Optimal trajectories for the preview control of dual-stage actuators

Preview Control for Dual-Stage Actuators (DSA) consists in allowing movement of the slow actuator while maintaining the system output at the reference point. This strategy is possible only if the fast actuator is able to compensate the movement of the slow one. The advantages of this control scheme are related to an improvement on the settling time of the output, a consequence of the fact that the slow actuator is allowed to move ahead of the reference transition time. This paper will discuss fundamental limitations that constrain the trajectories of the primary actuator to a feasible set, i.e., a set whose trajectories the secondary actuator is able to effectively compensate. From this initial discussion optimal trajectories will be devised via quadratic programming. Experimental results show the effectiveness of the proposed design