检索结果-南通市图书馆

A robust implicit high-order discontinuous Galerkin method for solving compressible Navier-Stokes equations on arbitrary grids

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 96-119页

作者： jia Yan Xiaoquan Yang Peifen Weng Shanghai Institute of Applied Mathematics and Mechanics Shanghai Key Laboratory of Mechanics in Energy EngineeringSchool of Mechanics and Engineering ScienceShanghai UniversityShanghai200072China College of Energy and Mechanical Engineering Shanghai University of Electric PowerShanghai200090China

The primary impediments impeding the implementation of high-order methods in simulating viscous flow over complex configurations are robustness and *** challenges impose significant constraints on computational efficiency,particularly in the domain of engineering *** address these concerns,this paper proposes a robust implicit high-order discontinuous Galerkin(DG)method for solving compressible Navier-Stokes(NS)equations on arbitrary *** method achieves a favorable equilibrium between computational stability and *** solve the linear system,an exact Jacobian matrix solving strategy is employed for preconditioning and matrix-vector generation in the generalized minimal residual(GMRES)*** approach mitigates numerical errors in Jacobian solution during implicit calculations and facilitates the implementation of an adaptive Courant-Friedrichs-Lewy(CFL)number increasing strategy,with the aim of improving convergence and *** further enhance the applicability of the proposed method for intricate grid distortions,all simulations are performed in the reference *** practice significantly improves the reversibility of the mass matrix in implicit calculations.A comprehensive analysis of various parameters influencing computational stability and efficiency is conducted,including CFL number,Krylov subspace size,and GMRES convergence *** computed results from a series of numerical test cases demonstrate the promising results achieved by combining the DG method,GMRES solver,exact Jacobian matrix,adaptive CFL number,and reference domain calculations in terms of robustness,convergence,and *** analysis results can serve as a reference for implicit computation in high-order calculations.

关键词： Discontinuous Galerkin method Exact Jacobian matrix GMRES solver Adaptive CFL number Reference domain High-order

Optimal V_(2)O_(3) extraction by sustainable vanadate electrolysis in molten salts

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Tungsten 2024年第1期6卷 141-149页

作者： Bao-Yan Feng Zhong-Hua Zhao Yong-Zheng jia jia-Liang An Ming-Yong Wang State Key Laboratory of Advanced Metallurgy University of Science and Technology BeijingBeijing 100083China Beijing Key Laboratory of Green Recovery and Extraction of Rare and Precious Metals University of Science and Technology BeijingBeijing 100083China

Recently,an environmentally friendly electrolysis process of soluble vanadates is proposed and successfully confirmed for V_(2)O_(3) *** this paper,the solubilities of various vanadates(***_(3),Na_(4)V_(2)O_(7) and Na_(3)VO_(4))in NaCl molten salt are *** dependences of V_(2)O_(3) extraction on vanadate form,anode material and molten salt system are studied.A long-term electrolysis is carried *** result indicates that the solubilities of all vanadates are high and meet electrolytic *** to Na_(4)V_(2)O_(7) and Na_(3)VO_(4),NaVO_(3) exhibits larger current efficiency and lower electricity *** using SnO_(2) anode instead of graphite anode,the current efficiency of NaVO_(3) electrolysis can be increased to 58.1%and carbon pollution is avoided at 700℃.The extraction rate of V_(2)O_(3) reaches 81.3%after long-term electrolysis of 51 *** work optimizes the electrochemical preparation process of V_(2)O_(3) from various vanadates and contributes to the improvement of current efficiency and extraction rate.

关键词： Molten salt electrolysis Vanadate Vanadium trioxide Current efficiency

Periapsis shift in spherically symmetric spacetimes and effects of electric interactions

在线全文

学校读者我要写书评

暂无评论

Chinese Physics C 2024年第8期48卷 224-239页

作者： Qianchuan Wang Junji jia School of Physics and Technology Wuhan UniversityWuhan 430072China Department of Astronomy&MOE Key Laboratory of Artificial Micro-and Nano-structures School of Physics and TechnologyWuhan UniversityWuhan 430072China

The periapsis shift of charged test particles in arbitrary static and spherically symmetric charged spacetimes are *** perturbative methods,the near-circular approximation and post-Newtonian methods,are developed and shown to be very accurate when the results are determined to high *** near-circular approximation method is more precise when eccentricity e of the orbit is small,whereas the post-Newtonian method is more effective when orbit semilatus rectum p is *** from these two methods are shown to agree when both e is small and p is *** results are then applied to the Reissner-Nordström spacetime,the Einstein-Maxwell-dilation gravity,and a charged wormhole *** effects of various parameters on the periapsis shift,particularly that of the electrostatic interaction,are carefully *** periapsis shift data of the solar-Mercury are then used to constrain the charges of the Sun and Mercury,and the data of the Sgr A^(*)-S2 periapsis shift are used to determine,for the first time using this method,the constraints of the charges of Sgr A^(*)and S2.

关键词： periapsis shift orbital precession post-Newtonian approximation perturbative method

A dynamical neural network approach for distributionally robust chance-constrained Markov decision process

在线全文

学校读者我要写书评

暂无评论

Science China Mathematics 2024年第6期67卷 1395-1418页

作者： Tian Xia jia Liu Zhiping Chen School of Mathematics and Statistics Xi’an Jiaotong UniversityXi’an710049China

In this paper,we study the distributionally robust joint chance-constrained Markov decision *** the logarithmic transformation technique,we derive its deterministic reformulation with bi-convex terms under the moment-based uncertainty *** cope with the non-convexity and improve the robustness of the solution,we propose a dynamical neural network approach to solve the reformulated optimization *** results on a machine replacement problem demonstrate the efficiency of the proposed dynamical neural network approach when compared with the sequential convex approximation approach.

关键词： Markov decision process chance constraints distributionally robust optimization moment-based ambiguity set dynamical neural network

Enhancing Image Description Generation through Deep Reinforcement Learning:Fusing Multiple Visual Features and Reward Mechanisms

在线全文

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第2期78卷 2469-2489页

作者： Yan Li Qiyuan Wang Kaidi jia School of Cyber Security Gansu University of Political Science and LawLanzhou730070China

Image description task is the intersection of computer vision and natural language processing,and it has important prospects,including helping computers understand images and obtaining information for the visually *** study presents an innovative approach employing deep reinforcement learning to enhance the accuracy of natural language descriptions of *** method focuses on refining the reward function in deep reinforcement learning,facilitating the generation of precise descriptions by aligning visual and textual features more *** approach comprises three key ***,it utilizes Residual Network 101(ResNet-101)and Faster Region-based Convolutional Neural Network(Faster R-CNN)to extract average and local image features,respectively,followed by the implementation of a dual attention mechanism for intricate feature ***,the Transformer model is engaged to derive contextual semantic features from textual ***,the generation of descriptive text is executed through a two-layer long short-term memory network(LSTM),directed by the value and reward *** with the image description method that relies on deep learning,the score of Bilingual Evaluation Understudy(BLEU-1)is 0.762,which is 1.6%higher,and the score of BLEU-4 is ***-based Image Description Evaluation(CIDEr)scored 0.998,Recall-Oriented Understudy for Gisting Evaluation(ROUGE)scored 0.552,the latter improved by 0.36%.These results not only attest to the viability of our approach but also highlight its superiority in the realm of image *** research can explore the integration of our method with other artificial intelligence(AI)domains,such as emotional AI,to create more nuanced and context-aware systems.

关键词： Image description deep reinforcement learning attention mechanism

A New Encrypted Traffic Identification Model Based on VAE-LSTM-DRN

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第1期78卷 569-588页

作者： Haizhen Wang Jinying Yan Na jia College of Computer and Control Engineering Qiqihar UniversityQiqihar161006China Heilongjiang Key Laboratory of Big Data Network Security Detection and Analysis Qiqihar UniversityQiqiharChina

Encrypted traffic identification pertains to the precise acquisition and categorization of data from traffic datasets containing imbalanced and obscured *** extraction of encrypted traffic attributes and their subsequent identification presents a formidable *** existing models have predominantly relied on direct extraction of encrypted traffic data from imbalanced datasets,with the dataset’s imbalance significantly affecting the model’s *** the present study,a new model,referred to as UD-VLD(Unbalanced Dataset-VAE-LSTM-DRN),was proposed to address above *** proposed model is an encrypted traffic identification model for handling unbalanced *** encoder of the variational autoencoder(VAE)is combined with the decoder and Long-short term Memory(LSTM)in UD-VLD model to realize the data enhancement processing of the original unbalanced *** enhanced data is processed by transforming the deep residual network(DRN)to address neural network gradient-related ***,the data is classified and *** UD-VLD model integrates the related techniques of deep learning into the encrypted traffic recognition technique,thereby solving the processing problem for unbalanced *** UD-VLD model was tested using the publicly available Tor dataset and VPN *** UD-VLD model is evaluated against other comparative models in terms of accuracy,loss rate,precision,recall,F1-score,total time,and ROC *** results reveal that the UD-VLD model exhibits better performance in both binary and multi classification,being higher than other encrypted traffic recognition models that exist for unbalanced ***,the evaluation performance indicates that the UD-VLD model effectivelymitigates the impact of unbalanced data on traffic *** can serve as a novel solution for encrypted traffic identification.

关键词： Data enhancement LSTM deep residual network VAE

Towards imbalanced motion:part-decoupling network for video portrait segmentation

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(Information Sciences) 2024年第7期67卷 197-210页

作者： Tianshu YU Changqun XIA jia LI State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBeihang University Peng Cheng Laboratory

Video portrait segmentation(VPS), aiming at segmenting prominent foreground portraits from video frames, has received much attention in recent years. However, the simplicity of existing VPS datasets leads to a limitation on extensive research of the task. In this work, we propose a new intricate large-scale multi-scene video portrait segmentation dataset MVPS consisting of 101 video clips in 7 scenario categories,in which 10843 sampled frames are finely annotated at the pixel level. The dataset has diverse scenes and complicated background environments, which is the most complex dataset in VPS to our best *** the observation of a large number of videos with portraits during dataset construction, we find that due to the joint structure of the human body, the motion of portraits is part-associated, which leads to the different parts being relatively independent in motion. That is, the motion of different parts of the portraits is imbalanced. Towards this imbalance, an intuitive and reasonable idea is that different motion states in portraits can be better exploited by decoupling the portraits into parts. To achieve this, we propose a part-decoupling network(PDNet) for VPS. Specifically, an inter-frame part-discriminated attention(IPDA)module is proposed which unsupervisedly segments portrait into parts and utilizes different attentiveness on discriminative features specified to each different part. In this way, appropriate attention can be imposed on portrait parts with imbalanced motion to extract part-discriminated correlations, so that the portraits can be segmented more accurately. Experimental results demonstrate that our method achieves leading performance with the comparison to state-of-the-art methods.

关键词： video portrait segmentation imbalanced motion unsupervised part decoupling motion correlation inter-frame attention

THE GLOBAL WELL-POSEDNESS OF SOLUTIONS TO COMPRESSIBLE ISENTROPIC TWO-FLUID MAGNETOHYDRODYNAMICS IN A STRIP DOMAIN

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Acta Mathematica Scientia 2024年第5期44卷 1997-2018页

作者： Zefu FENG Jing jia School of Mathematical Sciences Chongqing Normal UniversityChongqing401331China

In this paper,we consider a model of compressible isentropic two-fluid magneto-hydrodynamics without resistivity in a strip domain in three dimensional *** exploiting the two-tier energy method developed in[Anal PDE,2013,6:1429–1533],we prove the global well-posedness of the governing model around a uniform magnetic field which is non-parallel to the horizontal ***,we show that the solution converges to the steady state at an almost exponential rate as time goes to *** to the work of Tan and Wang[SIAM J Math Anal,2018,50:1432–1470],we need to overcome the difficulties caused by particles.

关键词： two-fluid magnetohydrodynamics global well-posedness compressiblility

Towards efficient generative AI and beyond-AI computing:New trends on ISSCC 2024 machine learning accelerators

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 Semiconductors 2024年第4期45卷 12-15页

作者： Bohan Yang jia Chen Fengbin Tu Department of Electronic and Computer Engineering The Hong Kong University of Science and TechnologyHong KongChina AI Chip Center for Emerging Smart Systems The Hong Kong University of Science and TechnologyHong KongChina School of the Gifted Young University of Science and Technology of ChinaHefei 230026Chin

Compared to the last decade when the convolution neu-ral network(CNN)dominated the research field,machine learn-ing(ML)algorithms have reached a pivotal moment called the generative artificial intelligence(AI)*** the emer-gence of large-scale foundation models[1],such as large multi-modal model(LMM)GPT-4[2]and text-to-image generative model DALL·E[3].

关键词： ISSCC beyond AI