Abstract
Disturbance compensation control is one of the useful methods for dynamic control of robot manipulators, in which the disturbance observer estimates interference torques such as friction and modelling error as the disturbance. This disturbance torque can be compensated to match the system dynamics with the model. The disturbance consists of external disturbances and that caused by perturbation such as system parameter change and unknown dynamics in the manipulator. The former can be suppressed by using the conventional disturbance compensation control, however, because the latter forms an inner feedback loop, its effects to the system stability and control performance should be considered. In addition, when a dynamic control method is used with the disturbance compensation control, interference between those two controls should also be taken into account.
In this paper, we consider the disturbance compensation control in joint-space. Especially the filter employed in the feedback loop of the estimated disturbance torque is designed by using μ-synthesis to make the system robust stable. Further, from the viewpoint of μ-analysis, we discuss the total system stability and control performance when the dynamic hybrid control method is used with the disturbance compensation control. We show that the combination of these two control methods can achieve high performance hybrid control even though the exact information about the manipulator dynamics is not obtained. Some numerical simulation results are provided to confirm the theoretical development.
