Adaptive receding horizon control for constrained nonlinear systems

Considers the problem of adaptive receding horizon (model predictive) control of nonlinear systems which are subject to control constraints and are linear in the unknown parameters. The ability of model predictive control to handle constraints, especially actuator constraints, is very important for applications. However, much of the literature on model predictive control does not adequately address the stability problem. In an attempt to solve the constrained, adaptive receding horizon problem, the authors restrict themselves to systems with accessible states. It is shown that a standard estimation procedure provides accurate prediction over a finite horizon even if the estimated parameter is not equal to the true parameter. The estimation procedure is then employed in a receding horizon controller. Global stability of the resultant closed-loop system is established