ON MIMO OPTIMAL PREVIEW TRACKING CONTROL FOR KNOWN TRAJECTORY MODELS

A class of discrete time feedforward tracking controllers is presented for linear multivariable plants subjected to time-varying exogenous inputs with known models. The controller design is based on linear quadratic optimization techniques and the augmented system approach. The optimal controllers use state feedback for regulation purposes. For the tracking objective the future set-points along the reference trajectory vector are used. Necessary and sufficient conditions for the existence of an optimal feedforward solution are presented in terms of the relationships between the eigenvalues of the closed-loop system and those of the trajectory models. For illustration a numerical example is included which is based on a dual-stage positioning carriage modelled as a linear multivariable plant.