Learning to school in the presence of hydrodynamic interactions

Abstract

Schooling, an archetype of collective behavior, emerges from the interactions of fish responding to visual and other informative cues mediated by their aqueous environment. In this context, a fundamental and largely unexplored question concerns the role of hydrodynamics. Here, we investigate schooling by modeling swimmers as vortex dipoles whose interactions are governed by the Biot-Savart law. When we enhance these dipoles with behavioral rules from classical agent based models we find that they do not lead robustly to schooling due to flow mediated interactions. In turn, we present dipole swimmers equipped with adaptive decision-making that learn, through a reinforcement learning algorithm, to adjust their gaits in response to non-linearly varying hydrodynamic loads. The dipoles maintain their relative position within a formation by adapting their strength and school in a variety of prescribed geometrical arrangements. Furthermore, we identify schooling patterns that minimize the individual and the collective swimming effort, through an evolutionary optimization. The present work suggests that the adaptive response of individual swimmers to flow-mediated interactions is critical in fish schooling.

18 pages, 12 figures

References in corpus (2)

• Novel type of phase transition in a system of self-driven particles
• Purely hydrodynamic ordering of rotating disks at a finite Reynolds number

Cited by in corpus (42)

• Machine Learning for Fluid Mechanics
• Flow Navigation by Smart Microswimmers via Reinforcement Learning
• Robust active flow control over a range of Reynolds numbers using an artificial neural network trained through deep reinforcement learning
• Model of collective fish behavior with hydrodynamic interactions
• Recent advances in applying deep reinforcement learning for flow control: perspectives and future directions
• Synchronised Swimming of Two Fish
• Zermelo's problem: Optimal point-to-point navigation in 2D turbulent flows using Reinforcement Learning
• A review on Deep Reinforcement Learning for Fluid Mechanics
• A review on deep reinforcement learning for fluid mechanics: an update
• Classifying vortex wakes using neural networks
• Controlling Rayleigh-Bénard convection via Reinforcement Learning
• Finding Efficient Swimming Strategies in a Three Dimensional Chaotic Flow by Reinforcement Learning
• DRLinFluids -- An open-source python platform of coupling Deep Reinforcement Learning and OpenFOAM
• Self-learning how to swim at low Reynolds number
• Learning to swim in potential flow
• Smart Inertial Particles
• Learning to Control Active Matter
• School cohesion, speed, and efficiency are modulated by the swimmers flapping motion
• Optimising low-Reynolds-number predation via optimal control and reinforcement learning
• Lattices of hydrodynamically interacting flapping swimmers
• Swimming strategy of settling elongated microswimmers by reinforcement learning
• Active Cloaking in Stokes Flows via Reinforcement Learning
• Streaming enhanced flow-mediated transport
• Shape curvature effects in viscous streaming
• Reciprocal swimming at intermediate Reynolds number
• Multi-curvature viscous streaming: flow topology and particle manipulation
• Discovering dynamic laws from observations: the case of self-propelled, interacting colloids
• Vortex phase matching of a self-propelled model of fish with autonomous fin motion
• Hunting active Brownian particles: Learning optimal behavior
• Long-distance migration with minimal energy consumption in a thermal turbulent environment
• An unrecognized force in inertial microfluidics
• Swimming through parameter subspaces of a simple anguilliform swimmer
• Efficient collective swimming by harnessing vortices through deep reinforcement learning
• Learning to swim efficiently in a nonuniform flow field
• Stabilising viscous extensional flows using Reinforcement Learning
• Deep reinforcement learning for tracking a moving target in jellyfish-like swimming
• A critical assessment of reinforcement learning methods for microswimmer navigation in complex flows
• On the stability of an in-line formation of hydrodynamically interacting flapping plates
• Pattern stabilisation in swarms of programmable active matter: a probe for turbulence at large length scales
• Learning swimming escape patterns for larval fish under energy constraints
• A reduced-scale autonomous morphing vehicle prototype with enhanced aerodynamic efficiency
• Symmetry reduction for deep reinforcement learning active control of chaotic spatiotemporal dynamics