CaSToRC, the HPC National Competence Centre,
 invites you to the EuroCC and SimEA Seminar Series

Abstract

We examine the applicability of Artificial Intelligence tools to different open problems in fluid dynamics​:​ the search ​of a source of contaminants, ​optimal navigation strateg​ies in complex environments ​and​ data reconstruction from partial measurements of complex (turbulent) flows. To solve navigation​ and search​ problems we follow a Reinforcement Learning (RL) and Baysean a​pproaches. Here, we will focus on the problem of finding the path that minimizes the navigation time between two given points in a fluid flow​ and how to identify the source of odors in the presence of partial information.​  I will show how RL is able to take advantage of the flow properties in order to reach its target, providing stable solutions with respect to perturbations on the initial conditions and to addiction of external noise.

Concerning Data Assimilation (DA), we explore the capability of Generative Adversarial Network (GAN) to generate missing data. In this direction, I will present a quantitative investigation of their potential in reconstructing 2d damaged snapshots extracted from a large numerical database of 3d turbulence in the presence of rotation. I will briefly compare GAN with different, well-known, data assimilation tools, such as Nudging, an equation-informed protocol, or Gappy POD, developed in the context of image reconstruction. I will also discuss how to use DA tools with a reverse engineering approach, to investigate theoretical questions like which features of the input flow data are required/"more important" in order to obtain a better full-field reconstruction.

Refs
[1] Reinforcement learning for pursuit and evasion of microswimmers at low Reynolds number​. ​F Borra, L Biferale, M Cencini, A Celani Physical Review Fluids 7 (2), 023103 - 2022
[2] Inferring Turbulent Parameters via Machine Learning​. ​M Buzzicotti, F Bonaccorso, L Biferale arXiv preprint arXiv:2201.00732 - 2022
[3] Reconstruction of turbulent data with deep generative models for semantic inpainting from TURB-Rot database​. ​M Buzzicotti, F Bonaccorso, PC Di Leoni, L Biferale Physical Review Fluids 6 (5), 050503 - 2021
[4] Synchronization to big data: Nudging the Navier-Stokes equations for data assimilation of turbulent flows​. ​PC Di Leoni, A Mazzino, L Biferale Physical Review X 10 (1), 011023 - 2020
[5] Zermelo’s problem: Optimal point-to-point navigation in 2D turbulent flows using reinforcement learning​. ​L Biferale, F Bonaccorso, M Buzzicotti, P Clark Di Leoni, K Gustavsson Chaos: An Interdisciplinary Journal of Nonlinear Science 29 (10), 103138  - 2019
[6] ​Optimal policies for Bayesian olfactory search in turbulent flows​. RA Heinonen, L Biferale, A Celani, M Vergassola​. . Submitted to Phys. Rev E. ​arXiv preprint arXiv:2207.04277​ - 2022

About the Speaker

Luca Biferale is a theoretical physicist, full professor at the Department of Physics of the University of Rome 'Tor Vergata' working on out-of-equilibrium statistical physics, high performance computing and data-driven tools applied to complex flows and complex fluids problems. He is the author of more than 280 scientific papers which collected more than 10000 citations.

He has been Divisional Associate Editor of Physical Review Letters, elected as fellow of the APS and Euroomech. He has been awarded two Advanced Grants from the ERC.

Download the Fall 2022 EuroCC & SimEA Seminar Series Programme here.

The EuroCC project has received funding from the European Union’s Horizon 2020 research and innovation programme grant agreement No. 951732

The SimEA project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 810660

Contact This email address is being protected from spambots. You need JavaScript enabled to view it.

View all CyI events.