Optimal projection methods for model order reduction of discrete-time systems

Abstract

AbstractNumerical algorithms are developed for model order reduction of discrete-time systems using both optimal projection and $H_2$-norm minimization. The state-space matrices of the reduced-order system are obtained via the solution of a convex optimization problem. Subsequently, the results are exploited for the design of non-linear reduced-order systems verifying the input-to-state stability property. Proofs of stability and error approximation bounds are provided along with numerical simulations to highlight the strengths and the validity of the theoretical results.