Design of an artificial bionic neural network to control fish-robot's locomotion

Abstract

An artificial bionic neural network to control fish-robot locomotion is presented. The fish-robot, which is inspired from "Gymnarchus niloticus", is modeled as a multi-joint dynamic system with parallel connections and composed of several motors. The principle of the central pattern generators (CPGs) governing the locomotion of fish is analyzed. An artificial neural network which has many comparability with the biological CPGs is designed to control the fish-robot. Experimental results of startup, stop, forward swimming and backward swimming show its validity and efficiency. © 2007 Elsevier B.V. All rights reserved.