Simultaneous design and nonlinear model predictive control under uncertainty: A back-off approach. (February 2022)

• Record Type: Journal Article
• Authors: Palma-Flores, Oscar; Ricardez-Sandoval, Luis A.
• Is Part Of: Journal of process control. Volume 110(2022)
• Page Start: 45
• Page End: 58
• Publication Date: 2022-02
• Subjects: Back-off method -- Integration of design and control -- Power series expansion -- Nonlinear model predictive control; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2021.12.007
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract: We propose the use of an iterative back-off approach for the integration of design and NMPC-based control under uncertainty. This methodology is based on the use of power series expansions (PSE) to feature the simplification of the resulting bilevel optimization formulation when nonlinear model predictive control (NMPC) is incorporated as the control scheme. To illustrate the benefits of the proposed approach, the optimal design and control of a wastewater treatment plant is formulated under different scenarios. The results show that the proposed methodology leads to improvements in control performance compared with the use of decentralized PI-based controllers and a linear MPC strategy to maintain the dynamic operation of the system in the presence of disturbances and parameter uncertainty. Thus, the present approach not only reduces the computational effort by solving single-level NMPC-based optimization problems but it also returns economically attractive and dynamically operable designs while using NMPC. Highlights: A back-off methodology for optimal design and NMPC-based control is presented. PSE-functions are used in the methodology to represent closed-loop behaviour. A Wastewater Treatment Plant is used to test the proposed methodology. Performance is improved when NMPC is considered for optimal process design.