Robustness Properties and Output Feedback of Optimization Based Sampled-data Open-loop Feedback

Rolf Findeisen, Frank Allgöwer

Abstract: This paper investigates the inherent robustness properties and the output-feedback problem of sampled-data open-loop feedback control. One typical sampled-data open-loop feedback scheme is nonlinear model predictive control. Specifically, the influence on the stability of the closed loop for model-plant mismatch, exogenous disturbances, unknown delays, numerical errors, and state estimation errors is examined. It is shown that if the decreasing function of the sampled-data open-loop feedback is continuous, then the closed-loop possesses inherent robustness properties. As one specific application of the derived result conditions for the semi-regional practical output-feedback stabilization via observer-based state-feedback are derived.