Modeling neural control of robotic fish with pectoral fins using a CPG-based network

Abstract

A three-dimensional locomotion control method for robotic fish with pectoral fins is presented in this paper. The approach employed a central pattern generator (CPG) network model, which can generate stable locomotive rhythms. The existence of periodic solution of the model has been proved. Parameter influence on the solution is also analyzed. The rhythm generators of the model not only produce rhythms, but also alter their frequencies and amplitudes separately. Robotic fish driven by pectoral fins is capable of implementing various motion patterns, such as pitching mode, backward motion, turning mode, etc, by using the CPG network model. The concrete procedure for using the model is described in this paper. Corresponding experimental results validated the neural control method, which can be extended to other robotic applications.