Building a detailed flow model of acity using Domain Decomposition, Convolutional Autoencoders and Adversarial Networks
ACSE-9 Independent Research Project
MSc Applied Computational Science and Engineering
Imperial College London
This project is a non-intrusive reduced order modelling (NIROM) on flows around buildings. High-fidelity representations of flows are compressed and modelled. In particular, proper orthogonal decomposition (POD) and convolutional autoencoders (CAE) are compared as methods for reducing order of the data, and an predictive adversarial network (PAN)(made by Zef Wolffs) is used to predict the compressed flow representations. Domain decopmosition is applied to predict on a larger domain using subdomains.
This project is developed in Python. Packages scipy, numpy, matplotlib, sklearn, tensorflow, keras, vtu etc. used.
To get a local copy up and running follow these simple steps.
This is an example of how to list things you need to use the software and how to install them.
- npm
npm install npm@latest -g
- Clone the repo
git clone https://github.com/github_username/repo_name.git
- Install NPM packages
npm install
Use this space to show useful examples of how a project can be used. Additional screenshots, code examples and demos work well in this space. You may also link to more resources.
For more examples, please refer to the Documentation
See the open issues for a list of proposed features (and known issues).
Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.
- Fork the Project
- Create your Feature Branch (
git checkout -b feature/AmazingFeature) - Commit your Changes (
git commit -m 'Add some AmazingFeature') - Push to the Branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Distributed under the MIT License. See LICENSE for more information.
Hanna Go - hg2917@ic.ac.uk / hannago2917@gmail.com
Project Link: https://github.com/acse-hg2917/DDGAN_buildings
I would like to thank my supervisors Dr.Claire Heaney and Prof.Christopher Pain for giving the opportunity to investigate this topic and providing great guidance. I would also like to thank our DD-GAN group, especially Xiangqi Liu for working together on how to approach the buildings project and Zef Wolffs and Jon Tomasson for providing the WGAN and PAN constructions.