Multi-Objective Model Predictive Control

Abstract

Multi-objective optimization design recently has attracted great attention of the researchers in solving engineering problems that have conflicting objectives.
Although several control specifications which are often irreconcilable can be considered in the single objective function, choosing the appropriate weighting functions are another challenge faced by control designers. In this paper, a new Model Predictive Control scheme based on the multi-objective optimization is proposed in which at each sampling time, the MPC control action is chosen automatically among the set of Pareto optimal solutions based on the Nash Bargaining Solution from Game Theory. This method is independent of the system type. It is applied on the nonlinear systems along with TP transformation to design multi-objective MPC. As a result, LMIs and convex optimization techniques can be utilized to provide an on-line solution for the multi-objective MPC design. The proposed method is executed on a complex nonlinear system.  It is shown through the examples that the proposed method can execute approvingly compared to other methods in the literature of the control systems.