Reference governor based model predictive contouring control for biaxial systems

This paper aims at improving the contouring accuracy in biaxial position control systems. A quadratic cost functional consisting of the contour error is formulated including both axes of the system. Linear constraints on the motor voltages and currents are considered. Modeling of the driving motors including the often neglected current dynamics enables precise position control. Even constraints on the contour error can be considered, e.g. to meet manufacturing tolerances. This so called model predictive contouring control is employed in a control structure consisting of a reference governor which adapts the path reference in order to meet constraints and a subordinate unconstrained controller for precise tracking of the reference. This flexible structure is simple to tune and makes it possible to decouple the computationally complex optimization from the subordinate controller.