Robust model predictive control for constrained networked nonlinear systems: An approximation-based approach.

• A robust approximation-based model predictive control scheme for constrained networked nonlinear systems is proposed. • A novel constraint tightening approach for the formulation of the FHOCP is proposed. • Both model error and inter-sampling behavior are considered. • The state and control input constraints are satisfied in continuous-time sense. • A guideline to determine the allowable sampling period is provided. In this paper, a robust approximation-based model predictive control (RAMPC) scheme for the constrained networked control systems (NCSs) subject to external disturbances is proposed. At each sampling instant, the approximate discrete-time model (DTM) is utilized for solving the optimal control problem online, and the control input applied to continuous-time systems can then be determined. Such RAMPC scheme enables to implement MPC for the continuous-time systems in the digital environment, and meanwhile, achieves the state and control input constraints satisfaction in continuous-time sense. Furthermore, we also provide a guideline to determine the allowable sampling period. Sufficient conditions for the feasibility of the RAMPC scheme as well as the associated stability are developed. Finally, the effectiveness of the RAMPC scheme is shown through a numerical simulation. [ABSTRACT FROM AUTHOR]