DING JianHao huang TieJun State Key Laboratory of Multimedia Information Processing School of Computer SciencePeking UniversityBeijing 100871China

The human visual system is a complex and interconnected network comprising billions of neurons. It plays an essential role in translating environmental light stimuli into information that guides and shapes human perception and action. Research on the visual system aims to uncover the underlying neural structure principles of human visual perception and their possible ***, there are two main approaches: biological system analysis and simulation, artificial intelligence models based on deep learning. Here we aim to discuss the two approaches to human-level vision systems. Deep learning has significantly impacted the field of vision with achievements in representation, modeling, and hardware design. However, there is still a significant gap between deep learning models and the human visual system in terms of scalability, transferability, and sustainability. The progress of the biological visual system can help fill the gap by further understanding the properties and functions of different components of the system. We take the efforts of reconstructing the retina as an example to illustrate that even if we are unable to replicate the visual system on a computer right now, we can still learn a lot by combining existing research outcomes in neuroscience. At the end of the paper, we suggest tracing back to gradually build visual systems from the computational counterpart of biological structures to achieve a human-level vision system in the future.

关键词： human-leveled vision system deep learning biological visual system

同方期刊数据库

Adaptive learning-based optimal tracking control system design and analysis of a disturbed nonlinear hypersonic vehicle model

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第6期67卷 1893-1906页

作者： AN Kai WANG ZhenGuo huang Wei Hypersonic Technology Laboratory National University of Defense TechnologyChangsha410073China

We propose an adaptive learning-based optimal control scheme for height-velocity control models considering model uncertainties and external disturbances of hypersonic winged-cone *** longitudinal nonlinear model is first established and transformed into the control-oriented error equations,and the control scheme is organized by a steady-compensation *** overcome and eliminate the impact of model uncertainties and external disturbances,an adaptive radial basis function neural network(RBFNN)is designed by a q-gradient *** the height-velocity error system with estimated uncertainties into account,the adaptive learning-based optimal tracking control(ALOTC)scheme is proposed by combining the critic-only adaptive dynamic programming(ADP)framework and parameter optimization of system settling ***,a novel weight update law is proposed to satisfy the online iteration requirements,and the algorithm convergence and closedloop stability are discussed by the Lyapunov ***,four simulation cases are provided to prove the effectiveness,accuracy,and robustness of the proposed scheme for the hypersonic longitudinal control system.

关键词： optimal tracking control adaptive dynamic programming RBFNN hypersonic vehicle

Laminar Flame Structure-Dependent Exergy Destruction Behavior at Auto-Ignition Time Scale:A Case Study of Dimethyl Ether(DME)

在线全文

学校读者我要写书评

暂无评论

Journal of Thermal Science 2024年第4期33卷 1257-1271页

作者： WU Honghuan huang Wenlin ZHAO Hao SUN Wuchuan huang Zuohua ZHANG Yingjia State Key Laboratory of Multiphase Flow in Power Engineering Xi’an Jiaotong UniversityXi’an710049China

Hybrid deflagration/auto-ignition flame structures coexist in the combustion of advanced *** exergy destruction caused by different irreversible processes under varied flame regimes is thus important for understanding engine *** this study,the flame propagation modes for the premixed DME/air mixtures are numerically investigated under engine-relevant *** entropy generation and exergy destruction characteristics are compared under different flame *** reveal that as the typical premixed flame transition towards auto-ignition front,the exergy destruction from heat conduction and species mass diffusion gradually vanish and are dominated by chemical *** distributions of temperature and species mole fraction in the flame domain are analyzed to clarify the exergy destruction behaviors caused by heat conduction and mass ***,by dividing the DME oxidation process into four stages,the main reaction channels that contribute to the increase in exergy destruction from chemical reaction have been *** is found that the production and consumption of CH_(2)O and HO_(2) radical dominate the exergy destruction behavior during DME *** this basis,the kinetic mechanism of low-temperature chemistry causing greater exergy destruction is ***,low-temperature chemistry leads to significant exergy destruction due to(a)the large irreversibility of itself and(b)(mainly)the enhancement of H_(2)O_(2)loop reactions by low-temperature reaction *** the reduction of combustion irreversibility is promising to be achieved by reasonably regulating the fuel oxidation path.

关键词： exergy destruction flame structures entropy generation low-temperature chemistry DME

Shock-induced energy localization and reaction growth considering chemical-inclusions effects for crystalline explosives

在线全文

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2024年第3期33卷 278-294页

作者： Ruqin Liu Yanqing Wu Xinjie Wang Fenglei huang Xiaona huang Yushi Wen State Key Laboratory of Explosion Science and Technology Beijing Institute of TechnologyBeijing100081China School of Power and Mechanical Engineering Wuhan UniversityWuhan430072HubeiChina Institute of Chemical Materials China Academy of Engineering Physics(CAEP)Mianyang621999SichuanChina

Chemical inclusions significantly alter shock responses of crystalline explosives in macroscale gap experiments but their microscale dynamics origin remains *** shock-induced energy localization,overall physical responses,and reactions in a-1,3,5-trinitro-1,3,5-triazinane(a-RDX)crystal entrained various chemical inclusions were investigated by the multi-scale shock technique implemented in the reactive molecular dynamics *** indicated that energy localization and shock reaction were affected by the intrinsic factors within chemical inclusions,i.e.,phase states,chemical compositions,and *** atomic origin of chemical-inclusions effects on energy localization is dependent on the dynamics mechanism of interfacial molecules with free space volume,which includes homogeneous intermolecular compression,interfacial impact and shear,and void collapse and *** introducing various chemical inclusions,the initiation of those dynamics mechanisms triggers diverse decay rates of bulk RDX molecules and hereby impacts on growth speeds of final *** chemical inclusions can reduce the effectiveness of the void during the shock *** the shockwave velocity of 9 km/s,the parent RDX decay rate in RDX entrained amorphous carbon decreases the most and is about one fourth of that in RDX with a vacuum void,and solid HMX and TATB inclusions are more reactive than amorphous carbon but less reactive than dry air or acetone *** lessdense shocking system denotes the greater increases in local temperature and stress,the faster energy liberation,and the earlier final reaction into equilibrium,revealing more pronounced responses to the present intense *** quantitative models associated with the relative system density(RD_(sys))were proposed for indicating energy-localization mechanisms and evaluating initiation safety in the shocked crystalline ***_(sys)is defined by the density ratio of defective RDX to perfec

关键词： Shock responses Energy localization Crystalline explosives Chemical inclusions Reactive molecular dynamics

Cyclic Pulsating Pressure Enhanced Segregating Structuration of Ultra-High Molecular Weight Polyethylene/Graphene Composites as High-performance Light-Weight EMI Shields

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Polymer Science 2024年第7期42卷 958-967,I0008页

作者： Yun-Zhi huang Xiao-Xiao Liu Lan-Wei Li Guang-Ming huang Zhao-Xia huang Jin-Ping Qu National Engineering Research Center of Novel Equipment for Polymer Processing Key Laboratory of Polymer Processing EngineeringMinistry of EducationGuangdong Provincial Key Laboratory of Technique and Equipment for Macromolecular Advanced ManufacturingSchool of Mechanical and Automotive EngineeringSouth China University of TechnologyGuangzhou 510641China School of Advanced Manufacturing Technology Guangdong Mechanical&Electrical PolytechnicGuangzhou 510550China

Currently,the enhancement in electromagnetic interference(EMI)performance of polymeric composite generally relies on either improving electrical conductivity(σ)for stronger electromagnetic(EM)reflections or tailoring structure for higher EM ***,we proposed a novel technique called cyclic pulsating pressure enhanced segregating structuration(CPP-SS),which can reinforce these two factors *** structural information was supplied by optical microscopy(OM)and scanning electron microscopy(SEM),both of which confirmed the formation and evolution of segregate structured ultra-high molecular weight polyethylene(UHMWPE)/graphene ***,the result showed that CPP-SS can significantly improve theσof ***,advanced specific EMI shielding efficiency of 31.1 d B/mm was achieved for UHMWPE/graphene composite at 1-mm thickness and a low graphene loading of 5 wt%.Meanwhile,it also confirmed that the intrinsic disadvantage of poor mechanical properties of conventional segregated structure composites can be *** work is believed to provide a fundamental understanding of the structural and performance evolutions of segregated structured composites prepared under CPPSS,and to bring us a simple and efficient approach for fabricating high-performance,strong and light-weight polymeric EMI shields.

关键词： Cyclic pulsating pressure Segregated structure Ultra-high molecular weight polyethylene Graphene Electromagnetic interface

Machine learning-assisted sparse observation assimilation for real-time aerodynamic field perception

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第5期67卷 1458-1469页

作者： ZHAO QingYu huang Jun GUO YuXin PAN YuXuan JI JingJing huang YongAn State Key Laboratory of Intelligent Manufacturing Equipment and Technology School of Mechanical Science and EngineeringHuazhong University of Science and TechnologyWuhan 430074China

Accurate aerodynamic distribution perception and real-time flight state evaluation are crucial for flight safety,e.g.,stall ***,the observations are usually sparse due to limitations in sensor mounting space and cost,and a reconstruction technology is urgently ***,a machine learning-assisted assimilation method based on sparse observations has been *** from the traditional reconstruction methods focusing on boundary condition correction,the proposed method formulates the flow field pressure distribution as a linear superposition of flow field modes,thereby forming a real-time reconstruction pattern that combines offline modal extraction using computational fluid dynamics(CFD)with real-time determination of modal weights using a neural *** this study,CFD simulations were conducted under 800different operating conditions for common modal extraction and model *** weights of these modes were determined online based on merely five observations for reconstructing the full pressure field.A pressure reconstruction with a relative error of 6.1%and a mean square error of 0.003 was achieved within the prescribed condition *** computational cost was just2 ms for each reconstruction run,significantly faster than the 20 min required by the classical reconstruction ensemble transform Kalman *** also showed that the method maintains almost the same accuracy amidst 1.5%measurement *** practical examples,shock waves and the change of lift coefficient were analyzed using the proposed method,providing remarkable evidence for the capability of the method in supporting stall *** validate the method’s effectiveness and explore its potential in real-time and accurate monitoring of an aircraft.

关键词： aerodynamic force sparse observation neural networks pressure field reconstruction

Superiority of a Convolutional Neural Network Model over Dynamical Models in Predicting Central Pacific ENSO

在线全文

学校读者我要写书评

暂无评论

Advances in Atmospheric Sciences 2024年第1期41卷 141-154页

作者： Tingyu WANG Ping huang Center for Monsoon System Research Institute of Atmospheric PhysicsChinese Academy of SciencesBeijing 100029China Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters(CIC-FEMD) Nanjing University of Information Science&TechnologyNanjing 210044China State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics Institute of Atmospheric PhysicsChinese Academy of SciencesBeijing 100029China University of Chinese Academy of Sciences Beijing 100049China

The application of deep learning is fast developing in climate prediction,in which El Ni?o–Southern Oscillation(ENSO),as the most dominant disaster-causing climate event,is a key *** studies have shown that deep learning methods possess a certain level of superiority in predicting ENSO *** present study develops a deep learning model for predicting the spatial pattern of sea surface temperature anomalies(SSTAs)in the equatorial Pacific by training a convolutional neural network(CNN)model with historical simulations from CMIP6 *** with dynamical models,the CNN model has higher skill in predicting the SSTAs in the equatorial western-central Pacific,but not in the eastern *** CNN model can successfully capture the small-scale precursors in the initial SSTAs for the development of central Pacific ENSO to distinguish the spatial mode up to a lead time of seven months.A fusion model combining the predictions of the CNN model and the dynamical models achieves higher skill than each of them for both central and eastern Pacific ENSO.

关键词： ENSO diversity deep learning ENSO prediction dynamical forecast system

Higher-dimensional integrable deformations of the classical Boussinesq-Burgers system

在线全文

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2024年第6期76卷 1-12页

作者： Xiaoyu Cheng Qing huang School of Mathematics Northwest UniversityXi'an 710127China

In this paper,the(1+1)-dimensional classical Boussinesq-Burgers(CBB)system is extended to a(4+1)-dimensional CBB system by using its conservation laws and the deformation *** Lax integrability,symmetry integrability and a large number of reduced systems of the new higher-dimensional system are ***,for illustration,an exact solution of a(1+1)-dimensional reduced system is constructed from the viewpoint of Lie symmetry analysis and the power series method.

关键词： classical Boussinesq-Burgers system higher-dimensional system deformation algorithm Lax pair conservation law

Enhanced boron neutron capture therapy(BNCT)through controlled drug release via boron-loaded nanofiber mats

在线全文

学校读者我要写书评

暂无评论

Nano Research 2024年第8期17卷 7479-7492页

作者： Wenyong huang Yongjin Yang Yong Pan Shiwei Jing Yanxin Qi Yubin huang Faculty of Chemistry Northeast Normal UniversityChangchun 130024China Key Laboratory of Sustainable Advanced Functional Materials of Jilin Province Changchun 130024China School of Physics Northeast Normal UniversityChangchun 130024China

Boron neutron capture therapy(BNCT)is a novel binary therapy combining boron targeted drugs and neutron irradiation,which can selectively and effectively kill cancer cells at the cellular *** release of boron drug and its accumulation in tumor sites are the crux of ***,we developed a^(10)B-boric acid(^(10)BA)-loaded nanofiber applying for BNCT by in situ *** nanofibers were obtained by electrospinning technique using polyethylene glycol/polylactide(PEO/PLA)block *** changing the ratio of hydrophilicity to hydrophobicity of the nanofibers,the controlled release and the effective accumulation of boron 10 isotope(^(10)B)were achieved in ***^(10)B content in tumor could reach to 2540μg/g,significantly exceeding the required level of 20–50μg/g for BNCT *** pertinent DNA damage experiments,direct evidence and quantified data of BNCT-induced DNA damage in tumor cells were obtained for the first *** sequencing was employed to predict the molecular mechanisms and potential signaling pathways of BNCT,providing theoretical basis for future combined *** antitumor efficiency of BNCT was demonstrated by establishing mice model of subcutaneous tumor and tumor *** research presents a novel boron-loaded nanofiber mats for BNCT,which enables controlled drug release and holds significant potential in the treatment of unresectable or postoperative residual tumors.

关键词： boron neutron capture therapy nanofibers controlled release in situ administration