ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Machine learning strategies for path-planning microswimmers in turbulent flows

Alageshan, JK and Verma, AK and Bec, J and Pandit, R (2020) Machine learning strategies for path-planning microswimmers in turbulent flows. In: Physical Review E, 101 (4).

[img] PDF
Phy_Rev_E_101_2020.pdf - Published Version
Restricted to Registered users only

Download (2MB) | Request a copy
Official URL: https://dx.doi.org/10.1103/PhysRevE.101.043110


We develop an adversarial-reinforcement learning scheme for microswimmers in statistically homogeneous and isotropic turbulent fluid flows, in both two and three dimensions. We show that this scheme allows microswimmers to find nontrivial paths, which enable them to reach a target on average in less time than a naïve microswimmer, which tries, at any instant of time and at a given position in space, to swim in the direction of the target. We use pseudospectral direct numerical simulations of the two- A nd three-dimensional (incompressible) Navier-Stokes equations to obtain the turbulent flows. We then introduce passive microswimmers that try to swim along a given direction in these flows; the microswimmers do not affect the flow, but they are advected by it. Two nondimensional control parameters play important roles in our learning scheme: (a) the ratio Ṽs of the microswimmer's bare velocity Vs and the root-mean-square (rms) velocity urms of the turbulent fluid and (b) the product B of the microswimmer-response time B and the rms vorticity �rms of the fluid. We show that the average time required for the microswimmers to reach the target, by using our adversarial-reinforcement learning scheme, eventually reduces below the average time taken by microswimmers that follow the naïve strategy.

Item Type: Journal Article
Publication: Physical Review E
Publisher: American Physical Society
Additional Information: Copyright for this article belongs to American Physical Society
Keywords: Learning systems; Motion planning; Reinforcement learning; Turbulent flow, Control parameters; Homogeneous and isotropic; Learning schemes; Pseudospectral direct numerical simulation; Root Mean Square; Three dimensions; Turbulent fluid flow; Turbulent fluids, Navier Stokes equations
Department/Centre: Division of Physical & Mathematical Sciences > Physics
Date Deposited: 19 Jun 2020 09:12
Last Modified: 19 Jun 2020 09:12
URI: http://eprints.iisc.ac.in/id/eprint/65498

Actions (login required)

View Item View Item