Motion Control of a Robot Fish Based on CPG

Abstract

In this paper, we developed a robot fish consisting of a fish body, a two-linked fish tail and two 2-DOF pectoral fins. In order to make robot fish detect obstacles and perceive its postures, three infrared sensors and a gyro are fixed on the fish body. And an artificial CPG controller consists of four neural oscillators with sensory information feedback have been designed. This controller can generate rhythm instructions that can drive pectoral fins and tail fin flapping repeatedly. According to the relationship between motion patterns and parameters of CPG model, two feedback strategies are designed to control pitching and turning locomotion respectively. Simulations we have conducted show the applicability and validity of CPG model and feedback strategies