A Practical Tuning Approach for Multivariable Model Predictive Control

This paper presents a practical off-line approach for tuning Model Predictive Control (MPC) parameters with constraints. The MPC parameters values are computed from the expected closed-loop performances. Our original approach based on the exponential data weighting can be applied to a wide set of controllable Multi-Input Multi-Output (MIMO) processes and plants. It guarantees the stability of closed-loop systems and permits the controller's designer to set a trade-off between accuracy and rapidity. Some simulation results shown emphasize its effectiveness when compared to several existing methods.