ON LINEAR-PROGRAMMING AND ROBUST MODEL-PREDICTIVE CONTROL USING IMPULSE-RESPONSES

Campo and Morari have derived a linear programming problem, with a potentially large number of constraints, which is equivalent to a min-max formulation for robust model-predictive control of linear systems. That formulation involves minimization, with respect to the controls, of the maximum, with respect to the system's impulse response (from a set of possible impulse responses), of the infinity norm of the error between the predicted and required system output sequences. Here an alternative linear programming problem is derived which has a smaller number of constraints and is therefore potentially more convenient for on-line control.