Bioinspired Neural-Based Control of Flexible Fish-Like Propulsors

Abstract

Controlling the motion of soft robotics systems can be a challenging task and researchers have turned to biology to shed light on this complex problem. In particular, researchers have used mathematical models of central pattern generators (CPGs) to control the periodic movement of robotic systems. A useful property of CPGs is that their output can be altered by sensory feedback. However, questions remain about how to acquire sensory feedback from a compliant structure, and how to use sensory feedback to produce desired performance. This chapter focuses on neural-based control of a flexible fish-like propulsor (fin) with spatially varying mechanical properties. We show how manipulation of the sensory feedback signal and architecture affect the entrainment state (i.e., the frequency and amplitude). Specifically, we explore how the location of the feedback signal and the sensory feedback topology affect the entrainment characteristics of the fin. Additionally, we show how the settings of the coupled system affect the resonance-entrainment characteristics of a fin. Our results offer a way in which to collect sensory information and use this information to modulate the entrainment state of a flexible fish-like propulsor.