Model matching approach to discrete-time polynomial optimization problems

This paper presents a solution to the H_∞ predictive control problem based on minimax prediction using a technique inspired by the solution of the H_∞ model matching problem. This solution methodology uses a Youla-Kucera parameter to consider only controllers which are in the set of stabilizing controllers. The final form of true equations which define the controller is simplified, resulting in a better numerical solution readily solvable by a generalized eigenvalue problem