About Neural Networks

Neural networks:

  • are used to define arbitrary mappings (mathematical functions) between a set of inputs and a corresponding set of outputs;

  • are standalone model data;

  • must be defined in the input file and assigned a unique ID;

  • can be of arbitrary size and configuration;

  • can be evaluated from any user subroutine using a set of dedicated utility routines;

  • are assumed to be “trained”; and

  • can be used for a variety of applications, including surrogate constitutive modeling and complex boundary and loading conditions.

This page discusses:

See Also
Evaluating Abaqus Neural Networks
In Other Guides
*NEURAL NETWORK
*NEURAL NETWORK DATA
Activation functions used by feed-forward neural networks
Plate with elliptical hole under thermal loading

Products Abaqus/Standard