A pytorch implementation of PINN-for-NS-eqation
This repository is actually a quite remote version of my code, but people have been constantly asking questions under this repository, so I decide to update this old repository.
Here I present an interesting example of reconstructing flow field of 2D unsteady flow past a circular cylinder at Re=3900. As you can see, PINNs can reconstruct the wake flow with only 36 measurement points available. The full dataset cylinder_Re3900_ke_all_100snaps.mat can be found and downloaded at https://drive.google.com/drive/folders/1thZ8X0d-DFc-bgGcw7G76UCBJSdsbfFl?usp=sharing.
The annotates are both in Chinese and English
You can plot comparsion pics and gifs in plot.py
Xu S, Sun Z, Huang R, et al. A practical approach to flow field reconstruction with sparse or incomplete data through physics informed neural network[J]. Acta Mechanica Sinica, 2023, 39(3): 322302.
Xu S, Yan C, Zhang G, et al. Spatiotemporal parallel physics-informed neural networks: A framework to solve inverse problems in fluid mechanics[J]. Physics of Fluids, 2023, 35(6).
Raissi M, Perdikaris P, Karniadakis G E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378: 686-707.