An H formulation of the QFT approach to robust performance under significant uncertainties

Abstract

The main problem investigated in this research deals with the synthesis of controllers for achieving a desired robust performance in the presence of significant plant uncertainties. The robust performance problem is formulated in the so called Quantitative Feedback Theory (QFT) setting. A new method of solving the classical QFT problem that is based on the H technique is proposed. As a by product, a new robust stability criterion, which uses both sensitivity and complementary sensitivity functions alternatively, is developed. Another problem studied is the synthesis of a robust stabilizing compensator for plants with structured uncertainties. For this problem we develop a loopshaping technique similar to QFT loopshaping in principle but is based on the Youla parameterization. The advantage here is that the need to count the Nyquist encirclements for verifying stability of a nominal loop can be removed. Also considered is the robust stability problem related to MIMO. QFT nonsequential design method. A new sufficiency condition is provided to ensure closed loop robust stability of designs achieved with the nonsequential method.