Robust Nonlinear Model Predictive Control with Bounded Disturbances Based on Zonotopic Constraint Tightening

• Published in: Journal of control, automation & electrical systems Vol. 33; no. 1; pp. 78 - 91
• Main Authors: Cunha, Victor M., Santos, Tito L. M.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01-02-2022; Springer Nature B.V
• Subjects: Control; Control and Systems Theory; Control stability; Differential calculus; Disturbances; Electrical Engineering; Engineering; Mechatronics; Nonlinear control; Prediction models; Predictive control; Propagation; Robotics; Robotics and Automation; Robust control; Voltage converters (DC to DC); Zonotopes; Constraint tightening; Robust model predictive control; Set-based state estimation; Nonlinear control
• ISSN: 2195-3880; 2195-3899
• DOI: 10.1007/s40313-021-00783-0

This paper proposes a robust nonlinear model predictive control based on nominal predictions with tighter constraints derived from a zonotope-based disturbance propagation. The worst-case disturbance reachable sets are computed from zonotopes combined with the mean-value theorem applied to the nonlinear model, which reduces the conservativeness of the tighter constraints. Mean-value disturbance estimates are also incorporated into the prediction model in order to mitigate the effect of asymptotic constant disturbances while maintaining the recursive feasibility and stability of the control policy. The proposed disturbance propagation technique is applied to simulations of a DC–DC converter and a continually stirred tank reactor benchmark case studies to illustrate the benefits of the proposed approach.