A Norm-Bounded based MPC strategy for uncertain systems under partial state availability

TL;DR

This paper proposes a robust MPC approach for uncertain discrete-time linear systems with partial state measurements, combining offline memoryless control design with online predictive control using semi-definite programming.

Contribution

It introduces a novel norm-bounded uncertainty handling method within MPC that accounts for partial state feedback and input rate constraints.

Findings

Effective control under partial state availability.

Use of semi-definite programming for control law synthesis.

Robustness against norm-bounded uncertainties.

Abstract

A robust model predictive control scheme for a class of constrained norm-bounded uncertain discrete-time linear systems is developed under the hypothesis that only partial state measurements are available for feedback. Off-line calculations are devoted to determining an admissible, though not optimal, linear memoryless controller capable to formally address the input rate constraint; then, during the on-line operations, predictive capabilities complement the off-line controller by means of N steps free control actions in a receding horizon fashion. These additive control actions are obtained by solving semi-definite programming problems subject to linear matrix inequalities constraints.