On the reduction of redundancy in MPC parameter estimation

It is shown that for open-loop stable plants the task of an MPC can be solved by identifying the plant model only instead of the redundant prediction model. Interesting structural considerations can be derived from comparison of the topologies and structural forms of regulators obtained by pole-placement design, the GTDOF scheme and the MPC. Analyzing the possible closed-loop ID errors a new simple method is introduced for adaptive and combined iterative ID and control based on minimal number of parameters