A New Tuning Approach for MPC Applied to a Disturbed DC Motor

Abstract

In order to get correct results using the Model Predictive Control (MPC), one must find the suitable values of its parameters. Despite a large number of papers in literature on MPC tuning methods, there is no available analytical approach permitting to identify explicitly the robustness area of a process independently of its order. Here we intend to overcome this limit thanks to an analytical one. The interest of our approach is to be applicable to a wide set of linear controllable and observable single-input single-output (SISO) processes. The issues of optimal closed-loop stability and the energy consumed are addressed in this paper. The proposed method is tested experimentally via a DC motor. Finally, a performance comparison is made with existing methods to emphasize its effectiveness.