Stabilizing Model Predictive Control for Nonlinear Systems in Input-Output quasi-LPV Form

Abstract

This paper present a Nonlinear Model Predictive Control (NMPC) scheme for systems in input-output quasi-Linear Parameter Varying (IO-qLPV) form. Stability is guaranteed by the use of terminal ingredients: a terminal cost associated with a control Lyapunov function, with a dual mode controller within a terminal constraint set. These terminal ingredients are computed offline by solving an LMI problem, using a recent result from IO-LPV literature. Online computational complexity is kept low by solving the nonlinear optimization problem as a sequence of Quadratic Programs. A simulation example on a 2-DOF robotic manipulator illustrates the approach and the computation of the terminal ingredients.