Skip to content

Inverstigation of Conditional Variational Autoencoder for Finite Element Method simulation

Notifications You must be signed in to change notification settings

ziyinq/linearFEM

Repository files navigation

linearFEM

Techniqal detail please refer to the report

Compile

  • This simulation code is built based on camke and C++
  • To compile the C++ code, create a new folder in the root directory, type
cmake ..

then type

make

and a executable file will generated in the src folder

  • modify the main.cpp to use linear or quadratic FEM simulation under different Young's modulus
  • create a folder named output, the simulation result .bgeo files will be written in this folder

CVAE code

  • the algorithm is written in python and PyTorch
  • the CVAE model is defined in CVAEcuda.py, and linearFEM_CVAE(l2l).py is the framework for training linear to linear, linearFEM_CVAE(l2q).py is the framework for training linear to quadratic.

About

Inverstigation of Conditional Variational Autoencoder for Finite Element Method simulation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published