A COMPUTATIONAL PROCEDURE FOR MULTIVARIABLE STATE FEEDBACK ROBUST CONTROLLER DESIGN

Document Type : Original Article

Author

Lecturer, Dept. of Engineering Mathematics & Physics, Faculty of Engineering Alexandria University, Alexandria, Egypt.

Abstract

This paper describes a procedure for multivariable state feedback robust controller design. The plant in state space is given by the operational point description or in terms of a vector of slow varying physical parameters. Through solution of the Sylvester matrix equation, a nonunique static feedback controller, which assigns the prespecified closed-loop spectrum, is calculated. In addition, all the remaining feedback degrees of freedom are utilized to optimize a multiobjective function that reflects further design properties. The robust feedback gains is calculated through a three-phase computational algorithm. Numerical examples show that under the robust state feedback control, the closed-loop systems can both achieve satisfied transient characteristics and greatly reduce state trajectory sensitivity to small or large parameter variations in the plant. The pro-posed procedure is still applied to a VTOL aircraft model.