A multiplier approach for the robust design of discrete-time control systems with real/complex uncertainties (Q1295036)

The authors present an algorithm for robustness analysis and design for discrete-time systems in the presence of mixed uncertainties based on a multiplier approach for computing suitable scaling of transfer matrices. Suitable dynamic multipliers could be found using LMI procedures and the conditions testing robustness specifications are formulated in terms of a convex minimization problem. The \(H_\infty\)-norm condition involves the real plant data and scalings computed via standard spectral factorizations of the multiplier, which enables to bypass the repeated sector transformations. The authors claim that their contribution is in developing the multiplier algorithm for robustness of discrete-time systems working exclusively in discrete time. They seem to forget that a real plant is always continuous in time and it is the weakest feature of their reasoning.