• Home
  • Search
  • Explainable Artificial Intelligence for Mechanics: Physics-Explaining Neural Networks for Constitutive Models
  • Cite Icon36
  • https://doi.org/10.3389/fmats.2021.824958Copy DOI Icon

Explainable Artificial Intelligence for Mechanics: Physics-Explaining Neural Networks for Constitutive Models

Show More
  • Abstract
  • Highlights & Summary
  • PDF
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

(Artificial) neural networks have become increasingly popular in mechanics and materials sciences to accelerate computations with model order reduction techniques and as universal models for a wide variety of materials. However, the major disadvantage of neural networks remains: their numerous parameters are challenging to interpret and explain. Thus, neural networks are often labeled as black boxes, and their results often elude human interpretation. The new and active field of physics-informed neural networks attempts to mitigate this disadvantage by designing deep neural networks on the basis of mechanical knowledge. By using this a priori knowledge, deeper and more complex neural networks became feasible, since the mechanical assumptions can be explained. However, the internal reasoning and explanation of neural network parameters remain mysterious. Complementary to the physics-informed approach, we propose a first step towards a physics-explaining approach, which interprets neural networks trained on mechanical data a posteriori. This proof-of-concept explainable artificial intelligence approach aims at elucidating the black box of neural networks and their high-dimensional representations. Therein, the principal component analysis decorrelates the distributed representations in cell states of RNNs and allows the comparison to known and fundamental functions. The novel approach is supported by a systematic hyperparameter search strategy that identifies the best neural network architectures and training parameters. The findings of three case studies on fundamental constitutive models (hyperelasticity, elastoplasticity, and viscoelasticity) imply that the proposed strategy can help identify numerical and analytical closed-form solutions to characterize new materials.

Loading PDF

Similar Papers
  • Conference Article
  • Citations34

Training of Artificial Neural Network Using PSO With Novel Initialization Technique

  • Nov 01, 2018
  • Hafiz Tayyab Rauf +3
  • Research Article

О нейросетевых моделях деформирования тонкостенных оболочек

  • Jan 01, 2025
  • LETI Transactions on Electrical Engineering & Computer Science
  • Iu N Zgoda
  • Research Article
  • Citations52

On the training and performance of high-order neural networks

  • Oct 01, 1995
  • Mathematical Biosciences
  • Nicolaos B Karayiannis +1
  • Research Article
  • Citations64

Accurate online training of dynamical spiking neural networks through Forward Propagation Through Time

  • May 08, 2023
  • Nature Machine Intelligence
  • Bojian Yin +2
  • Research Article

Potential of a neural network in the diagnosis of laryngeal tumors

  • Jul 03, 2024
  • Digital Diagnostics
  • Evgeniya A Safyannikova +10
  • Research Article

Abstract C006: DrugCell: A visible neural network to guide precision medicine

  • Dec 01, 2019
  • Molecular Cancer Therapeutics
  • Brent M Kuenzi +4
  • Research Article

ЖАДНОЕ ПОСЛОЙНОЕ ОБУЧЕНИЕ СВЕРТОЧНЫХ НЕЙРОННЫХ СЕТЕЙ

  • Jan 01, 2021
  • SOFT MEASUREMENTS AND COMPUTING
  • V.V Erokhin
  • Research Article
  • Citations9

Rule extraction based on PROMETHEE-assisted multi-objective genetic algorithm for generating interpretable neural networks

  • Dec 17, 2023
  • Applied Soft Computing
  • Dounia Yedjour +3
  • Research Article

Rapid Inverse Parameter Inference Using Physics-Informed Neural Networks

  • Aug 09, 2024
  • Electrochemical Society Meeting Abstracts
  • Malik Hassanaly +4
  • PDF
  • Research Article

A neural network approach to the identification of room occupationalness according to air parameters

  • Apr 17, 2024
  • System technologies
  • Huk K.G +1
  • Research Article
  • Citations15

Taguchi Based Optimisation of Artificial Neural Network to Establish a Direct Microstructure: Mechanical Property Correlation in a near-α Titanium Alloy

  • Mar 09, 2016
  • Transactions of the Indian Institute of Metals
  • I Balasundar +2
  • Conference Article
  • Citations7

Artificial Intelligence for the Diagnostics of Gas Turbines: Part I — Neural Network Approach

  • Jan 01, 2005
  • R Bettocchi +3
  • Research Article
  • Citations34

Application of artificial neural network for predicting plain strain fracture toughness using tensile test results

  • Feb 27, 2006
  • Fatigue & Fracture of Engineering Materials & Structures
  • J Y Kang +2
  • Conference Article
  • Citations4

Neural Network Controller for Attitude Control of Quadrotor

  • Nov 01, 2019
  • Ussama Maqbool +2
  • Book Chapter

Multi-Objective Training of Neural Networks

  • Jan 01, 2009
  • M P Cuéllar +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.