Underactuated Robotic Fish Control: Maneuverability and Adaptability Through Proprioceptive Feedback

Abstract

Bioinspired robots are a promising technology for minimizing environmental disruption during underwater inspection and exploration, as well as for improving animal farming conditions and protecting wildlife. In this research, we propose a control strategy for an underactuated robotic fish that mimics the oscillatory movement of a real fish's tail using only one DC motor. Our control strategy is bioinspired by Central Pattern Generators (CPGs) and integrates proprioceptive sensory feedback. Specifically, we incorporated the angular position of the tail as an input control variable to enhance the feedback loop of the CPGs. This enables the controller to adapt to changes in the tail structure, weight, or the environment in which the robotic fish swims, allowing it to change its swimming speed and steering angular speed. Our robotic fish can swim at a speed between 0.18 and 0.26 Body Length per second (BL/s), with a tail beating frequency between 1.7 and 2.3 Hz. It can also vary its steering angular speed in the range of 0.08 rad/s, resulting in a relative change in the curvature radius of 0.25 m. With modifications to the modular design, we can further improve the speed and steering performance while maintaining the developed control strategy. This research highlights the potential of bioinspired robotics to address pressing environmental challenges while improving solutions efficiency, reliability and reducing development costs.