Control based on numerical methods and recursive Bayesian estimation in a continuous alcoholic fermentation process

• Published in: Bioresources Vol. 4; no. 4; pp. 1372 - 1395
• Main Authors: Quintero, Olga L., Amicarelli, Adriana A., Scaglia, Gustavo, di Sciascio, Fernando
• Format: Journal Article
• Language: English
• Published: 01-11-2009
• ISSN: 1930-2126; 1930-2126
• DOI: 10.15376/biores.4.4.1372-1395

Biotechnological processes represent a challenge in the control field, due to their high nonlinearity. In particular, continuous alcoholic fermentation from Zymomonas mobilis (Z.m) presents a significant challenge. This bioprocess has high ethanol performance, but it exhibits an oscillatory behavior in process variables due to the influence of inhibition dynamics (rate of ethanol concentration) over biomass, substrate, and product concentrations. In this work a new solution for control of biotechnological variables in the fermentation process is proposed, based on numerical methods and linear algebra. In addition, an improvement to a previously reported state estimator, based on particle filtering techniques, is used in the control loop. The feasibility estimator and its performance are demonstrated in the proposed control loop. This methodology makes it possible to develop a controller design through the use of dynamic analysis with a tested biomass estimator in Z.m and without the use of complex calculations.