Direct Synthesis of Reduced-Order Controllers with Integrating Action for a Class of Norm-Bounded Uncertain Plants

Reduced-order controller design for uncertain plants is, in general, iterative in nature. This paper identifies a class of multivariable plant transfer functions having uncertain parameters for which, a reduced-order controller with integrating action can be designed following a direct synthesis method recently reported in literature for fixed multivariable plants. The features of observability canonical state-space representation of single-input-single-output plants are exploited to obtain such a class of multivariable plants. The plants are found to have (i) no finite zeros in the element transfer function, (ii) same pole polynomial in each element transfer function of a row, and (iii) order equal to the sum of the orders of all row pole polynomials. The proposed control design methodology is applied to a laboratory-based two degree-of-freedom helicopter system having uncertain parameters. The robust performance behavior of the proposed control is validated through real-time experiments.