Bayesian inference of non-linear multiscale model parameters accelerated by a deep neural network

DOI10.1016/j.cma.2019.112693zbMath1441.74192OpenAlexW2984016396WikidataQ126841969 ScholiaQ126841969MaRDI QIDQ2175257

Zoltan Major, Aitor Arriaga, Kepa Zulueta, Ludovic Noels, Ling Wu

Publication date: 29 April 2020

Published in: Computer Methods in Applied Mechanics and Engineering (Search for Journal in Brave)

Full work available at URL: https://orbi.uliege.be/handle/2268/240110

zbMATH Keywords

compositesneural networkBayesian inferencemultiscalenon-linear

Mathematics Subject Classification ID

Bayesian inference (62F15) Bounds on effective properties in solid mechanics (74Q20) Applications to the sciences (65Z05) Neural nets and related approaches to inference from stochastic processes (62M45)

Related Items (13)

Variational Bayesian approximation of inverse problems using sparse precision matrices ⋮ An intelligent nonlinear meta element for elastoplastic continua: deep learning using a new time-distributed residual U-net architecture ⋮ <scp>Data</scp>‐physics driven reduced order homogenization ⋮ Interfacing finite elements with deep neural operators for fast multiscale modeling of mechanics problems ⋮ A machine learning-based viscoelastic-viscoplastic model for epoxy nanocomposites with moisture content ⋮ A micromechanical mean‐field homogenization surrogate for the stochastic multiscale analysis of composite materials failure ⋮ Data-physics driven multiscale approach for high-pressure resin transfer molding (HP-RTM) ⋮ Probabilistic physics-guided transfer learning for material property prediction in extrusion deposition additive manufacturing ⋮ Numerical Analysis for Convergence of a Sample-Wise Backpropagation Method for Training Stochastic Neural Networks ⋮ A recurrent neural network-accelerated multi-scale model for elasto-plastic heterogeneous materials subjected to random cyclic and non-proportional loading paths ⋮ A variable-fidelity hybrid surrogate approach for quantifying uncertainties in the nonlinear response of braided composites ⋮ A neural network-aided Bayesian identification framework for multiscale modeling of nanocomposites ⋮ Recurrent neural networks (RNNs) with dimensionality reduction and break down in computational mechanics; application to multi-scale localization step

Uses Software

• Scikit
• PRMLT

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Adaptive proposal distribution for random walk Metropolis algorithm
• A Bayesian approach to selecting hyperelastic constitutive models of soft tissue
• A framework for data-driven analysis of materials under uncertainty: countering the curse of dimensionality
• Micromechanics of inelastic composites with misaligned inclusions: numerical treatment of orientation
• Continuum micro-mechanics of elastoplastic polycrystals
• Estimating fibres' material parameter distributions from limited data with the help of Bayesian inference
• Computational homogenization of nonlinear elastic materials using neural networks
• The determination of the elastic field of an ellipsoidal inclusion, and related problems
• Variational Principles for Inhomogeneous Non-linear Media
• Bounds and Self-Consistent Estimates for the Overall Properties of Nonlinear Composites
• A New Variational Principle and Its Application to Nonlinear Heterogeneous Systems

This page was built for publication: Bayesian inference of non-linear multiscale model parameters accelerated by a deep neural network