Averaged Modeling of Pectoral Fin-Actuated Robotic Fish

Abstract

Pectoral fins play an important role in the locomotion and maneuvering of robotic fish. Considering the cyclic nature of typical actuation modes, it is of interest to develop a dynamic average model that is amenable to controller design, where the control inputs are actuation pattern parameters. In this work, we propose a scaling-based approach to develop a nonlinear dynamic average model for a robotic fish propelled by a pair of rowing pectoral fins. In particular, the fin-generated hydrodynamic forces and moment, modeled using blade element theory, are scaled with functions of the fin-beat parameters, and classical averaging is then conducted over the corresponding modified dynamics. To determine proper scaling functions with minimal complexity, we propose a novel estimation scheme employing a nonlinear model predictive control formulation paired with a multivariate nonlinear regression scheme. Experimental and simulation results comparing the predictions from the dynamic and averaged models are presented to support the efficacy of the averaged modeling approach.