Simply Robust Control Strategy Based on Model Predictive Control

Abstract

A control engineer is required to design a controller that will stabilize a system, and satisfy specific performance under the influence of disturbance, observation noise, unmodeled dynamics. Controllers with strong robustness such as an H-infinity control and a Sliding Mode Control (SMC) theory have been proposed, but these controllers are difficult to apply to industries due to the difficulty of theory and problems such as chattering. On the other hand, a Model Predictive Control (MPC) provides a controller with intuitive and easy design and tuning. The MPC is one of the excellent controllers not only having this advantage but also considering constraints of inputs or state variables. However, the controller based on the MPC has insufficient robustness. We propose a new control theory that is the modified MPC to improve robustness by applying matching conditions. This proposed method not only has strong robustness but also Inherits the characteristics of the MPC with easy design and tuning. The new controller proposed in this study is named an Extended Model Predictive Control (EMPC). The theory of the EMPC is explained in this paper and verified effectiveness by numerical simulation.