Adaptive model predictive control for a class of constrained linear systems with parametric uncertainties

This paper investigates adaptive model predictive control (MPC) for a class of constrained linear systems with unknown model parameters. This is also posed as the dual control problem consisting of system identification and regulation. We first propose an online strategy for simultaneous unknown parameter identification and uncertainty set estimation based on the recursive least square technique. The designed strategy provides a contractive sequence of uncertain parameter sets, and the convergence of parameter estimates is achieved under certain conditions. Second, by integrating tube MPC with proposed estimation routine, the developed adaptive MPC provides a less conservative solution to handle multiplicative uncertainties. This is made possible by constructing the polytopic tube based on the consistently updated nominal system and uncertain parameter set. In addition, the proposed method is extended with reduced computational complexity by sacrificing some degrees of optimality. We theoretically show that both designed adaptive MPC algorithms are recursively feasible, and the perturbed closed-loop system is asymptotically stable under standard assumptions. Finally, numerical simulations and comparison are given to illustrate the proposed method.