Optimal Robust Controllers for Multibody Systems

Abstract

The proposed paper addresses control design problems for multibody systems (MBS) like robot manipulators or mechanical structures with active vibration damping. When performing motion tasks, such systems may be subject to even severe disturbances. A control design approach for motion stabilisation has to meet the increasing demands for faster response, higher position accuracy, and reduced energy consumption. A central role in solving such a complicated control optimisation problem plays the matrix that transfers the control inputs into mechanical accelerations. For MBS having as many control forces as controlled outputs, simple conditions on that matrix are found to be necessary and sufficient for such systems to be controllable in the presence of bounded random disturbances. There are proposed optimal trade-off relations for designing decentralised controllers with maximum degree of robustness. An interesting extension of these concepts to the important class of over-controlled MBS is proposed. Examples with a car body suspension and an elastic-joint manipulator are presented to show how the proposed control design approach can be applied and developed.