Abstract
This paper addresses the control problem of multi-degree-of-freedom actuators with uncertain torque models and unmodelled system parameters, such as the rotor’s moments of inertia. The control scheme relies on (i) the parameterization of the system nonlinearity in terms of two regression functions and the unknown parameters, and (ii) a control structure with filtered input torque and projection-based adaptation laws. The input–output mapping analysis rigorously shows that the proposed controller maintains the bounded deviation of the control system from a stable, non-adaptive reference system. The size of the deviation is inversely proportional to the filter bandwidth. Simulations using a standard model of spherical motors illustrate the tracking performance of the controlled rotor’s orientations.
Get full access to this article
View all access options for this article.
