TECHNIQUES FOR WARM STARTING FINITE ELEMENT ANALYSES WITH DEEP NEURAL NETWORKS

摘要

A simulation application receives simulation parameters associated with a simulation to be generated. The simulation parameters include geometry associated with the simulation and corresponding boundary conditions. The simulation engine processes the simulation parameters and then, using a neural network, generates a solution estimate. Based on the estimated solution, the simulation engine then executes a finite element analysis solver using the solution estimate as a starting point. The FEA solver iterates until a converged solution is reached. The converged solution is then provided to the end-user.

主权项

1. A computer-implemented method for generating a finite element analysis (FEA) simulation, the method comprising:
generating a solution estimate based on a set of simulation parameters that describe physical properties of a simulation, wherein the solution estimate comprises an approximate solution for the simulation; warm-starting a finite element analysis (FEA) solver based on the solution estimate; and starting from the solution estimate, causing the FEA solver to iteratively solve a set of governing equations associated with the simulation to generate a converged solution for the simulation.