A ROBUST DESIGN OF MODEL-BASED PREDICTIVE CONTROL SYSTEM

摘要

This paper presents a robust enhancement scheme for model-based predictive controllers applicable to both state-space and input-output representation of the process model. The novel scheme takes advantage of the fact that the dynamic characteristic of the system located in different parts of the stability disc inherits different dynamic properties. Alternatively, the robustness of some systems is more superior than others with their characteristic polynomial situated in another position. Based on the asymptotic properties of the class of optimal control algorithms, unstable dynamics may be reflected across the stability disc and thus, we are able to direct the closed loop system dynamics to the more robust region by introducing a fictitious disc targeting at the latter region. Simulation study shows that the proposed robust control system is effective and robustness can be enhanced.