Discrete model reduction preserving bounded realness (Q1361343)

In model reduction theory, it is required that the reduced-order system retains as many good properties of the original system as possible. There are some particular classes of systems, such as bounded real and positive real systems, for which it is not sufficient to minimize the \(H^\infty\)-norm error. In this paper, the authors propose a discrete model reduction method that preserves the bounded realness. They define the discrete bounded real balancing and obtain the reduced order discrete models from the generalized singular perturbation approximation and direct truncation of the bounded real balanced system. They show the properties of the reduced order systems and derive the \(H^\infty\) norm error bounds in some cases.