检索结果-内蒙古大学图书馆

Inverse Design of Airfoil Using a Deep Convolutional Neural Network

AIAA JOURNAL 2019年第3期57卷 993-1003页

作者： Sekar, Vinothkumar Zhang, Mengqi Shu, Chang Khoo, Boo Cheong Natl Univ Singapore Dept Mech Engn 9 Engn Dr 1 Singapore 117575 Singapore

This paper proposes an approach to perform the inverse design of airfoils using deep convolutional neural networks (CNNs). The conventional approaches are based on the solution of differential equations, which are either difficult to solve or take a tedious procedure to obtain the solution. As shown in this paper, the deep learning technique can be used effectively to obtain the airfoil shape from the coefficient of pressure distribution. More specifically, the CNN is applied due to its ability to handle any airfoil geometry without the need for complex parametrization. In the training phase, the pressure coefficient distribution is fed as input to the CNN to obtain a prediction model for the airfoil shape. In the testing phase, a new pressure coefficient distribution is given to the CNN model, generating an airfoil shape that is very close to the associated airfoil with an L2 error of less than 2% for most of the cases. An extensive investigation of the effects of various hyperparameters in the CNN architectures is performed. The results show that the CNN method is very efficient in terms of computational time and shows a competitive prediction accuracy. It thus constitutes an attractive new method in the inverse design of airfoils, especially when existing data can be readily exploited.

关键词： Airfoil Databases Deep Convolutional Neural Network Pressure Coefficient Airfoil Geometry Machine Learning Angle of Attack Aerodynamic Shape Optimization numerical algorithms Computational Fluid Dynamics Flow Conditions

来源：评论

学校读者我要写书评

暂无评论

Teaching a generative model: Mathematical formulation and solution framework

引用

COMPUTERS & INDUSTRIAL ENGINEERING 2019年 130卷 119-126页

作者： Wang, Honggang Zhang, Bo Univ La Verne Dept Appl Business Sci & Econ Coll Business & Publ Management La Verne CA 91750 USA IBM Thomas J Watson Res Ctr Yorktown Hts NY USA

We develop the mathematical formulation for teaching generative models to a learner whose learning processes and cognitive behaviors may be analytically intractable, but can be simulated by numerical processes. The model considers the learner's bias (prior knowledge) or memory process by using stochastic models. We also present an optimization framework for solving the involved non-convex, stochastic optimization problems associated with machine teaching. The algorithm design and the conditions and analysis are discussed for local convergence properties of the proposed optimization algorithms. In the paper, we discuss a number of example cases to illustrate the algorithmic ideas and demonstrate their efficiency.

关键词： Statistical learning Machine teaching Data driven optimization Computer simulation numerical algorithms

来源：评论

学校读者我要写书评

暂无评论

DSF chain for biorthogonal polynomials and its application to matrix eigenvalue problems

引用

JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS 2019年 347卷 273-295页

作者： Mukaihira, Atsushi Tsuboi, Yoshimasa Doshisha Univ Fac Sci & Engn Kyotanabe Kyoto Japan

A discrete-time integrable system, called the DSF chain, is derived from the compatibility condition of spectral equations for biorthogonal polynomials associated with the discrete Schur flow. The DSF algorithm for computing eigenvalues of a type of matrix eigenvalue problem is designed. A discrete analogue of the Miura transformation between the discrete Schur flow and the qd-algorithm is constructed. Two special solutions to the DSF chain related to q-special functions are presented. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Schur flow Qd-algorithm Matrix eigenvalue problems numerical algorithms Integrable systems Biorthogonal polynomials

来源：评论

学校读者我要写书评

暂无评论

Discrete Total Variation with Finite Elements and Applications to Imaging

引用

JOURNAL OF MATHEMATICAL IMAGING AND VISION 2019年第4期61卷 411-431页

作者： Herrmann, Marc Herzog, Roland Schmidt, Stephan Vidal-Nunez, Jose Wachsmuth, Gerd TU Chemnitz Fac Math D-09107 Chemnitz Germany Julius Maximilians Univ Wurzburg Fac Math & Comp Sci Lehrstuhl Math 6 Emil Fischer Str 40 D-97074 Wurzburg Germany Brandenburgische Tech Univ Cottbus Senftenberg Inst Math Chair Optimal Control Pl Deutsch Einheit 1 D-03046 Cottbus Germany

The total variation (TV)-seminorm is considered for piecewise polynomial, globally discontinuous (DG) and continuous (CG) finite element functions on simplicial meshes. A novel, discrete variant (DTV) based on a nodal quadrature formula is defined. DTV has favorable properties, compared to the original TV-seminorm for finite element functions. These include a convenient dual representation in terms of the supremum over the space of Raviart-Thomas finite element functions, subject to a set of simple constraints. It can therefore be shown that a variety of algorithms for classical image reconstruction problems, including TV-L-2 denoising and inpainting, can be implemented in low- and higher-order finite element spaces with the same efficiency as their counterparts originally developed for images on Cartesian grids.

关键词： Discrete total variation Dual problem Image reconstruction numerical algorithms

来源：评论

学校读者我要写书评

暂无评论

Deep drawing simulations using the finite element method embedding a multi-level crystal plasticity constitutive law: Experimental verification and sensitivity analysis

引用

COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING 2019年第Sep.1期354卷 245-270页

作者： Barrett, Timothy J. Knezevic, Marko Univ New Hampshire Dept Mech Engn 33 Acad WayKingsbury HallW119 Durham NH 03824 USA

This paper presents experimental verification of a multi-level simulation framework aimed at sheet metal forming analysis. Specifically, deep drawing of a cylindrical cup simulations from alloy AA6022-T4 sheets are carried out using a physically based elasto-plastic self-consistent (EPSC) polycrystalline homogenization model embedded in implicit finite elements and verified experimentally. The EPSC model takes into account the evolution of microstructure and directionality of deformation mechanisms acting at the single-crystal level in predicting material behavior. It is calibrated and validated as a standalone model using flow stress and R-ratio data as well as iso-shear contours measured along several directions of the sheet through uniaxial and plane strain tension experiments. Furthermore, the particularities pertaining to cyclic response including non-linear unloading and the Bauschinger effect are also calibrated through large strain tension-compression data. Consistent with experimental measurements, the process simulations in finite elements predict directionally dependent thinning of the cup, especially around the punch radius and variation in the cup height around the rim of the cup, referred to as earing. By comparing experiments and predictions, role of the R-ratio is revealed as critical for the accurate prediction of the cup height. Further sensitivity analysis shows that initial texture has a strong influence on the R-ratio, while introducing a minor effect on hardening. The analysis into the choice of finite element types in terms of their accuracy and efficiency shows that the 3D 8 nodal elements (C3D8R) and continuum shell 3D 8 nodal elements (SC8R) are superior over the planar shell 3D 4 nodal elements (S4R) with the former being the most accurate and the latter being computationally efficient. It is demonstrated that the simulation framework presented in this paper can be used to predict phenomena pertaining to material behavior and resulting geo

关键词： Microstructures Crystal plasticity Anisotropic material numerical algorithms AA6022-T4

来源：评论

学校读者我要写书评

暂无评论

Improved Sine-Fitting algorithms for Measurements of Complex Ratio of AC Voltages by Asynchronous Sequential Sampling

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2019年第6期68卷 1659-1665页

作者： Augustyn, Jerzy Kampik, Marian Kielce Univ Technol Dept Comp Sci Elect & Elect Engn PL-25314 Kielce Poland Silesian Tech Univ Dept Elect & Elect Engn PL-44100 Gliwice Poland

This paper presents a modified sine-fitting algorithms which can be used for high-accuracy sampling measurements of complex voltage ratio of sinusoidal signals. The proposed algorithms provide a significant reduction ... 详细信息

关键词： Asynchronous sampling complex voltage ratio (CVR) numerical algorithms phase-shift measurements precise measurements sampling measurements sine-fitting algorithms (SFAs)

来源：评论

学校读者我要写书评

暂无评论

Towards a greater understanding of serrated flows in an Al-containing high-entropy-based alloy

引用

INTERNATIONAL JOURNAL OF PLASTICITY 2019年 115卷 71-92页

作者： Brechtl, J. Chen, S. Y. Xie, X. Ren, Y. Qiao, J. W. Liaw, P. K. Zinkle, S. J. Univ Tennessee Bredesen Ctr Interdisciplinary Res & Educ Knoxville TN 37996 USA Univ Tennessee Dept Mat Sci & Engn Knoxville TN 37996 USA Argonne Natl Lab Xray Sci Div 9700 S Cass Ave Argonne IL 60439 USA Taiyuan Univ Technol Coll Mat Sci & Engn Taiyuan Shanxi Peoples R China Univ Tennessee Dept Nucl Engn Knoxville TN 37996 USA

A serrated flow, which occurs in a material undergoing mechanical deformation, is a complex process of great engineering significance. Here statistical, dynamical, and multifractal modeling and analyses were performed on the stress-time series to characterize and model the stress-drop behavior of an Al0.5CoCrCuFeNi high-entropy alloy (HEA). Results indicate that the spatiotemporal dynamics of the serrated flow is affected by changes in the strain rate and test temperature. The sample entropy, in general, was found to be the highest in the samples tested at 500 degrees C. The higher complexity in the serrated flow at this temperature appeared to be associated with the stress-drop behavior that had intermediate values in terms of the maximum stress drop, the multifractality of the data set, and the histogram distributions. Moreover, the sample entropy was the lowest for the samples tested at 600 degrees C. The lower complexity values were associated with a wider multifractal spectrum and a less uniform and sparser distribution of the stress-drop magnitudes. In terms of the serration types, Type-C serrations were related to the lowest complexity values, widest multifractal spectra, and higher probability of exhibiting larger stress drops. Conversely, Type-A and B serrations were associated with the higher complexity, narrower spectra, and lower probability of higher stress drops. Furthermore, the body-centered-cubic (BCC) structure and the fully-ordered L1(2) nano-particles were found to emerge in the samples at 600 degrees C and are thought to be linked to the decreased spatiotemporal correlations in the stress-drop behavior. A serrated flow, which occurs in a material undergoing mechanical deformation, is a complex process of great engineering significance. Here statistical, dynamical, and multifractal modeling and analyses were performed on the stress-time series to characterize and model the stress-drop behavior of an Al0.5CoCrCuFeNi high-entropy alloy (HEA). Resul

关键词： High entropy alloys Dislocations Serrated flow Mechanical testing numerical algorithms

来源：评论

学校读者我要写书评

暂无评论

Fuel-Optimal Rocket Landing with Aerodynamic Controls

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2019年第1期42卷 65-77页

作者： Liu, Xinfu Beijing Inst Technol Sch Aerosp Engn Beijing 100081 Peoples R China

Aerodynamic forces are not negligible for a reusable rocket returning back to Earth. How the aerodynamic controls and propulsion should be coordinated to realize fuel-optimal precise landing is addressed in this paper. To this end, a model-based optimal control problem is formulated with the rocket's angle of attack and thrust as control inputs, and constraints on the controls are included to reflect the capabilities of the vehicle. Precise landing requires the (highly nonlinear and nonconvex) problem to be solved onboard in real time. This ability of online computation is becoming increasingly desired in aerospace guidance and control for autonomous missions. Hence, this paper presents how to solve the rocket landing problem via convex optimization that has guaranteed polynomial-time complexity. Specifically, a novel methodology of handling the rocket nonlinear dynamics is introduced, and a relaxation technique used to convexify nonconvex constraints is theoretically proved to be valid. High efficiency of the proposed method, with potential for online computation, is demonstrated by numerical examples and comparisons with other methods.

关键词： Rocket Trajectories Fuels Lift Coefficient Aerodynamic Force Nonlinear Programming Trajectory Optimization Nonlinear Dynamics Earth Flight Path Angle numerical algorithms

来源：评论

学校读者我要写书评

暂无评论

Efficient Derivative Evaluation for Rigid-Body Dynamics Based on Recursive algorithms Subject to Kinematic and Loop Constraints

引用

IEEE CONTROL SYSTEMS LETTERS 2019年第3期3卷 619-624页

作者： Kudruss, Manuel Manns, Paul Kirches, Christian Visteon Elect Germany GmbH ADAS Dept D-76227 Karlsruhe Germany Tech Univ Carolo Wilhelmina Braunschweig Inst Math Optimizat D-38106 Braunschweig Germany

Simulation and control of robotic and bio-mechanical systems depend on a mathematical model description, typically a rigid-body system connected by joints, for which efficient algorithms to compute the forward or inverse dynamics exist. Gradient-based optimization and control methods require derivatives of the dynamics, often approximated by numerical differentiation (FD). However, they benefit from accurate gradients, which promote faster convergence, less iterations, and improved handling of nonlinearities or ill-conditioning of the problem formulations, which are particularly observed when kinematic constraints are involved. In this letter, we apply algorithmic differentiation (AD) to propagate sensitivities through dynamics algorithms. To this end, we augment the computational graph of these algorithms with derivative information. We provide analytic derivatives for elementary operations, in particular matrix factorizations of the descriptor form of the equation of motions, which yields a very efficient derivative evaluation for constrained dynamics. The proposed approach is implemented within the free software package rigid body dynamics library (RBDL), which heavily employs so-called spatial transformations in its implementation of the dynamics algorithms. Thus, manipulations of spatial transformations are treated as elementary operations. The efficiency is improved further by sparsity exploitation. We validate and benchmark the implementation against its FD counterpart for a lifting motion of a human model.

关键词： numerical algorithms robotics control applications

来源：评论

学校读者我要写书评

暂无评论

Modelling recrystallization textures driven by intragranular fluctuations implemented in the viscoplastic self-consistent formulation

引用

ACTA MATERIALIA 2019年 164卷 530-546页

作者： Zecevic, Miroslav Lebensohn, Ricardo A. McCabe, Rodney J. Knezevic, Marko Univ New Hampshire Dept Mech Engn 33 Acad WayKingsbury HallW119 Durham NH 03824 USA Los Alamos Natl Lab Theoret Div Los Alamos NM 87544 USA Los Alamos Natl Lab Mat Sci & Technol Div Los Alamos NM 87544 USA

This paper presents a recrystallization model driven by intragranular orientation gradients and strain energy fields calculated by means of the viscoplastic self-consistent (VPSC) formulation. The VPSC model is extended for calculation of the coupling between intragranular stress fluctuations with corresponding second moments of lattice spin and misorientation fields in the grains. Access to these quantities allows modelling of transition bands and nucleation kinetics. In the proposed recrystallization model, grain growth is assumed to be proportional to the difference between the stored energy of each grain and that of the effective medium. Recrystallization textures for several cubic metals are simulated, showing good agreement with corresponding experiments. The model reveals the importance of considering appropriate, microstructurally-based and orientation-dependent recrystallization nucleation mechanisms. The recrystallization texture of heavily rolled copper with a strong cube texture component is found to be a consequence of nucleation at transition bands, which is also the cause of the recrystallization textures in compressed iron and drawn copper wire. In contrast, the recrystallization texture of rolled interstitial free steel is found to be caused by grain boundary nucleation occurring in grains with the highest strain energy. (C) 2018 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

关键词： Microstructures Texture Crystal plasticity Recrystallization numerical algorithms

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：