A new multivariable control concept for the falling film evaporator process. (October 2021)

• Record Type: Journal Article
• Authors: Hofmann, Julian; Ponomarev, Anton; Hagenmeyer, Veit; Gröll, Lutz
• Is Part Of: Journal of process control. Volume 106(2021)
• Page Start: 72
• Page End: 83
• Publication Date: 2021-10
• Subjects: Falling film evaporator -- Multivariable model-based control -- Time delay -- Digital twin; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2021.08.015
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract: The paper presents a new multivariable control concept for falling film evaporators (FFEs). Our concept solves the major challenges encountered in modern FFE control: large transport delays, additional control of the output mass flow, coupling of controlled variables, and disturbances due to time-varying input dry matter content. The challenges are addressed together, for the first time, by the following control design. Based on a dynamic nonlinear input–output model, we consider a linearizing output transformation to enable application of classical linear control methods composed of feedforward design, disturbance rejection, and a decoupling network. Due to these features, we are able to design robust PID and PI controllers that substantially compensate plant-model mismatches. Connecting our concept to a digital twin of the plant yields good performance, which encourages future application of the design in the real-world process. Highlights: The main control challenges w.r.t. falling film evaporators (FFEs) are addressed. Robustness w.r.t. parametric plant-model mismatch is shown in a simulation study. Robustness w.r.t. structural plant-model mismatch is shown via a digital twin. An automated ramp-up strategy for the FFE process is proposed and validated. Simple implementation of the concept to standard digital control systems is enabled.