Optimal constant scaling matrix for H/sup infinity / control and mu -synthesis-state feedback case

A property of a constant scaling matrix for mu -synthesis is clarified for a state feedback case. The property of a set of constant scaling matrices for which the H/sup infinity / norm can be made less than gamma by state feedback is examined. The convexity of the set and the uniqueness of the optimal value are shown. An algorithm for determining the optimal value is shown for the case of two uncertainties, and a novel DK-iteration is proposed based on the algorithm.<
>