检索结果-内蒙古大学图书馆

AIAA SciTech Forum

作者： Sperry, Mark Z. Kondap, Kavish Hwang, John T. Univ Calif San Diego Dept Mech & Aerosp Engn 9500 Gilman Dr La Jolla CA 92093 USA Francis Parker Sch 6501 Linda Vista Rd San Diego CA 92111 USA

ISBN: (数字)9781624107115

ISBN: (纸本)9781624107115

Multidisciplinary design optimization (MDO) models are built by assembling sub-models that represent varying disciplines and design conditions. Consequently, such models are often computationally expensive, requiring their evaluation in a parallel computing environment to decrease solution time. MDO modeling frameworks, which are software tools to allow construction of MDO models, generally require manual user input to specify where to parallelize their code. However, this process can be tedious if the model is large or complex. In this paper, we propose a method heavily inspired by scheduling theory that aims to fully automate parallelization of MDO models. We use static task scheduling to partition mathematical operations in the model to different processors prior to model execution. Additionally, our approach supports adjoint-based derivative computation to enable the use of gradient-based optimization algorithms. Our results show a significant performance improvement when applying our method to several MDO applications. We believe our approach enables MDO modelers to solve computationally expensive problems in a more efficient manner.

关键词： Multidisciplinary Design and Optimization parallel computing environment Optimization Algorithm Software Tool Uncertainty Quantification computing Vortex Lattice Method Blade Element Momentum Theory Asynchronous Communications Dynamic Scheduling Algorithm

来源：评论

学校读者我要写书评

暂无评论

Application of Multi-fidelity Methods to Rotorcraft Performance Assessment

Application of Multi-fidelity Methods to Rotorcraft Performa...

引用

AIAA SciTech Forum

作者： Walizer, Laura Haehnel, Robert Allen, Luke Wenren, Yonghu US Army Engn Res & Dev Ctr Geotech & Struct Lab AIAA Vicksburg MS 39180 USA US Army Engn Res & Dev Ctr Cold Regions Res & Engn Lab Hanover NH 03755 USA US Army Engn Res & Dev Ctr Cold Regions Res & Engn Lab AIAA Hanover NH 03755 USA

ISBN: (数字)9781624107115

ISBN: (纸本)9781624107115

We present a Python-based multi-fidelity tool to estimate rotorcraft performance metrics. We use Gaussian-Process regression (GPR) methods to adaptively build a surrogate model using a small number of high-fidelity CFD points to improve estimates of performance metrics from a medium-fidelity comprehensive analysis model. To include GPR methods in our framework, we used the EmuKit Python package. Our framework adaptively chooses new high-fidelity points to run in regions where the model variance is high. These high-fidelity points are used to update the GPR model;convergence is reached when model variance is below a pre-determined level. To efficiently use our framework on large computer clusters, we implemented this in Galaxy Simulation Builder, an analysis tool that is designed to work on large parallel computing environments. The program is modular, and is designed to be agnostic to the number and names of dependent variables and to the number and identifying labels of the fidelity levels. We demonstrate our multi-fidelity modeling framework on a rotorcraft collective sweep (hover) simulation and compare the accuracy and time savings of the GPR model to that of a simulation run with CFD only.

关键词： Rotorcrafts Gaussian Process Regression Uncertainty Quantification Python Galaxies Computational Fluid Dynamics Computer Cluster parallel computing environment Power Coefficient High Performance computing

来源：评论

学校读者我要写书评

暂无评论

A COMPARISON OF TWO parallel RANKING AND SELECTION PROCEDURES

A COMPARISON OF TWO PARALLEL RANKING AND SELECTION PROCEDURE...

引用

Winter Simulation Conference

作者： Ni, Eric C. Henderson, Shane G. Hunter, Susan R. Cornell Univ Operat Res & Informat Engn Ithaca NY 14853 USA Purdue Univ Sch Ind Engn W Lafayette IN 47907 USA

ISBN: (纸本)9781479974863

Traditional solutions to ranking and selection problems include two-stage procedures (e.g., the NSGS procedure of Nelson et al. 2001) and fully-sequential screening procedures (e.g., Kim and Nelson 2001 and Hong 2006). In a parallel computing environment, a naively-parallelized NSGS procedure may require more simulation replications than a sequential screening procedure such as that of Ni, Hunter, and Henderson (2013) (NHH), but requires less communication since there is no periodic screening. The parallel procedure NHH may require less simulation replications overall, but requires more communication to implement periodic screening. We numerically explore the trade-offs between these two procedures on a parallel computing platform. In particular, we discuss their statistical validity, efficiency, and implementation, including communication and load-balancing. Inspired by the comparison results, we propose a framework for hybrid procedures that may further reduce simulation cost or guarantee to select a good system when multiple systems are clustered near the best.

关键词： mathematics computing nonlinear programming parallel algorithms resource allocation NSGS procedure fully-sequential screening procedure load balancing parallel computing environment parallel ranking procedure parallel selection procedure sequential screening procedure Abstracts

来源：评论

学校读者我要写书评

暂无评论

Development of Common parallel Programming Platform for MPI and PVM

Development of Common Parallel Programming Platform for MPI ...

引用

International Conference on Artificial Intelligence and Evolutionary Algorithms in Engineering Systems (ICAEES)

作者： Swamy, Veerabasavantha Sampath, S. Nanjesh, B. R. Sagar, Bharat Bhushan Prodha Devaraya Inst Technol Dept CS & E Hospet Karnataka India Shri Venkateshwara Univ Dept CS & E Gajraula Uttar Pradesh India Adichunchanagiri Inst Technol Dept IS & E Chikmagalur Karnataka India Birla Inst Technol Dept IS & E Noida Uttar Pradesh India

ISBN: (纸本)9788132221357;9788132221340

parallel virtual machine (PVM) and message-passing interface (MPI) are the most successful message-passing libraries to map parallel algorithm onto parallel computing platform. Configuration of MPI and PVM in the nodes present in parallel computing environment (desktop PC's interconnected using ethernet LAN) is the time-consuming task for a user. This configuration procedure requires lot of knowledge about the steps to be followed to make them work properly. Configuration becomes the difficult task when there are more number of nodes in the parallel computing environment. Our work aims on developing a common parallel programming platform, which allows a user to get MPI and PVM in the nodes without any time-consuming configuration steps. This is done by integrating the recent version of PVM: PVM3.4.6 into the MPI's MPICH2 packages in order to simplify the time-consuming task of configuring them separately.

关键词： parallel virtual machine parallel computing environment LINUX platform

来源：评论

学校读者我要写书评

暂无评论

Application of Model Center to Real World Distributed and parallel Execution Challenges

Application of Model Center to Real World Distributed and Pa...

引用

AIAA Science and Technology Forum and Exposition (AIAA SciTech)

作者： Haisma, Michael Ko, Andy Levy, Mason Phoenix Integrat Blacksburg VA 24060 USA Lockheed Martin Palmdale CA 93599 USA

ISBN: (数字)9781624105951

ISBN: (纸本)9781624105951

Phoenix Integration worked with Lockheed Martin to apply ModelCenter software in a distributed and parallel computing context using ad hoc resources to the execution of MDO workflows for conceptual design of an aircraft. In the course of this work, the authors identified issues with automation of analysis tools in a parallel context, and developed methods to overcome those issues. The authors also performed a series of studies and to understand and characterize the performance of workflow execution using load balancing systems and ad hoc compute resources. Results of these studies indicate that while lighter workloads appear to scale well, more intensive workloads featuring heavy file I/O operations appear to be resource limited and prone to failure at higher levels of parallelization on each compute node, while scaling to higher number of compute nodes would be a more effective application of available resources.

关键词： computing Resources parallel computing environment CATIA Computer Systems Graphical User Interface Catastrophic Failure Black Box Air Force Research Laboratory Aircraft Conceptual Design Design of Experiments

来源：评论

学校读者我要写书评

暂无评论

GCRO with Dynamic Deflated Restarting for Solving Adjoint Systems of Equations for Aerodynamic Shape Optimization

GCRO with Dynamic Deflated Restarting for Solving Adjoint Sy...

引用

AIAA Aviation Forum and Exposition

作者： Chen, Chih-Hao Nadarajah, Siva McGill Univ Dept Mech Engn Montreal PQ H3A 2S6 Canada

ISBN: (数字)9781624105890

ISBN: (纸本)9781624105890

We present a dynamically adjusted deflated restarting procedure for the Generalized Conjugate Residual method with inner Orthogonalization (GCRO). The proposed method employs a GCR solver for the outer iteration and the generalized minimal residual (GMRES) with deflated restarting in the inner iteration. Approximate eigenpairs are evaluated at the end of each inner GMRES restart cycle. Our approach determines the number of vectors to be deflated from the spectrum based on the number of negative Ritz values, k*. We show that the approach restores convergence to cases where GMRES with restart failed and compare the approach against standard GMRES with restarts and deflated restarting. Efficiency is demonstrated for a two-dimensional NACA 0012 airfoil and a three-dimensional Common Research Model (CRM) wing. In addition, numerical experiments confirm the scalability of the solver.

关键词： Aerodynamic Shape Optimization NACA 0012 Reynolds Averaged Navier Stokes Turbulence Models Angle of Attack Lagrange Multipliers Flow Conditions Eddy Viscosity parallel computing environment Aircraft Configurations

来源：评论

学校读者我要写书评

暂无评论

MULTISECTION: parallelIZED BISECTION

MULTISECTION: PARALLELIZED BISECTION

引用

Winter Simulation Conference

作者： Pallone, Stephen Frazier, Peter I. Henderson, Shane G. Cornell Univ Operat Res & Informat Engn Ithaca NY 14853 USA

ISBN: (纸本)9781479974863

We consider a one-dimensional bisection method for finding the zero of a function, where function evaluations can be performed asynchronously in a parallel computing environment. Using dynamic programming, we characterize the Bayes-optimal policy for sequentially choosing points at which to query the function. In choosing these points, we face a trade-off between aggressively reducing the search space in the short term, and maintaining a desirable spread of queries in the long-term. Our results provide insight on how this trade-off is affected by function evaluation times, risk preferences, and computational budget.

关键词： Bayes methods dynamic programming parallel processing risk analysis search problems Bayes-optimal policy computational budget function evaluation times one-dimensional bisection method parallel computing environment parallelized bisection risk preferences search space Computational modeling Dynamic programming Heuristic algorithms Optimization parallel processing Random variables Stacking

来源：评论

学校读者我要写书评

暂无评论

Optimal three-dimensional nozzle shape design using CFD and parallel simulated annealing

引用

JOURNAL OF PROPULSION AND POWER 2002年第1期18卷 217-221页

作者： Wang, X Damodaran, M Nanyang Technol Univ Sch Mech & Prod Engn Ctr Adv Numer Engn Simulat Singapore 639798 Singapore

The optimal three-dimensional nozzle shape design was studied using computational fluid dynamics (CFD) and parallel simulated annealing (PSA). The three dimensional CFD solver for solving the Euler and Navier-Stokes equations was used to compute the internal flowfield from which the objective function was evaluated for each iterated aerodynamic shape. The computed results showed that parallel simulated annealing (PSA) appeared to be a robust and simple approach for the design of complex problems and that the evaluations of objective function could be reduced to the same level as generic gradient based methods.

关键词： Nozzle Walls Annealing Ion Thruster Magnetic Materials Total Variation Diminishing Navier Stokes Equations Sensitivity Analysis parallel computing environment Message Passing Interface Genetic Algorithm

来源：评论

学校读者我要写书评

暂无评论

Toward cost-effective aeroelastic analysis on advanced parallel computing systems

引用

JOURNAL OF AIRCRAFT 1999年第4期36卷 710-715页

作者： Goodwin, SA Weed, RA Sankar, LN Raj, P Lockeed Martin Aeronaut Syst Computat Fluid Dynam Flight Sci Div Marietta GA 30063 USA Mississippi State Univ Natl Sci Fdn Res Ctr Computat Field Simulat Nichols Res Inc Vicksburg MS 39180 USA Georgia Inst Technol Sch Aerosp Engn Atlanta GA 30332 USA

Progress toward the development and validation of a fast, accurate, and cost effective aeroelastic method for several advanced parallel computing platforms is presented. The ENSAERO code, developed at NASA Ames Research Center, was selected for this research effort. To improve the performance, the capabilities of the constituent modules have been enhanced. The relative merits of four parallel computing environments are investigated: SGI workstation cluster, SGI Power Challenge machine, SGI Power Challenge cluster, and the IBM SP2 system. To assess the computational performance of the enhanced ENSAERO code, the results of unsteady Navier-Stokes simulations of the transonic how over a fully flexible aeroelastic arrow-wing body configuration are reported. In addition, to ensure that the modified version of ENSAERO performs to specifications and to evaluate the benefits of the enhancements, unsteady Euler/Navier-Stokes calculations for an AGARD standard aeroelastic configuration are done. To evaluate the level of accuracy of the enhanced system, the ENSAERO results are compared with established computational solutions and experimental data.

关键词： Dynamic Aeroelastic Analysis parallel computing Systems parallel computing Platform Freestream Mach Number Navier Stokes Calculations Dynamic Pressure NASA Langley Research Center parallel computing environment Angle of Attack CFL3D

来源：评论

学校读者我要写书评

暂无评论

Random packs and their use in modeling heterogeneous solid propellant combustion

引用

JOURNAL OF PROPULSION AND POWER 2001年第4期17卷 883-891页

作者： Kochevets, S Buckmaster, J Jackson, TL Hegab, A Univ Illinois Dept Mech & Ind Engn Urbana IL 61801 USA Univ Illinois Dept Aeronaut & Astronaut Engn Urbana IL 61801 USA Univ Illinois Ctr Simulat Adv Rockets Urbana IL 61801 USA

It is shown that random packs of spheres of various sizes can be constructed that model ammonium-perchloratein-binder propellants in the sense that both the size distributions and the packing fractions of industrial propellant packs can be matched, Strategies for dealing with fractional numbers of large particles are addressed, as are strategies for dealing with a large number of very fine particles (fine powder), Fine powder is necessary in a three-dimensional pack to achieve the required stoichiometric ratio of ammonium perchlorate to fuel binder, but is not necessary in a two-dimensional (disk) pack, Some preliminary calculations of the two-dimensional combustion field supported by a disk pack are presented, in which full coupling between the gas phase, the condensed phase, and the retreating nonplanar propellant surface is accounted for.

关键词： Heterogeneous Solid Propellant Aluminum Combustion Number of Particles Scatter Plot parallel computing environment Navier Stokes Equations Reaction Rate Constant Universal Gas Constant Thermal Properties Surface Properties

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：