Content area
Abstract
This thesis deals with the design of a feedback stabilizer for a linear time invariant plant. The focus of the development is on obtaining (a) a low order controller and (b) a controller that is robust i.e. preserves closed loop stability, with respect to perturbations of a physical parameter vector. Both these aspects of controller design have been largely overlooked in the control literature but are of fundamental importance in the design of practical systems of even a modest degree of complexity. The order of compensators designed by the available methods of modern control theory are unnecessarily high and forces the designer to work in a controller parameter space that is of excessively high dimension. This makes the task of readjusting the controller parameters to satisfy various design specifications prohibitive. We develop here a state space based algorithm and a transfer function based algorithm that attempt to alleviate this problem by designing low order stabilizers. The robustness results given here deal with structured perturbations in contrast with the theory available in the literature dealing with unstructured perturbations. The computational algorithms developed are illustrated by various examples and the many open areas of research are pointed out.
