The Improved Robust Model Predictive Control with Mixed H2/H∞ Control Approach

Abstract

Considering the linear time invariant (LTI) system with additive disturbance, a synthesis method of robust model predictive control (RMPC) is proposed, which is based on the mixed H2/H∞ approach so as to balance the control performance and the capability of disturbance rejection. In order to reduce the conservativeness, a multi-step control strategy is introduced to improve the control performance and enlarge the feasible region of MPC. Meanwhile, an efficient version of RMPC is also developed to reduce the online computational burden, which makes the design more practical. Numerical examples illustrate the effectiveness of the proposed methods.