Control Separation Based Model Predictive Control for Rejection of Unmatched Input Disturbances

A control separation based LQ control design has the capability for achieving both optimal tracking and disturbance rejection. In this paper, a control separation based model predictive control (MPC) scheme is developed for rejection of unmatched input disturbances of possibly nonminimum-phase time-varying systems. A control separation based MPC framework is formulated for disturbance rejection, which has the enhanced capability for handling system parameter variations by a closed-loop system operation. Such a framework is built in the development of three components: a baseline model predictive control design for linear timevarying systems, a control separation based LQ control design for unmatched input disturbances, and a control separation based MPC disturbance rejection design for linear time-varying systems. The analysis of system stability and output regulation is presented, and a comparison to traditional LQ based designs is given. Simulation study is conducted for the turbulence compensation problem of the hypersonic vehicles, and its results verify the effectiveness of the new control separation based MPC design for disturbance rejection.