检索结果-内蒙古大学图书馆

7th International Conference Intelligent computation Technology Automation (ICICTA)

作者： Huang Xin-cheng Xiao Ai-ling Tarim Univ Coll Mech & Elect Engn Alar 843300 Peoples R China

ISBN: (纸本)9781479966363

With the rapid development and popularization of CFD technology, requirements on data amount and computing speed have also increased a lot. The computing ability of single CPU is unable to meet the requirement. Therefore, iteration and parallel computation are the most important methods for CFD problems with large grid scales. This paper introduces the most popular iteration methods, classical iteration methods and their parallel computation strategies. On the basis of it, the parallel computing simulation environment is also constructed. It comprehensively applies symmetric multi-processing and distributed technologies and Windows XP 64 bit system. Research on CFD has focus on the parallel computation, which can meet the practical demand for computer capacity and computing time.

关键词： computation Fluid Dynamics Classical Iteration parallel computation

来源：评论

学校读者我要写书评

暂无评论

An efficient approach for the equivalent linearization of frame structures with plastic hinges under nonstationary seismic excitations

引用

Earthquake Engineering and Engineering Vibration 2024年第3期23卷 677-690页

作者： Huang Huan Li Yingxiong Li Yuyu College of Water Conservancy and Civil Engineering South China Agricultural UniversityGuangzhou 510642China Institute of Smart City and Intelligent Transportation Southwest Jiaotong UniversityChengdu 611756China

An efficient approach is proposed for the equivalent linearization of frame structures with plastic hinges under nonstationary seismic *** concentrated plastic hinges,described by the Bouc-Wen model,are assumed to occur at the two ends of a linear-elastic beam *** auxiliary differential equations governing the plastic rotational displacements and their corresponding hysteretic displacements are replaced with linearized differential ***,the two sets of equations of motion for the original nonlinear system can be reduced to an expanded-order equivalent linearized equation of motion for equivalent linear *** solve the equation of motion for equivalent linear systems,the nonstationary random vibration analysis is carried out based on the explicit time-domain method with high ***,the proposed treatment method for initial values of equivalent parameters is investigated in conjunction with parallel computing technology,which provides a new way of obtaining the equivalent linear systems at different time *** on the explicit time-domain method,the key responses of interest of the converged equivalent linear system can be calculated through dimension reduction analysis with high *** examples indicate that the proposed approach has high computational efficiency,and shows good applicability to weak nonlinear and medium-intensity nonlinear systems.

关键词： nonstationary random vibration plastic hinge equivalent linearization method explicit time-domain method parallel computation

来源：评论

学校读者我要写书评

暂无评论

Integration of work exchange network within distillation process optimization

引用

SEPARATION AND PURIFICATION TECHNOLOGY 2025年 362卷

作者： Lan, Yang Tian, Chenshan Lyu, Hao Sun, Jinsheng Tianjin Univ Sch Chem Engn & Technol Tianjin 300072 Peoples R China Hong Kong Univ Sci & Technol Guangzhou Sustainable Energy & Environm Thrust Guangzhou 511400 Peoples R China

Process integration is a well-established approach within process intensification (PI) strategies, aimed at increasing process efficiency to support more sustainable designs, for example, using heat integration (HI) and work integration (WI) to lower the thermal and electrical energy demands of distillation processes. While HI is widely utilized, WI remains less explored. In WI studies, mathematical programming is commonly employed to optimize the work exchange network (WEN). Although these methods often assume ideal gas behavior or overlook fluid phase changes. Consequently, they fail to address the complexities inherent in real industrial processes adequately. To find a solution, this paper turns to an improved, simulation-based WEN method embedded within a population-distributed differential evolution (DDE) algorithm. This method facilitates the simultaneous optimization of both design parameters and WI, which is more in line with on-site industrial applications. The effectiveness of this approach is demonstrated through a case study on methanol distillation, which is then extended to a cryogenic distillation process for capturing H2S and CO2 from the integrated gasification combined cycle (IGCC) syngas.

关键词： Work exchange network Process intensification Process optimization Process integration parallel computation Capture H 2 S and CO 2

来源：评论

学校读者我要写书评

暂无评论

A nonoverlapping domain decomposition method for extreme learning machines: Elliptic problems

引用

COMPUTERS & MATHEMATICS WITH APPLICATIONS 2025年 189卷 109-128页

作者： Lee, Chang-Ock Lee, Youngkyu Ryoo, Byungeun Korea Adv Inst Sci & Technol Dept Math Sci Daejeon 34141 South Korea Brown Univ Div Appl Math Providence RI 02912 USA

Extreme learning machine (ELM) is a methodology for solving partial differential equations (PDEs) using a single hidden layer feed-forward neural network. It presets the weight/bias coefficients in the hidden layer with random values, which remain fixed throughout the computation, and uses a linear least squares method for training the parameters of the output layer of the neural network. It is known to be much faster than Physics informed neural networks. However, classical ELM is still computationally expensive when a high level of representation is desired in the solution as this requires solving a large least squares system. In this paper, we propose a nonoverlapping domain decomposition method (DDM) for ELMs that not only reduces the training time of ELMs, but is also suitable for parallel computation. We introduce local neural networks, which are valid only at corresponding subdomains, and an auxiliary variable at the interface. We construct a system on the variable and the parameters of local neural networks. A Schur complement system on the interface can be derived by eliminating the parameters of the output layer. The auxiliary variable is then directly obtained by solving the reduced system after which the parameters for each local neural network are solved in parallel. A method for initializing the hidden layer parameters suitable for high approximation quality in large systems is also proposed. Numerical results that verify the acceleration performance of the proposed method with respect to the number of subdomains are presented.

关键词： Extreme learning machine Nonoverlapping domain decomposition method Elliptic problem parallel computation Weight initialization

来源：评论

学校读者我要写书评

暂无评论

Structural partitioning for parallel construction of geometric nonlinear reduced-order models

引用

INTERNATIONAL JOURNAL OF NON-LINEAR MECHANICS 2025年 175卷

作者： Bui, Tuan Anh Park, Junyoung Kim, Jun-Sik Kumoh Natl Inst Technol Dept Aeronaut Mech & Elect Convergence Engn 61 Daehak Ro Gumi 39177 Gyeongbuk South Korea Kumoh Natl Inst Technol Sch Mech Syst Engn 61 Daehak Ro Gumi 39177 Gyeongbuk South Korea

High-fidelity finite element models are widely used to predict the mechanical behavior of structures with complex geometries. While these models provide accurate results, they often require significant computational time, particularly when predicting nonlinear dynamic behavior. To address this, model order reduction techniques have been developed to reduce the computational time. However, constructing non-intrusive reducedorder models from high-fidelity finite element models still requires considerable computational time. This paper proposes a fully parallel process to accelerate the construction of geometrically nonlinear reduced-order models. In this approach, the structure is divided into multiple partitions, each assigned to a separate processor. The reduction basis for each partition ensures displacement consistency at partition interfaces without requiring additional modal coordinates at these interfaces. The parallel process operates without inter-processor communication, making it robust and straightforward to implement. It is compatible with various nonintrusive model order reduction techniques and achieves high computational efficiency. Notably, this approach introduces no additional errors, i.e., the reduced-order model constructed through the parallel process is identical to that obtained via traditional methods.

关键词： parallel computation Model order reduction Geometric nonlinearity Structural partition Non-intrusive reduction

来源：评论

学校读者我要写书评

暂无评论

Efficient GPU Cloth Simulation with Non-distance Barriers and Subspace Reuse

引用

ACM TRANSACTIONS ON GRAPHICS 2024年第6期43卷 1-16页

作者： Lan, Lei Lu, Zixuan Long, Jingyi Yuan, Chun Li, Xuan He, Xiaowei Wang, Huamin Jiang, Chenfanfu Yang, Yin Univ Utah Salt Lake City UT 84112 USA Univ Calif Los Angeles Los Angeles CA USA Chinese Acad Sci Inst Software Beijing Peoples R China Style3D Res Shanghai Peoples R China

This paper pushes the performance of cloth simulation, making the simulation interactive even for high-resolution garment models while keeping every triangle untangled. The penetration-free guarantee is inspired by the interior point method, which converts the inequality constraints to barrier potentials. We propose a major overhaul of this modality within the projective dynamics framework by leveraging an adaptive weighting mechanism inspired by barrier formulation. This approach does not depend on the distance between mesh primitives, but on the virtual life span of a collision event and thus keeps all the vertices within feasible region. Such a non-distance barrier model allows a new way to integrate collision resolution into the simulation pipeline. Another contributor to the performance boost comes from the subspace reuse strategy. This is based on the observation that low-frequency strain propagation is near orthogonal to the deformation induced by collisions or self-collisions, often of high frequency. Subspace reuse then takes care of low-frequency residuals, while high-frequency residuals can also be effectively smoothed by GPU-based iterative solvers. We show that our method outperforms existing fast cloth simulators by at least one order while producing high-quality animations of high-resolution models.

关键词： GPU simulation cloth animation collision detection parallel computation

来源：评论

学校读者我要写书评

暂无评论

Explicit Frechet Derivatives for 3-D Frequency-Domain Seismic Full-Waveform Inversion in Viscoelastic Tilted Transversely Isotropic Media and Fully parallel Implementations

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2025年 63卷

作者： Ma, Guoqi Zhou, Bing Al Suwaidi, Aisha Riahi, Mohamed Kamel Zemerly, Mohamed Jamal Liu, Xu Khalifa Univ Sci & Technol Dept Earth Sci Abu Dhabi U Arab Emirates Khalifa Univ Sci & Technol Dept Math Abu Dhabi U Arab Emirates Khalifa Univ Sci & Technol Dept Comp Sci Abu Dhabi U Arab Emirates King Fahd Univ Petr & Minerals Coll Petr Engn & Geosci Dhahran 31261 Saudi Arabia

The Fr & eacute;chet derivatives or sensitivity kernels of the observed seismograms are fundamental to seismic full-waveform inversion (FWI). They quantitatively measure the seismogram variations caused by any physical parameter perturbation of the Earth's subsurface. The 3-D viscoelastic tilted transversely isotropic (TTI) media are often encountered in practices due to the presence of dip thin layers, joints, fractures or cracks, orientated grains or crystallization, and water or gas saturation. To image such subsurface, we have derived explicit 3-D frequency-domain Fr & eacute;chet derivatives of the seismogram spectrum with respect to 13 independent physical parameters of TTI rock, which include density, five elastic moduli, five Q-factors, and inclination and declination angles of the symmetric axis of rock structure. We have demonstrated a fully parallel implementation to compute the Fr & eacute;chet derivatives and conduct synthetic subsurface imaging experiments, in which the 13 independent parameters of the subsurface targets are successfully reconstructed. The experimental results have verified the correctness and validity of the derived 3-D Fr & eacute;chet derivatives for imaging viscoelastic TTI media.

关键词： 3-D Fr & eacute chet derivatives frequency domain full-waveform inversion (FWI) full-waveform inversion (FWI) parallel computation parallel computation tilted transverse isotropy tilted transverse isotropy viscoelasticity viscoelasticity viscoelasticity

来源：评论

学校读者我要写书评

暂无评论

A novel fully explicit GPU-accelerated discrete element method coupled with moving particle hydrodynamics (DEM-MPH) for simulating the underwater particle sedimentation behavior

引用

PROGRESS IN NUCLEAR ENERGY 2025年 180卷

作者： Yokoyama, Ryo Xu, Yihua Miwa, Shuichiro Okamoto, Koji Univ Tokyo Nucl Profess Sch 2-22 Shirane Shirakata Tokai mura Ibaraki 3191188 Japan Univ Tokyo Dept Nucl Engn & Management 7-3-1 Hongo Bunkyo ku Tokyo 1138654 Japan

Ensuring the safety integrity of the core catcher system during a hypothetical core disruptive accident (CDA) in a sodium fast reactor (SFR) is crucial, particularly regarding the formation of the debris bed. This study presents a novel, fully explicit Discrete Element Method (DEM) coupled with Moving Particle Hydrodynamics (MPH), accelerated on a graphics processing unit (GPU), to calculate debris bed formation with various particle sizes. The DEM approach is applied to the solid phase, while the liquid phase is calculated using the MPH method. The DEM-MPH method is explicitly updated to avoid significant differences in time steps and to efficiently accelerate calculations on the GPU. For validation, a fundamental dam break simulation with tiny particle sizes (0.5 mm) was conducted. The simulation results show good agreement with experimental data in terms of the angle of repose and spread length. Additionally, the GPU-accelerated calculations demonstrated a calculation efficiency of several hundred times improvement compared to a single CPU. The DEM-MPH method was also compared to a previous experiment conducted by Sun Yat-Sen University, demonstrating the capability of handling almost 10 million particles. The results indicate that particle size significantly impacts debris bed formation. As particle size decreases, the debris bed shape becomes flatter due to the more significant convection within the liquid pool and the drag force between the solid phase and the liquid. All calculations showed agreement with experimental results regarding debris bed height. Overall, the present DEM-MPH method shows great potential for understanding debris bed formation, which is indispensable for ensuring SFR safety.

关键词： computational fluid dynamics Particle method Discrete element method parallel computation Debris bed Core disruptive accident Core catcher In-vessel retention Sodium fast reactor

来源：评论

学校读者我要写书评

暂无评论

Concise semi-analytical solution for group piles impedance function and generalized group factor analysis considering SSI effect

引用

ENGINEERING ANALYSIS WITH BOUNDARY ELEMENTS 2025年 176卷

作者： Hou, Hongwei Li, Jianbo Li, Zhiyuan Lin, Gao Dalian Univ Technol State Key Lab Coast & Offshore Engn Dalian 116024 Peoples R China Dalian Univ Technol Inst Earthquake Engn Dalian 116024 Peoples R China China Inst Water Resources & Hydropower Res Earthquake Engn Res Ctr Beijing 100038 Peoples R China

It is foremost to investigate group piles dynamic response for large nuclear island structure considering highprecision soil structure interaction (SSI). However, the existing group piles effect obtained using static stiffness cannot reflect to the frequency dependence, and the spatial coupling characteristics of far-field artificial boundary limits the efficient and fine impedance function computation for large piles under complex foundation conditions. In present paper, a refined impedance function procedure for large-scale group piles is built on the scaled boundary finite element method (SBFEM). The far-field stiffness matrix is made to localize the spatial coupling using decoupling technique, a Runge-Kutta search strategy on frequency with zero-frequency starting is developed to obtain the complete stiffness matrix, and a generalized group factor calculation is proposed. The accuracy and effectiveness are proved with the cases of impedance function for small-scale group piles. At length, some fine large-scale nuclear island group piles model is investigated. All cases demonstrate that the pile impedance function, especially under the condition of layered ground, presents very obvious frequency correlations in the simulation of large-scale nuclear island group piles, which further confirms that the provided method has good engineering applicability.

关键词： Group factor Impedance function Partitioned decoupling Nuclear island structure Modified SBFEM parallel computation

来源：评论

学校读者我要写书评

暂无评论

High-quality computer-generated holography based on Vision Mamba

引用

OPTICS AND LASERS IN ENGINEERING 2025年 184卷

作者： Yang, Lei Xu, Shengyuan Yang, Chunzheng Chang, Chenliang Hou, Qichao Song, Qiang Jiangsu Univ Sch Mech Engn Zhenjiang 212013 Peoples R China Univ Shanghai Sci & Technol Sch Opt Elect & Comp Engn Shanghai 200093 Peoples R China Hunan Univ Greater Bay Area Inst Innovat Guangzhou 511300 Peoples R China

Deep learning, especially through model-driven unsupervised networks, offers a novel approach for efficient computer-generated hologram (CGH) generation. However, current model-driven CGH generation models are primarily built on the convolutional neural networks (CNNs), which struggle to achieve high-quality hologram reconstruction due to limited receptive fields. Although Vision Transformers (ViTs) excel at processing more distant visual information, they are burdened with huge computational load. The recent emergence of Vision Mamba (ViM) presents a promising avenue to address these challenges. In this study, we introduce the CVMNet, a lightweight model that combines the precision of convolutional layers for local feature extraction and the longrange modeling abilities of state-space models (SSMs) to enhance the quality of CGHs. By employing parallel computation for the ViM to handle feature channels, the CVMNet effectively reduces the number of model parameters. Numerical reconstruction and optical experiments demonstrate that the CVMNet can generate 1080P high-quality holograms in just 16 ms, boosting an average PSNR of over 30 dB and effectively suppressing speckle noise in reconstructed images. Additionally, the CVMNet showcases robust generalization capabilities.

关键词： Computer-generated holography Deep learning Vision mamba parallel computation Long-range modeling

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：