Controller Reduction Methods Maintaining Performance and Robustness

n reducing high order controller des~gns, such ns arise from }{m or LQG techniques, to more practical low order ones, a re~onable objective is to preserve the controller robustness fperforlna”ce properties. In this paper, standard balanced truncation or Hanke] norm model approximation methods are applied toaugmentatimso fthec ontroller which emerge when characterizing theclas.s ofallstabilizing controllersin ternwof an arbitrary pmpers table transfer function. In the method, scaling par.smeters are at the disposal of the engineer to achieve an appropriate compromise b.etwee” prtxerving performance for the nominal plant and a certain type of robustnes to plant variations. There are a number of unique features of tbe approach of tbe paper. One feature is that a straightforward re-optimiz ation of a reduced-order controller is possibl ewithinth eframewor koftlmmetl]od. Asecond feature of the paper is that for controllers designed for simultaneous stabilization of a number of plants, the method seeks to preserve the performance/robustness of the reduced order controller for each plant.