Control and Optimization of a Bionic Robotic Fish Through a Combination of CPG model and PSO

Abstract Swimming speed and propulsive efficiency are critical indicators of a self-propelled bionic robotic fish in terms of swimming performance. This paper is devoted to achieving a maximum swimming speed and a higher propulsive efficiency for a four-joint robotic fish. A Newton–Euler method based dynamic model in conjunction with a central pattern generator (CPG) network serves to theoretically estimate fish-like swimming performance. In particular, each CPG unit based on limit cycles serves to drive one joint of the robotic fish. The overall CPG model is able to generate rhythmic signals for multimode swimming. To obtain the maximum average speed, a particle swarm optimization (PSO) algorithm is further utilized to optimize the feature parameters of the CPG model. Furthermore, the higher propulsive efficiency is sought within the same control framework. Simulations and experiments on the actual robotic fish demonstrate the improved propulsive performance and the effectiveness of the proposed control framework.