The main idea behind model predictive control is to solve an optimization problem online. On one hand, in reality, model/plant mismatches, exogenous disturbances, numerical errors and state measurement errors are present. On the other hand, the MPC control law provides a feedback control only at specific sampling instant and the system is controlled open-loop during adjacent sampling instants. Therefore, robust analysis and synthesis of MPC are of significant theoretical and practical importance.
In this talk, we are interested in the analysis of the inherent robustness properties of nominal MPC of nonlinear systems with input constraints, where the disturbances are persistent but bounded and the optimization problem has a terminal constraint. It is worth noting that the analysis does not assume the continuity of the optimal cost functional or of the control law, and hence the results are both more general and of greater practicality than previous ones. It is shown that the degree of robustness depends on the terminal set and the terminal penalty function, the prediction horizon, the upper bound on the disturbances and the logarithmic norm of the considered system.
Shuyou Yu is a Post-doctoral Research Associate at the Institute for Systems Theory and Automatic Control, University of Stuttgart. He received his Ph.D. from University of Stuttgart in 2011, and his B.Eng. and M.Eng from Jilin University, China, in 1997 and 2005, respectively.