skip to main content

This content will become publicly available on December 1, 2022

Title: A numerical study of fish adaption behaviors in complex environments with a deep reinforcement learning and immersed boundary–lattice Boltzmann method
Abstract Fish adaption behaviors in complex environments are of great importance in improving the performance of underwater vehicles. This work presents a numerical study of the adaption behaviors of self-propelled fish in complex environments by developing a numerical framework of deep learning and immersed boundary–lattice Boltzmann method (IB–LBM). In this framework, the fish swimming in a viscous incompressible flow is simulated with an IB–LBM which is validated by conducting two benchmark problems including a uniform flow over a stationary cylinder and a self-propelled anguilliform swimming in a quiescent flow. Furthermore, a deep recurrent Q-network (DRQN) is incorporated with the IB–LBM to train the fish model to adapt its motion to optimally achieve a specific task, such as prey capture, rheotaxis and Kármán gaiting. Compared to existing learning models for fish, this work incorporates the fish position, velocity and acceleration into the state space in the DRQN; and it considers the amplitude and frequency action spaces as well as the historical effects. This framework makes use of the high computational efficiency of the IB–LBM which is of crucial importance for the effective coupling with learning algorithms. Applications of the proposed numerical framework in point-to-point swimming in quiescent flow and position holding more » both in a uniform stream and a Kármán vortex street demonstrate the strategies used to adapt to different situations. « less
; ; ; ;
Award ID(s):
Publication Date:
Journal Name:
Scientific Reports
Sponsoring Org:
National Science Foundation
More Like this
  1. Some anguilliform swimmers such as eels and lampreys swim near the ground, which has been hypothesized to have hydrodynamic benefits. To investigate whether swimming near ground has hydrodynamics benefits, two large-eddy simulations of a self-propelled anguilliform swimmer are carried out—one swimming far away from the ground (free swimming) and the other near the ground, that is, midline at 0.07 of fish length (L) from the ground creating a gap of 0.04 L . Simulations are carried out under similar conditions with both fish starting from rest in a quiescent flow and reaching steady swimming (constant average speed). The numerical resultsmore »show that both swimmers have similar speed, power consumption, efficiency, and wake structure during steady swimming. This indicates that swimming near the ground with a gap larger than 0.04 L does not improve the swimming performance of anguilliform swimmers when there is no incoming flow, that is, the interaction of the wake with the ground does not improve swimming performance. When there is incoming flow, however, swimming near the ground may help because the flow has lower velocities near the ground.« less
  2. Fiber-based flexible piezoelectric composites with interdigitated electrodes, namely Macro-Fiber Composite (MFC) structures, strike a balance between the deformation and actuation force capabilities for effective underwater bio-inspired locomotion. These materials are also suitable for vibration-based energy harvesting toward enabling self-powered electronic components. In this work, we design, fabricate, and experimentally characterize an MFC-based bio-inspired swimmer-energy harvester platform. Following in vacuo and in air frequency response experiments, the proposed piezoelectric robotic fish platform is tested and characterized under water for its swimming performance both in free locomotion (in a large water tank) and also in a closed-loop water channel under imposed flow.more »In addition to swimming speed characterization under resonant actuation, hydrodynamic thrust resultant in both quiescent water and under imposed flow are quantified experimentally. We show that the proposed design easily produces thrust levels on the order of biological fish with similar dimensions. Overall it produces thrust levels higher than other smart material-based designs (such as soft material-based concepts), while offering geometric scalability and silent operation unlike large scale robotic fish platforms that use conventional and bulky actuators. The performance of this untethered swimmer platform in piezoelectric energy harvesting is also quantified by underwater base excitation experiments in a quiescent water and via vortex induced-vibration (VIV) experiments under imposed flow in a water channel. Following basic resistor sweep experiments in underwater base excitation experiments, VIV tests are conducted for cylindrical bluff body configurations of different diameters and distances from the leading edge of the energy harvesting tail portion. The resulting concept and design can find use for underwater swimmer and sensor applications such as ecological monitoring, among others.« less
  3. We investigate the self-propulsion of an inertial swimmer in a linearly density stratified fluid using the archetypal squirmer model which self-propels by generating tangential surface waves. We quantify swimming speeds for pushers (propelled from the rear) and pullers (propelled from the front) by direct numerical solution of the Navier–Stokes equations using the finite volume method for solving the fluid flow and the distributed Lagrange multiplier method for modelling the swimmer. The simulations are performed for Reynolds numbers ( $Re$ ) between 5 and 100 and Froude numbers ( $Fr$ ) between 1 and 10. We find that increasing the fluidmore »stratification strength reduces the swimming speeds of both pushers and pullers relative to their speeds in a homogeneous fluid. The increase in the buoyancy force experienced by these squirmers due to the trapping of lighter fluid in their respective recirculatory regions as they move in the heavier fluid is one of the reasons for this reduction. With increasing the stratification, the isopycnals tend to deform less, which offers resistance to the flow generated by the squirmers around them to propel themselves. This resistance increases with stratification, thus, reducing the squirmer swimming velocity. Stratification also stabilizes the flow around a puller keeping it axisymmetric even at high $Re$ , thus, leading to stability which is otherwise absent in a homogeneous fluid for $Re$ greater than $O(10)$ . On the contrary, a strong stratification leads to instability in the motion of pushers by making the flow around them unsteady and three-dimensional, which is otherwise steady and axisymmetric in a homogeneous fluid. A pusher is a more efficient swimmer than a puller owing to efficient convection of vorticity along its surface and downstream. Data for the mixing efficiency generated by individual squirmers explain the trends observed in the mixing produced by a swarm of squirmers.« less
  4. Abstract

    Through billions of years of evolution, microorganisms mastered unique swimming behaviors to thrive in complex fluid environments. Limitations in nanofabrication have thus far hindered the ability to design and program synthetic swimmers with the same abilities. Here we encode multi-behavioral responses in microscopic self-propelled tori using nanoscale 3D printing. We show experimentally and theoretically that the tori continuously transition between two primary swimming modes in response to a magnetic field. The tori also manipulated and transported other artificial swimmers, bimetallic nanorods, as well as passive colloidal particles. In the first behavioral mode, the tori accumulated and transported nanorods; inmore »the second mode, nanorods aligned along the toriʼs self-generated streamlines. Our results indicate that such shape-programmed microswimmers have a potential to manipulate biological active matter, e.g. bacteria or cells.

    « less
  5. Wirelessly powered and controllable microscale propulsion in 3-D space is of critical importance to micro swimming drones serving as an active and maneuverable in vivo cargo for medical uses. This aritcle describes a 3-D micro swimming drone navigating in 3-D space, propelled by unidirectional microstreaming flow from acoutsically oscillating bubbles. 3-D propulsion is enabled by multiple bubbles with different lengths embedded in different orientations inside the drone body. Each bubble generats propulsion by applying acoustic field at its resonance frequency. Therefore, 3-D propulsion in any direction is achievable by resonating bubbles individually or jointly. The drone with such a complexmore »design was fabricated by a two-photon polymerization 3-D printer. For stable maneuverability, a non-uniform mass distribution of the drone is designed to restore the drone to the designated posture under any disturbances. The restoration mechanism is formulated by a mathematical model, predicting the restoring time and shows an excellent agreemnt with the experimental results. This 3-D micro swimning drone proves its robustness as a manueverable microrobot navigating along programmble path in a 3-D space through selective and joint actuation of microbubbles.« less