LMI solution to gain-constrained robust deadbeat pole assignment

A novel optimization approach to robust deadbeat control is proposed. The design problem is cast into a convex programming task in which a special measure of closed-loop eigenvalue sensitivity is minimized. Advantages of the proposed method include: (1) global optimality is guaranteed when the solution set is nonempty; (2) constraints on the feedback gain can be catered naturally; (3) minimum-gain deadbeat control design can be readily treated.published_or_final_versio