检索结果-内蒙古大学图书馆

SSRN 2023年

作者： Zhang, Jianzhong Cao, Hongmei Department of Mathematics Nanjing University of Aeronautics and Astronautics Nanjing211106 China MIIT Key Laboratory of Mathematical Modelling and High Performance Computing of Air Vehicles Nanjing211106 China

In this paper, we investigate the Cauchy problem of compressible Euler-Korteweg equations with damping. The global well-posedness is established in L2-Lp critical Besov spaces. In our results, the existence theorem provides us with bounds that are independent of the relaxation parameter Ε and capillary coefficient κ. As a consequence, we rigorously justify the relaxation limit and study the effect of the Korteweg-type dispersion on the relaxation limit. Specially, when κ=0, our theorems reduce to the results in [7,8] on the Euler system with damping, and the smallness assumption for low-frequency initial data of velocity is removed in the sense of limit. © 2023, The Authors. All rights reserved.

关键词： Damping

来源：评论

学校读者我要写书评

暂无评论

Reducibility of linear quasi-periodic Hamiltonian derivative wave equations and half-wave equations under the Brjuno conditions

arXiv

引用

arXiv 2023年

作者： Lou, Zhaowei School of Mathematics Nanjing University of Aeronautics and Astronautics Nanjing211106 China MIIT Key Laboratory of Mathematical Modelling and High Performance Computing of Air Vehicles Nanjing211106 China

In this paper, we prove the reducibility for some linear quasi−periodic Hamiltonian derivative wave and half−wave equations under the Brjuno−Rüssmann non−resonance conditions. This generalizes KAM theory by Pöschel in [38] from the finite dimensional Hamiltonian systems to Hamiltonian *** Codes 37K55, 35L05, 35Q55 © 2023, CC BY-NC-SA.

关键词： Wave equations

来源：评论

学校读者我要写书评

暂无评论

SelectiveEC: Selective reconstruction in erasure-coded storage systems 12

SelectiveEC: Selective reconstruction in erasure-coded stora...

引用

12th USENIX Workshop on Hot Topics in Storage and File Systems, HotStorage 2020, co-located withUSENIX ATC 2020

作者： Xu, Liangliang Lyu, Min Li, Qiliang Xie, Lingjiang Xu, Yinlong University of Science and Technology of China Anhui Province Key Laboratory of High Performance Computing USTC China

Erasure coding has been a commonly used approach to provide high reliability with low storage cost. But the skewed load in a recovery batch severely slows down the failure recovery process in storage systems. To this end, we propose a balanced scheduling module, SelectiveEC, which schedules reconstruction tasks out of order by dynamically selecting some stripes to be reconstructed into a batch and selecting source nodes and replacement nodes for each reconstruction task. So it achieves balanced network recovery traffic, computing resources and disk I/Os against single node failure in erasure-coded storage systems. Compared with conventional random reconstruction, SelectiveEC increases the parallelism of recovery process up to 106% and averagely bigger than 97% in our simulation. Therefore, SelectiveEC not only speeds up recovery process, but also reduces the interference of failure recovery on the front-end applications. © HotStorage 2020 - 12th USENIX Workshop on Hot Topics in Storage and File Systems, co-located with USENIX ATC 2020. All rights reserved.

关键词： Recovery

来源：评论

学校读者我要写书评

暂无评论

Insights into the nucleation,grain growth and phase transformation behaviours of sputtered metastableβ-W films

引用

Journal of Materials Science & technology 2021年第31期90卷 66-75页

作者： Shuqun Chen Jinshu Wang Ronghai Wu Zheng Wang Yangzhong Li Yiwen Lu Wenyuan Zhou Peng Hua Hongyi Lia Key Laboratory of Advanced Functional Materials Education Ministry of ChinaFaculty of Materials and ManufacturingBeijing University of TechnologyBeijing 100124China School of Mechanics Civil Engineering and ArchitectureNorthwestern Polytechnical UniversityXi’an 710129China High Performance Computing Department National Supercomputing Center in ShenzhenShenzhen 518055China China Ship Development and Design Center Wuhan 430074China

Metastable phase in tungsten film is of great interests in recent years due to its giant spin Hall effects,however,little information has been known on its nucleation,growth and phase *** this paper,a 900 nm-thick tungsten film with double-layer structure(α-W underlayer andβ-W above it)was produced on SiO_(2)/Si substrate by high vacuum magnetron sputtering at room *** structural properties ofβ-W were systemically investigated by X-ray diffraction,transmission electron microscopy,thermodynamic calculation,first-principle and phase-field *** is found that theβ-W nucleation is energetically favoured on the SiO_(2)surface compared to theα-W *** the film thickening proceeds,β-W[211]turns to be preferred direction of growth owing to the elastic strain energy minimization,which is verified by phase-field ***,theβ→αphase transformation takes place near the film-substrate interface while the rest of the film keeps theβ-W phase,leading to a doublelayer *** localized phase transition is induced by lower Gibbs free energy ofα-W phase at larger grain sizes,which can be confirmed by thermodynamic *** in-situ heating TEM analysis of the as-deposited film reveals that theβ→αphase transformation is fulfilled byα/βinterface propagation rather than local atomic *** findings offer valuable insights into the intrinsic properties of metastable phase in tungsten.

关键词： Metastable phases Tungsten Nucleation and growth Phase field modelling Phase transformation

来源：评论

学校读者我要写书评

暂无评论

Decision Mamba: Reinforcement Learning via Hybrid Selective Sequence Modeling

arXiv

引用

arXiv 2024年

作者： Huang, Sili Hu, Jifeng Yang, Zhejian Yang, Liwei Luo, Tao Chen, Hechang Sun, Lichao Yang, Bo Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education China School of Artificial Intelligence Jilin University China Institute of High Performance Computing A*STAR Singapore Lehigh University BethlehemPA United States

Recent works have shown the remarkable superiority of transformer models in reinforcement learning (RL), where the decision-making problem is formulated as sequential generation. Transformer-based agents could emerge with self-improvement in online environments by providing task contexts, such as multiple trajectories, called in-context RL. However, due to the quadratic computation complexity of attention in transformers, current in-context RL methods suffer from huge computational costs as the task horizon increases. In contrast, the Mamba model is renowned for its efficient ability to process long-term dependencies, which provides an opportunity for in-context RL to solve tasks that require long-term memory. To this end, we first implement Decision Mamba (DM) by replacing the backbone of Decision Transformer (DT). Then, we propose a Decision Mamba-Hybrid (DM-H) with the merits of transformers and Mamba in high-quality prediction and long-term memory. Specifically, DM-H first generates high-value sub-goals from long-term memory through the Mamba model. Then, we use sub-goals to prompt the transformer, establishing high-quality predictions. Experimental results demonstrate that DM-H achieves state-of-the-art in long and short-term tasks, such as D4RL, Grid World, and Tmaze benchmarks. Regarding efficiency, the online testing of DM-H in the long-term task is 28× times faster than the transformer-based baselines. Copyright © 2024, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Adaptive OrderWENO Reconstructions for the Semi-Lagrangian Finite Difference Scheme for Advection Problem

引用

Communications in Computational Physics 2021年第6期30卷 67-96页

作者： Jiajie Chen Xiaofeng Cai Jianxian Qiu Jing-Mei Qiu Department of Mathematical Sciences University of DelawareNewarkDE19717USA School of Mathematical Sciences and Fujian Provincial Key Laboratory of Mathematical Modeling and High-Performance Scientific Computing Xiamen UniversityXiamenFujian 361005P.R.China

We present a new conservative semi-Lagrangian finite difference weighted essentially non-oscillatory scheme with adaptive *** is an extension of the conservative semi-Lagrangian(SL)finite difference WENO scheme in[Qiu and Shu,JCP,230(4)(2011),pp.863-889],in which linear weights in SL WENO framework were shown to not exist for variable coefficient ***,the order of accuracy is not optimal from reconstruction *** this paper,we incorporate a recent WENO adaptive order(AO)technique[Balsara et al.,JCP,326(2016),pp.780-804]to the SL WENO *** new scheme can achieve an optimal high order of accuracy,while maintaining the properties of mass conservation and non-oscillatory capture of solutions from the original SL *** positivity-preserving limiter is further applied to ensure the positivity of ***,the scheme is applied to high dimensional problems by a fourth-order dimensional *** demonstrate the effectiveness of the new scheme by extensive numerical tests on linear advection equations,the Vlasov-Poisson system,the guiding center Vlasov model as well as the incompressible Euler equations.

关键词： Semi-Lagrangian weighted essentially nonoscillatory WENO adaptive order reconstruction finite difference mass conservation Vlasov-Poisson incompressible Euler

来源：评论

学校读者我要写书评

暂无评论

Ispliter: An Intelligent and Automatic Surface Mesh Generator Using Neural Networks and Splitting Lines

SSRN

引用

SSRN 2022年

作者： Liu, Zengsheng Chen, Shizhao Gao, Xiang Zhang, Xiang Gong, Chunye Xu, Chuanfu Liu, Jie State Key Laboratory of High Performance Computing National University of Defense Technology Changsha410073 China College of Computer National University of Defense Technology Changsha410073 China

In this paper, we present a novel surface mesh generation approach based on neural networks and splitting lines that splits B-rep geometry models into isotropic triangular meshes. In the first stage, a recursive method is designed to generate plentiful data to train the neural network model offline. In the second stage, the implemented mesh generator, ISpliter, maps each surface patch into the parameter plane, and then the trained neural network model is applied to select the optimal splitting line to divide the patch into subdomains continuously until they are all triangles. In the third stage, ISpliter remaps the 2D mesh back to the physical space and optimize it. Several typical cases are compared to demonstrate the mesh quality generated by ISpliter and two baselines, Gmsh and NNW-GridStar. The results show that ISpliter can generate isotropic triangular meshes with high average quality, and the generated meshes are not inferior to those generated by the other two software under the same configuration. © 2022, The Authors. All rights reserved.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Render unto Numerics: Orthogonal Polynomial Neural Operator for PDEs with Non-periodic Boundary Conditions

arXiv

引用

arXiv 2022年

作者： Ziyuan, Liu Haifeng, Wang Kaijun, Bao Xu, Qian Hong, Zhang Songhe, Song College of Science National University of Defense Technology Changsha410073 China State Key Laboratory of High Performance Computing National University of Defense Technology Changsha410073 China

By learning the mappings between infinite function spaces using carefully designed neural networks, the operator learning methodology has exhibited significantly more efficiency than traditional methods in solving complex problems such as differential equations, but faces concerns about their accuracy and reliability. To overcomes these limitations, combined with the structures of the spectral numerical method, a general neural architecture named spectral operator learning (SOL) is introduced, and one variant called the orthogonal polynomial neural operator (OPNO), developed for PDEs with Dirichlet, Neumann and Robin boundary conditions (BCs), is proposed later. The strict BC satisfaction properties and the universal approximation capacity of the OPNO are theoretically proven. A variety of numerical experiments with physical backgrounds show that the OPNO outperforms other existing deep learning methodologies, as well as the traditional 2nd-order finite difference method (FDM) with a considerably fine mesh (with the relative errors reaching the order of 10−6), and is up to almost 5 magnitudes faster than the traditional method. Copyright © 2022, The Authors. All rights reserved.

关键词： Numerical methods

来源：评论

学校读者我要写书评

暂无评论

New Insights into the Microstructure and Tribological Behaviors of the Boronized Layer on Inconel 718 Alloy Surface

SSRN

引用

SSRN 2024年

作者： Wu, Zhenjiang Chen, Shuqun Shao, Mengxue Liu, Yue Li, Yangzhong Peng, Jian Li, Hongyi Wang, Jinshu Key Laboratory of Advanced Functional Materials Education Ministry of China College of Materials and Engineering Beijing University of Technology Beijing100124 China Beijing University of Technology China High Performance Computing Department National Supercomputing Center in Shenzhen Guangdong Shenzhen518055 China State Key Lab of Advanced Technology for Materials Synthesis and Processing Wuhan University of Technology Wuhan430070 China

Boronizing of nickel based superalloys is of great interests in aerospace industry due to the excellent wear resistance, however, little information has been known on its phase formation and load-dependent wear mechanism. In this work, high quality boronized layer was developed on the Inconel 718 alloy by powder pack-boriding process at 850 °C for 6 h under argon atmosphere. The microstructure characteristics and load-dependent tribological behaviors of the boronized layers were systematically investigate. It is found that the main boronized layer on Inconel 718 alloy consists of a ~11.3 um thick compound layer with nanocrystalline (Ni,Fe)23B6 and (Cr,Fe)B grains, and a ~10.9 um thick diffusion layer with dendritic Cr2B precipitation within the matrix. The phase evolution of the boronizing layer are explainable by thermodynamic modeling of Ni-Cr-B system and the phase stability for metastable Ni-B compounds is confirmed by DFT simulation. Due to the good interfacial bonding strength and high microhardness value, the boronized layer displays excellent wear resistance against Si3N4 ceramic balls at room temperature. In particular, the wear rates of the specimens under applied loads of 10-20 N are higher than those in the 50-100 N loads case. The latter is mainly facilitated by the formation of lubricating boron oxide tribolayer on friction ball surface. We believe the findings in present work will contribute to a deeper understanding of the boronizing treatment in nickel-based superalloys. © 2024, The Authors. All rights reserved.

关键词： Wear resistance

来源：评论

学校读者我要写书评

暂无评论

Ispliter: An Intelligent and Automatic Surface Mesh Generator Using Neural Networks and Splitting Lines

SSRN

引用

SSRN 2022年

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：