Iterative identification and model predictive control: Theory and application

ICIC International ©2021 ISSN.Although system identification and model predictive control are two differ-ent research fields separately, one gap exists between these two subjects. To alleviate this gap between them, system identification and model predictive control are combined to be one iterative identification and model predictive control strategy in this paper, where the process of system identification and model predictive control will be carried out iteratively many times until convergence. Based on some priori information about the asymptotic analysis for predictor error identification, variance analysis corresponding to closed loop output response is derived to show the tracking performance for this proposed iterative strategy. From this derived variance analysis, some factors can be chosen to guarantee the perfect tracking performance, such as input spectrum, and noise filter. Furthermore, to extend this iterative strategy to more general cases in industry, one model predictive control based reference governor is studied to provide proportional-integral-differential (PID) controller. Finally, several simulation experiments about flight control for heli-copter have been performed to demonstrate the effectiveness of our proposed theories.

Iterative identification and model predictive control: Theory and application Academic Article in Scopus