Dynamics and Control of Bioinspired Swimming, Schooling, and Pursuit

Thumbnail Image

Publication or External Link





Understanding the benefits of the behaviors of aquatic animals can improve the capabilities of robotic systems. Aquatic species such as the zebrafish swim with discrete motions that alternate between perception and action while avoiding predators and swimming in schools, and other species such as the lionfish use their pectoral fins to herd and trap prey. This work seeks to model these bioinspired behaviors (i.e., schooling, swimming with intermittent sensing and actuation, and pursuit and evasion in a structured environment) and enhance our understanding of their benefits.

A hybrid dynamic model is derived with two phases; namely a burst phase during which each particle applies a control input and a coast phase during which each particle performs state estimation. This model provides a way to investigate how having non-overlapping sensing and control affects a multi-agent system's ability to achieve collective behavior such as steering to some desired direction. By evaluating the stability properties of the equilibrium points for the collective behavior, investigators can determine parameter values that exhibit exponentially stable behavior.

Aside from swimming intermittently, fish also need to avoid predators. Inspired by observations of predation attempts by lionfish (Pterois sp.), a pursuit-evasion game is derived in a bounded environment to study the interaction of an advanced predator and an intermittently steering prey. The predator tracks the prey with a pure-pursuit strategy while using a bioinspired tactic to minimize the evader's escape routes, i.e, to trap the prey. Specifically, the predator employs symmetric appendages inspired by the large pectoral fins of lionfish, but this expansion increases its drag. The prey employs a bioinspired randomly-directed escape strategy to avoid capture and collisions with the boundary known as the protean strategy. This game investigates the predator's trade-off of minimizing the work to capture the prey and minimizing the prey's escape routes. Using the predator's expected work to capture as a cost function determines when the predator should expand its appendages as a function of the relative distance to the evader and the evader's proximity to the boundary.

Prey fish also swim in schools to protect themselves from predators. To drive a school of fish robots into a parallel formation, a nonlinear steering controller is derived and implemented on a robotic fish platform. These robotic fish are actuated with an internal reaction wheel driven by a DC motor. Implementation of the proposed parallel formation control law on an actual school of soft robotic fish is described, including system identification experiments to identify motor dynamics and the design of a motor torque-tracking controller to follow the formation torque control. Experimental results demonstrate a school of four robotic fish achieving parallel formations starting from random initial conditions.