Computational mechanics enhanced by deep learning

Abstract The present paper describes a method to enhance the capability of, or to broaden the scope of computational mechanics by using deep learning, which is one of the machine learning methods and is based on the artificial neural network. The method utilizes deep learning to extract rules inherent in a computational mechanics application, which usually are implicit and sometimes too complicated to grasp from the large amount of available data A new method of numerical quadrature for the FEM stiffness matrices is developed by using the proposed method, where a kind of optimized quadrature rule superior in accuracy to the standard Gauss–Legendre quadrature is obtained on the element-by-element basis. The detailed formulation of the proposed method is given with the sample application above, and an acceleration technique for the proposed method is discussed

[1]  T. Hughes,et al.  Isogeometric analysis : CAD, finite elements, NURBS, exact geometry and mesh refinement , 2005 .

[2]  K. Giannakoglou,et al.  Aerodynamic shape design using evolutionary algorithms and new gradient-assisted metamodels , 2006 .

[3]  Sanjay R. Arwade,et al.  Response classification of simple polycrystalline microstructures , 2008 .

[4]  Eric Darve,et al.  Assembly of finite element methods on graphics processors , 2011 .

[5]  Ruben Sevilla,et al.  Numerical integration over 2D NURBS-shaped domains with applications to NURBS-enhanced FEM , 2011 .

[6]  Genki Yagawa,et al.  INELASTIC CONSTITUTIVE PARAMETER IDENTIFICATION USING AN EVOLUTIONARY ALGORITHM WITH CONTINUOUS INDIVIDUALS , 1997 .

[7]  Bernhard A. Schrefler,et al.  Artificial Neural Networks in numerical modelling of composites , 2009 .

[8]  E. Oñate,et al.  Neural networks for variational problems in engineering , 2008 .

[9]  A. Elnashai,et al.  Self-learning simulation method for inverse nonlinear modeling of cyclic behavior of connections , 2008 .

[10]  T. Belytschko,et al.  Element‐free Galerkin methods , 1994 .

[11]  S. Rajendran,et al.  A technique to develop mesh-distortion immune finite elements , 2010 .

[12]  Marc'Aurelio Ranzato,et al.  Building high-level features using large scale unsupervised learning , 2011, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[13]  Markus Zimmermann,et al.  Computing solution spaces for robust design , 2013 .

[14]  T. Hughes,et al.  Efficient quadrature for NURBS-based isogeometric analysis , 2010 .

[15]  Krzysztof Banaś,et al.  Finite element numerical integration for first order approximations on multi- and many-core architectures , 2016 .

[16]  Sung-Kie Youn,et al.  Isogeometric analysis with trimming technique for problems of arbitrary complex topology , 2010 .

[17]  Antonio Huerta,et al.  3D NURBS‐enhanced finite element method (NEFEM) , 2008 .

[18]  M. Lefik,et al.  Artificial neural network as an incremental non-linear constitutive model for a finite element code , 2003 .

[19]  Manolis Papadrakakis,et al.  Structural optimization using evolution strategies and neural networks , 1998 .

[20]  Rafael Stubs Parpinelli,et al.  A comparison of swarm intelligence algorithms for structural engineering optimization , 2012 .

[21]  Victor M. Calo,et al.  Optimal quadrature rules for odd-degree spline spaces and their application to tensor-product-based isogeometric analysis , 2016 .

[22]  Antonio Delgado,et al.  Damage detection on crates of beverages by artificial neural networks trained with finite-element data , 2004 .

[23]  Rachid Ait-Haddou,et al.  Explicit Gaussian quadrature rules for C1 cubic splines with symmetrically stretched knot sequences , 2015, J. Comput. Appl. Math..

[24]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[25]  Jamshid Ghaboussi,et al.  Autoprogressive training of neural network constitutive models , 1998 .

[26]  Manolis Papadrakakis,et al.  An adaptive neural network strategy for improving the computational performance of evolutionary structural optimization , 2005 .

[27]  Genki Yagawa,et al.  Three-dimensional elasto-static analysis of 100 million degrees of freedom , 2001 .

[28]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[29]  Yoshua Bengio,et al.  Maxout Networks , 2013, ICML.

[30]  Yoram Singer,et al.  Adaptive Subgradient Methods for Online Learning and Stochastic Optimization , 2011, J. Mach. Learn. Res..

[31]  Genki Yagawa,et al.  Neural networks in computational mechanics , 1996 .

[32]  Jens Markus Melenk,et al.  Fully discrete hp-finite elements: fast quadrature , 2001 .

[33]  Adi Shraibman,et al.  Rank, Trace-Norm and Max-Norm , 2005, COLT.

[34]  Youssef M A Hashash,et al.  Numerical implementation of a neural network based material model in finite element analysis , 2004 .

[35]  Genki Yagawa,et al.  NEW REGULARIZATION BY TRANSFORMATION FOR NEURAL NETWORK BASED INVERSE ANALYSES AND ITS APPLICATION TO STRUCTURE IDENTIFICATION , 1996 .

[36]  Debaditya Dutta,et al.  A computationally efficient approach for inverse material characterization combining Gappy POD with direct inversion , 2015 .

[37]  Jamshid Ghaboussi,et al.  Characterizing rate-dependent material behaviors in self-learning simulation , 2006 .

[38]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[39]  B. Schrefler,et al.  ANN approach to sorption hysteresis within a coupled hygro‐thermo‐mechanical FE analysis , 2001 .

[40]  Bernard Haasdonk,et al.  Surrogate modeling of multiscale models using kernel methods , 2015 .

[41]  Marco Lúcio Bittencourt,et al.  Tensor‐based Gauss–Jacobi numerical integration for high‐order mass and stiffness matrices , 2009 .

[42]  Steffen Freitag,et al.  Modeling of materials with fading memory using neural networks , 2009 .

[43]  G. Yagawa,et al.  Free mesh method: A new meshless finite element method , 1996 .

[44]  Robert E. Smith,et al.  Classifier systems in combat: two-sided learning of maneuvers for advanced fighter aircraft , 2000 .

[45]  N. Sukumar,et al.  Generalized Gaussian quadrature rules on arbitrary polygons , 2010 .

[46]  N. S. Mera,et al.  The use of neural network approximation models to speed up the optimisation process in electrical impedance tomography , 2007 .

[47]  Ken-ichi Funahashi,et al.  On the approximate realization of continuous mappings by neural networks , 1989, Neural Networks.

[48]  N. Huber,et al.  A neural network tool for identifying the material parameters of a finite deformation viscoplasticity model with static recovery , 2001 .

[49]  John A. Gunnels,et al.  Large scale drop impact analysis of mobile phone using ADVC on Blue Gene/L , 2006, SC.

[50]  Thomas J. R. Hughes,et al.  Reduced Bézier element quadrature rules for quadratic and cubic splines in isogeometric analysis , 2014 .

[51]  M. Kikuchi.  Application of the symbolic mathematics system to the finite element program , 1989 .

[52]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[53]  Youssef M. Marzouk,et al.  A new network approach to Bayesian inference in partial differential equations , 2015 .

[54]  Manolis Papadrakakis,et al.  Innovative seismic design optimization with reliability constraints , 2008 .

[55]  Giancarlo Sangalli,et al.  Nonlinear static isogeometric analysis of arbitrarily curved Kirchhoff-Love shells , 2015, International Journal of Mechanical Sciences.

[56]  P. S. Koutsourelakis,et al.  Sparse Variational Bayesian Approximations for Nonlinear Inverse Problems: applications in nonlinear elastography , 2014, 1412.0473.

[57]  Genki Yagawa,et al.  A parallel finite element method with a supercomputer network , 1993 .

[58]  Leo Grady,et al.  Impact of geometric uncertainty on hemodynamic simulations using machine learning , 2015 .

[59]  Pritish Narayanan,et al.  Deep Learning with Limited Numerical Precision , 2015, ICML.

[60]  Yong Hyup Kim,et al.  A predictor–corrector method for structural nonlinear analysis , 2001 .

[61]  P. Hansbo.  A new approach to quadrature for finite elements incorporating hourglass control as a special case , 1998 .

[62]  Genki Yagawa,et al.  Quantitative nondestructive evaluation with ultrasonic method using neural networks and computational mechanics , 1995 .

[63]  Renato Vitaliani,et al.  Shortened quadrature rules for finite elements , 1982 .

[64]  Pietro Marco Congedo,et al.  Shape optimization of an airfoil in a BZT flow with multiple-source uncertainties , 2011 .

[65]  Genki Yagawa,et al.  Neural Network-Based Inverse Analysis for Defect Identification with Laser Ultrasonics , 2001 .

[66]  Beatriz Souza Leite Pires de Lima,et al.  Bio‐inspired algorithms for the optimization of offshore oil production systems , 2012 .

[67]  Adil Amirjanov,et al.  Investigation of a changing range genetic algorithm in noisy environments , 2008 .

[68]  Genki Yagawa,et al.  Implicit constitutive modelling for viscoplasticity using neural networks , 1998 .

[69]  G. Yagawa,et al.  A numerical integration scheme for finite element method based on symbolic manipulation , 1990 .

[70]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[71]  Ryuusuke Kawamura,et al.  Optimization of material composition of nonhomogeneous hollow sphere for thermal stress relaxation making use of neural network , 1999 .

[72]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[73]  Hyun-Jung Kim,et al.  Isogeometric analysis for trimmed CAD surfaces , 2009 .

[74]  Georgios E. Stavroulakis,et al.  Neural crack identification in steady state elastodynamics , 1998 .

[75]  J. L. Marcelin,et al.  Genetic optimization of stiffened plates without the FE mesh support , 2002 .

[76]  J. L. Marcelin,et al.  Genetic optimization of stiffened plates and shells , 2001 .

[77]  T. Hughes,et al.  A Simple Algorithm for Obtaining Nearly Optimal Quadrature Rules for NURBS-based Isogeometric Analysis , 2012 .

[78]  C. Poloni,et al.  Hybridization of a multi-objective genetic algorithm, a neural network and a classical optimizer for a complex design problem in fluid dynamics , 2000 .

[79]  Wing Kam Liu,et al.  A multiple-quadrature eight-node hexahedral finite element for large deformation elastoplastic analysis , 1998 .

[80]  Yoshua Bengio,et al.  Low precision storage for deep learning , 2014 .

[81]  Trenton Kirchdoerfer,et al.  Data-driven computational mechanics , 2015, 1510.04232.

[82]  D. E. Goldberg,et al.  Genetic Algorithms in Search , 1989 .

[83]  Genki Yagawa,et al.  Design window search using continuous evolutionary algorithm and clustering––its application to shape design of microelectrostatic actuator , 2002 .

[84]  Benjamin Peherstorfer,et al.  Dynamic data-driven reduced-order models , 2015 .

[85]  Antonio Rovira,et al.  A new methodology to solve non‐linear equation systems using genetic algorithms. Application to combined cyclegas turbine simulation , 2005 .

[86]  Martín Abadi,et al.  TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems , 2016, ArXiv.

[87]  Jin H. Huang,et al.  Detection of cracks using neural networks and computational mechanics , 2002 .

[88]  José Manuel García-Aznar,et al.  Computational evaluation of different numerical tools for the prediction of proximal femur loads from bone morphology , 2014 .

[89]  Shinobu Yoshimura,et al.  A New Local Contact Search Method Using a Multi-Layer Neural Network , 2007 .

[90]  Shinobu Yoshimura,et al.  Neural network approach to estimate stable crack growth in welded specimens , 1995 .

[91]  T. Furukawa,et al.  Pareto‐based continuous evolutionary algorithms for multiobjective optimization , 2002 .

[92]  Kjetil André Johannessen,et al.  Optimal quadrature for univariate and tensor product splines , 2017 .

[93]  Manolis Papadrakakis,et al.  Reliability-based structural optimization using neural networks and Monte Carlo simulation , 2002 .

[94]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[95]  Genki Yagawa,et al.  Recent developments of free mesh method , 2000 .

[96]  Tiangang Cui,et al.  Multifidelity importance sampling , 2016 .

[97]  Dan Givoli,et al.  XFEM‐based crack detection scheme using a genetic algorithm , 2007 .

[98]  Ionel M. Navon,et al.  Truncated-Newton training algorithm for neurocomputational viscoplastic model , 2003 .

[99]  Ran El-Yaniv,et al.  Binarized Neural Networks , 2016, NIPS.

[100]  M. Kikuchi.  Application of the symbolic mathematics system to the finite element program: evaluation of the stiffness matrix of a 4-nodes isoparametric element , 1989 .

[101]  Tomonari Furukawa,et al.  Neural network constitutive modelling for non‐linear characterization of anisotropic materials , 2011 .

[102]  David J. Benson,et al.  On the numerical integration of trimmed isogeometric elements , 2015 .