Design and implementation of a novel CPG-based locomotion controller for robotic dolphin

Abstract

This paper presents a novel locomotion controller based on the Central Pattern Generator (CPG) model with nearest coupling for a multilink robotic dolphin. Each CPG in the model is modeled with two input saturation functions, a nonlinear oscillating neuron, and an output amplification function. Its oscillating frequency and amplitude are independently modulated. The CPG model can be amenable to rapid computation for the real-time control and to generate sufficient rhythmic patterns for dolphin-like swimming. The CPG model can be easily integrated into a microcontroller without extra hardware burden via an iterative algorithm. Design and implementation issues of the CPG-based locomotion controller are brought forth and further applied to achieve multimodal swimming. Experimental results confirm the effectiveness of the CPG-based locomotion controller in dolphin-like propulsion and maneuvering.