Uncertainty, Robustness and Sensitivity Reduction in the Design of Single Input Control Systems

Abstract: Feedback control of systems modelled as single input, time invariant, linear and continuous time systems is the subject of this thesis. A growing interest in the design of feedback control systems that can cope with model uncertainties has been the motivation for the study. Some commonly used design approches for various kinds of systems are investigated with respect to robustness properties.The Smith controller, often propsed to controlling open loop stable plants with an input delay, state feedback control based on linear quadratic optimization techniques and output feedback control using a Kalman filter are considered.Also some new design strategies are given. A graphical design philosphy suitable for open loop stable plants with no modelled input delay, feedback stabilization of systems with a known number of unstable poles and modification of the Kalman filter based controller are proposed. The objective in common for these control strategies is to obtain control systems with good robustness properties.