Model predictive control

Model predictive control (MPC) has become the accepted standard for complex constrained multivariable control problems in the process industries. Over the last decades, a solid theoretical foundation for MPC has emerged so that in real life, large-scale MIMO applications controllers with non-conservative stability guarantees can be designed routinely and with ease. The big drawback of the MPC is the relatively formidable on-line computational effort, which limits its applicability to relatively slow and/or small problems.

Our research efforts aim at minimizing the on-line computational effort by

• developing a PWL control law based on a suboptimal solution to the MPC problem;
• implementing the control action on digital circuits

Involved People:

• Marco Storace
• Tomaso Poggi
• Alberto Oliveri

Main Publications:

• Bemporad A, Oliveri A, Poggi T and Storace M (2011), "Ultra-Fast Stabilizing Model Predictive Control via Canonical Piecewise Affine Approximations", Automatic Control, IEEE Transactions on. Vol. 56(12), pp. 2883 -2897.

  [BibTeX] [DOI]

Funding:

• This research activity was partially supported by the European Community through the MOBY-DIC project (FP7-IST-248858)