Assessing control strategies for the supercritical extraction from coffee beans: process-based control versus proportional integral derivative

• Published in: Journal of food process engineering Vol. 28; no. 5; pp. 494 - 505
• Main Authors: Riverol, C, Cooney, J
• Format: Magazine Article
• Language: English
• Published: Oxford, UK and Malden, USA Blackwell Science Inc 01.10.2005; Blackwell
• Subjects: Biological and medical sciences; carbon dioxide; coffee beans; Coffee, tea and other stimulative beverage industries; dynamic models; dynamic programming; Food engineering; Food industries; food processing quality; Fundamental and applied biological sciences. Psychology; General aspects; neural networks; supercritical fluid extraction; Extraction; Vegetables; Coffee; Process control; Bean
• ISSN: 0145-8876; 1745-4530
• DOI: 10.1111/j.1745-4530.2005.030.x

ABSTRACT This article addresses the formulation of a generic model control (GMC) in a supercritical fluid extraction unit (SFE). This alternative control is solved by formulating performance objectives subject to constraints on manipulated, controlled and associated variables. This study shows the performance of the control strategy depends on how the constraints are formulated. The batch extraction of caffeine is the problem proposed in this article. The presence of constraints in the optimization problem makes the overall system nonlinear even though the process dynamics are assumed linear. Hence, the “best” control strategy was determined using field data obtained directly from the pilot plant and compared with the decoupled proportional integral derivative control system actually installed in the plant.