Lexicographic NMPC Algorithms with Changing Criteria Prioritization

This paper considers the multiobjective control problem with different criteria priorities and we propose a lexicographic multiobjective model predictive control (MO-MPC) scheme for a changing criteria prioritization of constrained nonlinear systems. Using the lexicographic approach, the solution to the MO-MPC problem is obtained by solving a sequence of single objective, constrained optimal control problem over a receding horizon. The proposed controller is always guaranteed feasibility for any changing criteria prioritization by the general terminal constraint. Asymptotic stability of the closed-loop systems is also proved by a Lyapunov function of the positive definite function with the highest priority. An example of a two-tank system control is used to illustrate the effectiveness of the scheme proposed here.