New machine learning method predicts turbulence | FREE NEWS

New machine learning method predicts turbulence

Researchers at the University of Illinois have developed a new AI training method that can more accurately predict turbulence.

We don’t know how to mathematically write all turbulence. There are unknowns that cannot be represented on a computer. If you try to train an AI without considering the basic known equations of physics, it won’t work. We combined them and it worked.

Jonathan Freund, Professor and Head of Aerospace Engineering

When designing an airship or spacecraft, Freund said this method will help engineers predict whether the design will provide a safe flight.

They will be able to make changes, test a plane, for example, and see if the forecast has improved.

Anyone who wants to simulate physical phenomena can use this new method. You can just use our algorithm, load the data into the software. This is a method that allows for other unknown physics. And we can work with it in different planes.

Jonathan Freund, Professor and Head of Aerospace Engineering

The work was performed using a supercomputer at the National Center for Supercomputers at UIUC. This made the simulation faster and therefore more economical.

Freund said that while the work has been done at a research level, it could potentially affect the industry.

or as guest:
Comments: 0