Robust and reliable control in discrete time

This thesis presents new methods of robust control synthesis. Using the Frobenius-Hankel (FH) norm as an optimization criterion, a straightforward approach is obtained for computing controllers that have constraints on the order of the controller, for the centralized and decentralized feedback cases.A synthesis approach that relies on the use of discrete algebraic Riccati equations (DARE) is also used to obtain state-feedback, centralized output feedback, and decentralized output feedback controllers that guarantee stability and a predetermined level of disturbance attenuation as measured by the ${\cal H}\sb\infty$ norm.The design equations for the basic state-feedback, output-feedback and decentralized controllers may be modified to guarantee additional other properties of the closed-loop system such as robustness in the presence of plant (modelling) uncertainty, or reliability despite control-component outages.A convexity property of a certain matrix Riccati operator allows parameterization of families of control laws with the same desired properties. Each value of the parameter results in controller realizations of the same order as the plant.U of I OnlyETDs are only available to UIUC Users without author permissio