Date of Award

1993

Degree Name

M.S. in Electrical Engineering

Abstract

Classical feedback control theories are traditionally concerned with issues like stability and performance, however, they typically fail to address issues such as robustness and plant perturbation. This thesis is concerned with the robust stability and the robust performance of single-input single-output plants. The basic issue under analysis is how to realize the benefits of the usual feedback control structure in the presence of model uncertainty. This is accomplished by seeking feedback controllers providing robust stability and performance by minimizing weighted sensitivity functions of a linear system represented by its transfer function. A characterization of models for plants with unstructured uncertainty is introduced. Specifications and measures of stability and performance for robust controllers and the necessary and sufficient conditions to test the robust stability and the robust performance conditions of a control design are explored. A parametrization of feedback controllers that guarantee closed loop stability for both stable and unstable plants is shown and a systematic procedure for synthesizing robust controllers, known in the literature as HK controllers, is presented. These systematic algorithms are based on the theory of interpolation by analytic functions and the solution to the model matching problem. A case study of the inverted pendulum positioning system is developed to illustrate the concepts of robust analysis and the design algorithms. The controller is compared to a classic state variable feedback solution.

Keywords

Feedback control systems

Rights Statement

Copyright © 1993, author

Share

COinS