Skip to content

A Deep Learning based Approach to Reduced Order Modeling for Turbulent Flow Control using LSTM Neural Networks. arXiv:1804.09269

License

Notifications You must be signed in to change notification settings

mahdikherad/LSTM_ROM_Arxiv

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LSTM ROM for turbulent flow control - arxiv 1804.09269
A Deep Learning based Approach to Reduced Order Modeling
for Turbulent Flow Control using LSTM Neural Networks
https://arxiv.org/abs/1804.09269
Arvind T. Mohan & Datta V. Gaitonde
The Ohio State University


1) Dataset: Please check the paper for details on dataset and how it was processed to obtain POD time coefficients. The data in this
repo consists of the POD \alpha_{t} computed from the datasets obtained at http://turbulence.pha.jhu.edu/.
All scripts for downloading slices of the DNS datasets can be a found at the link above.

2) The codes (.py files and Ipython notebooks) are provided to be run directly from the directory

3) The training folder contains codes to train the LSTM/BiLSTM model for the given input

4) Analysis folder contains Ipython notebooks to visualize results of the LSTM/BiLSTM models.


The .h5 are the trained models which can be directly used for modeling from the Ipython notebooks.
The .npz files are compressed files which show the loss vs training epochs - can be visualized from the
analysis codes to study training performance.

This is a research code, and not production level software :) - so please feel free to leave me a note for comments, improvements/suggestions etc. Also, if you find the codes useful in your research please cite our work.

About

A Deep Learning based Approach to Reduced Order Modeling for Turbulent Flow Control using LSTM Neural Networks. arXiv:1804.09269

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 98.4%
  • Python 1.6%