Skip to content

This repository contains the official code for the paper : Investigating the Ability of PINNs To Solve Burgers' PDE Near Finite-Time BlowUp

License

Notifications You must be signed in to change notification settings

Dibyakanti/PINNs-Solving-Burgers-Near-Finite-Time-Blow-Up

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 

Repository files navigation

PINNs Solving Burgers' Near Finite-Time Blow-Up

This repository contains the official code for the paper : Investigating the Ability of PINNs To Solve Burgers' PDE Near Finite-Time BlowUp

1. Training the PINNs

1.1 (1+1)-Burgers'

To train the PINN and save the models for the 1+1 Burgers PDE

python3 ./models/train1d.py \
--lr 1e-4 \
--wt_decay 0 \
--steps 100000 \
--boundary_points 300 \
--collocation_points 20000 \
--lambd_list 1 1 1 \
--delta_range 0.950 0.999 0.004 \
--width_range 4 8 1  \
--seed 1234

Argument details :
lr: learnign rate
wt_decay: weight decay for the optimizer (regularization) 
steps: number of training epochs 
boundary_points: number of points for intial and boundary condition each 
collocation_points: number of collocation points 
lambd_list: weight for the different terms of the boundary and inital loss 
delta_range: range of delta min, max and the step ( np.arange(min,max,step) ) 
width_range: range of width min, max and the step ( np.arange(min,max,step) )  
seed: seed for the random intialization (torch)

1.2 (2+1)-Burgers'

To train the PINN and save the models for the 2+1 Burgers PDE

python3 ./models/train2d.py \
--lr 1e-4 \
--wt_decay 0 \
--steps 100000 \
--boundary_points 300 \
--collocation_points 20000 \
--time_range 0.950 0.999 0.004 \
--width_range 4 8 1  \
--seed 1234

Argument details :
lr: learnign rate
wt_decay: weight decay for the optimizer (regularization) 
steps: number of training epochs 
boundary_points: number of points for intial and boundary condition each 
collocation_points: number of collocation points 
time_range: range of delta min, max and the step ( np.arange(min,max,step) ) 
width_range: range of width min, max and the step ( np.arange(min,max,step) )  
seed: seed for the random intialization (torch)

2. Analysis

2.1 (1+1)-Burgers'

To calculate the LHS and RHS of Equation 11 of Theorem 4.2

python3 ./analysis/analysis1d.py \
--width 30 \
--file_dir "../models/seed1234/"

Argument details :
width: width of the PINN
file_dir: directory from where to load the saved models

2.2 (2+1)-Burgers'

To calculate the LHS and RHS of Equation 5 of Theorem 4.1

python3 ./analysis/analysis2d.py \
--width 30 \
--file_dir "../models/seed1234/"

Argument details :
width: width of the PINN
file_dir: directory from where to load the saved models

About

This repository contains the official code for the paper : Investigating the Ability of PINNs To Solve Burgers' PDE Near Finite-Time BlowUp

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages