Biological Modeling Control of a Multilegged Walking Robot

Abstract

This article focus on the design of a biomimetic control architecture, which can improve the capability of agility and independence of multilegged walking robots. Since this kind of robots are contrived from multiped animals, the thinking and behaviour manners of multiped animals are first researched and are classified into several types, which are the theoretical base of the biological model. The proposed conceptual model and physical model of the biomimetic control architecture, which are software hardware codesigned, are made up of four blocks, that is information sensing, information communication, behavior generation, and joint driver blocks. Furthermore, the architecture model can realize the evolution between different behavior control layers based various behaviour manners. Through simulation and experiment, it is proved that this biomimetic model makes multilegged walking robots adapt the irregular environment well