Kinetic Energy Maximization on Elastic Joint Robots Based on Feedback Excitation Control and Excitation Limit Hypersurface

Abstract

This paper describes amethod of realizing high kinetic energy utilizing mechanical elasticity within the joint limit ranges of multiple-joint robots. By utilizing series elastic elements, a robot obtains high kinetic energy compared with a rigid robot. In this paper, we propose feedback excitation control that realizes high kinetic energy utilizing series elastic joints. Robot motion has to be kept within the joint limit range. We propose a control method based on an excitation limit hypersurface to realize robot motion within the joint limit range. The feasibility of the method was evaluated in experiments. We performed a ball throwing task as an application of the method.

Design and modal analysis of feedback excitation control system for vertical series elastic manipulator

“…The feedback excitation controller for horizontal systems is derived by using energy integral of which anchoring point is gear shaft angles θ G since the equilibrium point of horizontal series elastic systems is always equal to gear shaft angles [4]. On vertical systems, however, the equilibrium position θo is not equal to θ G in general.…”

“…It is expected to realize a motion that does not depend only on actuator power by utilizing them. In the case of a robot that has series elastic elements [1] as shown in Fig.1, for example, speeds of links becomes instantaneously higher than that of a rigid robot with same motors [2], [3], [4]. This means that it is possible to obtain high kinetic energy by using low power driving systems and it is useful for "explosive" tasks in robots such as pitching and jumping.…”

“…In our previous work [4], we proposed feedback excitation control and the excitation limit hypersurface to obtain high kinetic energy on multiple-joint series-elastic robots within joint limit ranges. It is difficult, however, to use this method for robots that is influenced by gravity since a horizontal robot is considered in [4].…”

“…It is difficult, however, to use this method for robots that is influenced by gravity since a horizontal robot is considered in [4]. Feedback excitation control should be extended to treat gravity.…”

“…By focusing on the energy increase, however, we obtain a feedback controller for nonlinear systems. In the case of an Ν-DOF manipulator, total energy increases by controlling the equilibrium position moving velocity of each joint 0 Oi based on the relative position of each joint 0 i -0 Oi [4].…”