检索结果-南通市图书馆

Filtering and regret network for spacecraft component segmentation based on gray images and depth maps

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第7期37卷 439-449页

作者： Xiang LIU Hongyuan WANG Zijian WANG Xinlong CHEN Weichun CHEN Zhengyou XIE institute of Aeronautics Harbin Institute of TechnologyHarbin 150001China Qian Xuesen Laboratory of Space technology China Academy of Space TechnologyBeijing 100080China

Identifying and segmenting spacecraft components is vital in many on-orbit space missions,such as on-orbit maintenance and component *** depth maps with visual images has been proven effective in improving segmentation ***,existing methods ignore the noise and fallacy in collected depth maps,which interfere with the network to extract representative features,decreasing the final segmentation *** this end,this paper proposes a Filtering and Regret Network(FRNet)for spacecraft component *** FRNet incorporates filtering and regret mechanisms to suppress the abnormal depth response in shallow layers and selectively reuses the filtered cues in deep layers,avoiding the detrimental effects of low-quality depth information while preserving the semantic context inherent in depth ***,a two-stage feature fusion module is proposed,which involves information interaction and *** module effectively explores the feature correlation and unifies the multimodal features into a comprehensive ***,a large-scale spacecraft component recognition dataset is constructed for training and evaluating spacecraft component segmentation *** results demonstrate that the FRNet achieves a state-of-the-art performance with a mean Intersection Over Union(mIOU)of 84.13%and an average inference time of 133.2 ms when tested on an NVIDIA RTX 2080 SUPER GPU.

关键词： Spacecraft component recognition Multi-modal feature fusion Satellite dataset Intelligent systems Deep learning

维普期刊数据库

AutoQNN: An End-to-End Framework for Automatically Quantizing Neural Networks

在线全文

学校读者我要写书评

暂无评论

Journal of Computer Science & technology 2024年第2期39卷 401-420页

作者：龚成卢冶代素蓉邓倩杜承昆李涛 College of Software Nankai UniversityTianjin 300350China College of Computer Science Nankai UniversityTianjin 300350China State Key Laboratory of Processors Institute of Computing TechnologyChinese Academy of Sciences Beijing 100190China

Exploring the expected quantizing scheme with suitable mixed-precision policy is the key to compress deep neural networks(DNNs)in high efficiency and *** exploration implies heavy workloads for domain experts,and an automatic compression method is ***,the huge search space of the automatic method introduces plenty of computing budgets that make the automatic process challenging to be applied in real *** this paper,we propose an end-to-end framework named AutoQNN,for automatically quantizing different layers utilizing different schemes and bitwidths without any human *** can seek desirable quantizing schemes and mixed-precision policies for mainstream DNN models efficiently by involving three techniques:quantizing scheme search(QSS),quantizing precision learning(QPL),and quantized architecture generation(QAG).QSS introduces five quantizing schemes and defines three new schemes as a candidate set for scheme search,and then uses the Differentiable Neural Architecture Search(DNAS)algorithm to seek the layer-or model-desired scheme from the *** is the first method to learn mixed-precision policies by reparameterizing the bitwidths of quantizing schemes,to the best of our *** optimizes both classification loss and precision loss of DNNs efficiently and obtains the relatively optimal mixed-precision model within limited model size and memory *** is designed to convert arbitrary architectures into corresponding quantized ones without manual intervention,to facilitate end-to-end neural network *** have implemented AutoQNN and integrated it into *** experiments demonstrate that AutoQNN can consistently outperform state-of-the-art *** 2-bit weight and activation of AlexNet and ResNet18,AutoQNN can achieve the accuracy results of 59.75%and 68.86%,respectively,and obtain accuracy improvements by up to 1.65%and 1.74%,respectively,compared with state-of-the-art ***,c

关键词： automatic quantization mixed precision quantizing scheme search quantizing precision learning quan-tized architecture generation

The swirl and pyrolysis reaction synergistically enhance solid-solid heat transfer and product separation in cyclone pyrolyzer Author links open overlay panel

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Particuology 2024年第8期91卷 118-127页

作者： Haorui Niu Jingxuan Yang Wenhao Lian Nan Zhang Yongguo Li Hongzheng Ren Xiaogang Hao Guoqing Guan College of Chemical Engineering and technology Taiyuan University of TechnologyTaiyuan030024 china School of Chemical Engineering and technology North University of ChinaTaiyuan030051 china CNNC Key Laboratory on Radioactive Gas Purification technology China Institute for Radiation ProtectionTaiyuan030006 china Energy Conversion Engineering Laboratory Institute of Regional InnovationHirosaki University3-BunkyochoHirosaki036-8561Japan

Cyclone pyrolyzer is a novel type of downer that combines centrifugal force field and double-layer cyclone *** on transfer behavior is helpful to optimize the pyrolyzer to meet the needs of *** this study,the Computational Particle Fluid Dynamics(CPFD)model is used to analyze the transfer behavior of binary particles,and finds that the swirl and reaction have a synergistic *** effect can increase the heating rate of the particles to the range of flash pyrolysis,and its mechanism lies in the flow field structure of the *** to the centrifugal force field,the particles gather to the near *** rapid swirl,which facilitates intense gas-solid heat transfer,leads to the rapid heating and pyrolysis of biomass *** the pyrolysis proceeds,the mass of the biomass particles becomes smaller and they are more easily affected by the gas flow in *** the action of gas flow,char particles serve as new heat carrier to form the inner cycle of particles,which strengthens the heating *** pyrolysis products are discharged from the exhaust port in time with the flow field of the pyrolyzer to achieve separation from the heat carrier and inhibit the occurrence of secondary reactions.

关键词： Biomass pyrolysis Cyclone pyrolyzer Intra-particle circulation Synergistic effect

Mechanical design and analysis of bio-inspired reentrant negative Poisson’s ratio metamaterials with rigid-flexible distinction

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Smart and Nano Materials 2024年第1期15卷 1-20页

作者： Xinchun Zhang Junyu Wang Qidong Sun Jingyang Li Sheng Zhou Junfeng Qi Ran Tao Hebei Key Laboratory of Electric Machinery Health Maintenance&Failure Prevention North China Electric Power UniversityBaodingChina institute of Advanced Structure technology Beijing Institute of TechnologyBeijingChina Beijing Spacecrafts Manufacturing Factory BeijingChina

Aiming at achieving tunable reentrant structures with rigidity and uniformity,respectively,the C-shaped and S-shaped reentrant metamaterials were proposed by the bionic design of animal *** beam theory and energy methodology,the analytical expressions of the equivalent elastic modulus of the metamaterials were *** in deformation modes,mechanical properties,and energy absorption capacities were characterized by using experiments and the finite element analysis *** effects of ligament angle and thickness on the mechanical characteristics of two novel metamaterials were investigated by using a parametric *** results show that the stiffness,deformation mode,stress-strain curve,and energy absorption effects of three metamaterials are significantly *** design philosophy can be extended from 2D to 3D and is applicable at multiple dimensions.

关键词： Re-entrant mechanical metamaterials negative poisson’s ratio enhanced stiffness rigid-flexible distinction energy absorption

Extracting Integrated Features of Electronic Medical Records Big Data for Mortality and Phenotype Prediction

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第3期33卷 776-792页

作者： Fei LI Yiqiang CHEN Yang GU Yaowei WANG institute of computing technology University of Chinese Academy of Sciences Pengcheng Laboratory

The key to synthesizing the features of electronic medical records(EMR) big data and using them for specific medical purposes,such as mortality and phenotype prediction,is to integrate the individual medical event and the overall multivariate time series feature extraction automatically,as well as to alleviate data imbalance *** paper provides a general feature extraction method to reduce manual intervention and automatically process large-scale *** processing uses two variational auto-encoders(VAEs) to automatically extract individual and global *** avoids the well-known posterior collapse problem of Transformer VAE through a uniquely designed "proportional and stabilizing" mechanism and forms a unique means to alleviate the data imbalance *** conducted experiments using ICU-STAY patients' data from the MIMIC-Ⅲ database and compared them with the mainstream EMR time series processing *** results show that the method extracts visible and comprehensive features,alleviates data imbalance problems and improves the accuracy in specific predicting tasks.

关键词： Databases Time series analysis MIMICs Manuals Big Data Feature extraction Transformers

Robust reinforcement learning with UUB guarantee for safe motion control of autonomous robots

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science china(Technological Sciences) 2024年第1期67卷 172-182页

作者： ZHANG RuiXian HAN YiNing SU Man LIN ZeFeng LI HaoWei ZHANG LiXian School of Astronautics Harbin Institute of TechnologyHarbin 150001China School of Management Harbin Institute of TechnologyHarbin 150001China Beijing institute of Tracking and Telecommunication technology Beijing 100094China

This paper addresses the issue of safety in reinforcement learning(RL)with disturbances and its application in the safety-constrained motion control of autonomous *** tackle this problem,a robust Lyapunov value function(rLVF)is *** rLVF is obtained by introducing a data-based LVF under the worst-case disturbance of the observed *** the rLVF,a uniformly ultimate boundedness criterion is *** criterion is desired to ensure that the cost function,which serves as a safety criterion,ultimately converges to a range via the policy to be ***,to mitigate the drastic variation of the rLVF caused by differences in states,a smoothing regularization of the rLVF is *** train policies with safety guarantees under the worst disturbances of the observed states,an off-policy robust RL algorithm is *** proposed algorithm is applied to motion control tasks of an autonomous vehicle and a cartpole,which involve external disturbances and variations of the model parameters,*** experimental results demonstrate the effectiveness of the theoretical findings and the advantages of the proposed algorithm in terms of robustness and safety.

关键词： motion control reinforcement learning robustness stability

维普期刊数据库

Credit assignment for trained neural networks based on Koopman operator theory

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2024年第1期18卷 251-253页

作者： Zhen LIANG Changyuan ZHAO Wanwei LIU Bai XUE Wenjing YANG Zhengbin PANG institute for Quantum Information&State Key Laboratory of High Performance computing National University of Defense TechnologyChangsha 410000China institute of Software Chinese Academy of SciencesBeijing 100190China College of Computer Science and technology National University of Defense TechnologyChangsha 410000China

1 Introduction Artificial neural networks(ANNs,also NNs)have recently emerged as leading candidate models for deep learning,popularly used in various areas[1–3].Behind the enormous success,ANNs are generally with complicated structures,there being an intricate data flow through multiple linear or nonlinear components between the input layer and the output ***,it is pressing to evaluate how much a specific component contributes to the final output,termed the Credit Assignment Problem(CAP)[4]in this paper.

关键词： layer assignment operator

An improved elastoplastic model for rocks and application to cyclic loading and unloading triaxial compression tests

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Acta Mechanica Sinica 2024年第8期40卷 179-193页

作者： Qiuxin Gu Qiang Zhang Yapeng Li Peinan Wu Guilei Han State Key Laboratory for Geomechanics and Deep Underground Engineering School of Mechanics and Civil EngineeringChina University of Mining and TechnologyXuzhou221116China north china Engineering Investigation institute Shijiazhuang050021China

The recoverable strain of rock is completely classified as elastic strain in the conventional elastic-plastic theory,which often results in poor agreement between theoretical and experimental *** work proposes an improved elastoplastic model of rock materials considering the evolutions of crack deformation and elastic modulus to better characterize the nonlinear mechanical behavior of rock in the post-peak *** this model,the recoverable strain is assumed to be a combination of elastic and crack strain,and the constitutive relationship between crack strain and rock stress is *** on the proposed assumption,the evolutions of the mechanical parameters including strength parameters,elastic,plastic,and crack deformation parameters versus the plastic strain and confining stress were *** developed elastoplastic model was validated by comparing the theoretical values with the results of the triaxial cyclic loading and unloading *** theoretical calculation results show a good agreement with the laboratory test,which indicates that the improved elastoplastic model can effectively reflect the nonlinear mechanical behavior of the rock *** research results are expected to provide a valuable reference for further understanding the evolution of rock crack deformation.

关键词： Crack strain Elastic modulus deterioration Novel elastoplastic model Model validation

Unraveling the atomic interdiffusion mechanism of NiFe_(2_)O_(4)oxygen carriers during chemical looping CO_(2)conversion

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Carbon Energy 2024年第8期6卷 105-121页

作者： Da Song Yan Lin Shiwen Fang Yang Li Kun Zhao Xinfei Chen Zhen Huang Fang He Zengli Zhao Hongyu Huang Fanxing Li Guangzhou institute of Energy Conversion Chinese Academy of SciencesGuangzhouChina School of Energy Science and Engineering University of Science and Technology of ChinaHefeiChina College of Mechanical and Electrical Engineering Zhongkai University of Agriculture and EngineeringGuangzhouChina College of Chemistry and Bioengineering Guilin University of TechnologyGuilinChina north Carolina State University RaleighNorth CarolinaUSA

By employing metal oxides as oxygen carriers,chemical looping demonstrates its effectiveness in transferring oxygen between reduction and oxidation environments to partially oxidize fuels into syngas and convert CO_(2)into ***,NiFe_(2_)O_(4)oxygen carriers have demonstrated remarkable efficiency in chemical looping CO_(2)***,the intricate process of atomic migration and evolution within the internal structure of bimetallic oxygen carriers during continuous high‐temperature redox cycling remains ***,the lack of a fundamental understanding of the complex ionic migration and oxygen transfer associated with energy conversion processes hampers the design of high‐performance oxygen ***,in this study,we employed in situ characterization techniques and theoretical calculations to investigate the ion migration behavior and structural evolution in the bulk of NiFe_(2_)O_(4)oxygen carriers during H2 reduction and CO_(2)/lab air oxidation *** discovered that during the H2 reduction step,lattice oxygen rapidly migrates to vacancy layers to replenish consumed active oxygen species,while Ni leaches from the material and migrates to the *** the CO_(2)splitting step,Ni migrates toward the core of the bimetallic oxygen carrier,forming Fe–Ni *** the air oxidation step,Fe–Ni migrates outward,creating a hollow structure owing to the Kirkendall effect triggered by the swift transfer of lattice *** metal atom migration paths depend on the oxygen transfer *** discoveries highlight the significance of regulating the release–recovery rate of lattice oxygen to uphold the structures and reactivity of oxygen *** work offers a comprehensive understanding of the oxidation/reduction‐driven atomic interdiffusion behavior of bimetallic oxygen carriers.

关键词： chemical looping CO_(2)splitting density functional theory in situ characterization ionic migration