Optimal dynamic design of anthropomorphic robot module with redundant actuators

Abstract

In this paper, a study on the optimal dynamic design for an anthropomorphic robot module with redundant actuators is performed. Musculoskeletal structure of human body is a typical example of redundantly actuated mechanism, and provides superior features than general robotic mechanisms. An anthropomorphic robot module that resembles the structure of human upper limb is introduced to utilize the advantages of redundant actuation system. Optimal dynamic design of the proposed robot module that follows optimal kinematic design is carried out to maximize the advantages. Five design indices are introduced, which are associated with inertia matrix, inertia power array representing nonlinear terms and gravity terms of the dynamic modeling equation. A concept of composite design index based on max-min principle of fuzzy theory is employed to deal with multi-criteria based design. As a result of dynamic optimization, a set of dynamic parameters, representing optimal mass distribution of the manipulator is obtaind. It is shown that the dynamic optimization yields a notable enhancement in dynamic performances, as compared to the case of kinematic optimization only.