Optimal tracking strategies in a turbulent flow

We show how to apply optimal control theory to catch a passive drifting target in a turbulent flow by an autonomous flowing agent with limited maneuverability. For the case of a perfect knowledge of the environment, we show that optimal control theory can overcome chaotic dispersion capturing the Lagrangian target in the shortest possible time [1]. We also provide baselines using heuristic strategies based on local-only hydrodynamic cues. How to extend this approach to model-free data-driven tools is also briefly discussed [2]. One possible application of the present results is the control of microswimmers/minirobots at small scales, while it may be of interest to extend the present work to multi-agent systems.

Data are open downloadable from TURB-Lagr [3], a database of more than 300K three-dimensional trajectories of tracer particles advected by a fully developed homogeneous and isotropic turbulent flow.

[1] Calascibetta, C., Biferale, L., Borra, F., Celani, A. and Cencini, M. Optimal tracking strategies in a turbulent flow - arXiv preprint arXiv:2305.04677, (2023).

[2] Calascibetta, C., Biferale, L., Borra, F., Celani, A. and Cencini, M. Taming Lagrangian chaos with multi-objective reinforcement learning. Eur. Phys. J. E 46, 9 (2023).

[3] https://smart-turb.roma2.infn.it/