Homothetic Tube Model Predictive Control for Nonlinear Systems

Abstract

Homothetic tube model predictive control is generalized to the setting of nonlinear discrete time systems. The developed generalization is natural in the sense that the utilized assumptions in the nonlinear setting are suitably generalized variants of the related assumptions deployed in the linear case. The main analysis is carried out for Lipschitz systems, and it is subsequently extended to set–valued systems and other classes of nonlinear systems as well as particularized for parameterically uncertain linear systems. All desirable computational and structural properties of homothetic tube MPC are preserved.