Using Fast Enumeration to Re-Discover The Navier-Stokes Equations from Lattice Boltzmann Simulation
ORAL
Abstract
Theoretically, laws of statistical mechanics imply laws of hydrodynamics. The Lattice Boltzmann Method (LBM), an algorithm built on the principles of statistical mechanics, provides an alternative way to simulate fluid flow. This paper shows that using LBM simulation of 2D low Reynolds number flow past a cylinder, the machine is able to re-discover the Navier-Stokes equations for incompressible flow within 2 seconds. This is done by TheoSea, a Julia program that uses fast enumeration and singular value decomposition to re-discover smooth, differential theories from data.
*This research is funded by the Gordon and Betty Moore Foundation through Grant GBMF4915 to the Caltech Center for Data-Driven Discovery.
–
Presenters
-
Chao Ju
- Minerva Schools at KGI