Connecting parameter set estimation to robust control design

Abstract

The focus of this investigation is on the use of knowledge of an ellipsoidal bound on system parameters for the identification of modeling uncertainty information vital to the H/sub /spl infin// robust control designer. The characterization of parametric uncertainty as a coprime factor perturbation to a nominal plant model is addressed in a worst-case analysis; a scheme to interpolate samples of the perturbation bounding function is introduced. High order perturbation weightings complicate the computation of a robust controller; a technique for identification of reduced-order perturbation weightings, essential for practical control design, is presented. An example serves to illustrate the success of the developed technique.