Sporala red del conocimiento
EN
Universidad Nacional de Colombia

Manizales - Ingeniería y Arquitectura - Maestría en Ingeniería - Ingeniería Química · 2021

Modelamiento estocástico de biorreactores

Modeling and simulation have become over time a key tool for the continuous improvement of designs, control systems, and understanding of industrial processes, including biotechnology. Processes that respond to the need to make use of alternative sources to those based on petroleum, for the production of consumer products of all kinds. This is why this work focuses on modeling biotechnological processes such as enzymatic transformation and fermentation. The modeling was investigated using stochastic and deterministic approaches for two case studies, enzymatic reactions with and without inhibition, as well as the xylitol bioproduction from glucose and xylose. The deterministic approach was based on the solution of the ordinary differential equations obtained from the mass balances for a batch reactor and the chemical kinetics of the mass action law and the Michaelis-Menten type. On the other hand, the stochastic approximation was based on the application of the Gillespie stochastic solution algorithm, from the stochastic chemical kinetics and the stochastic approximation of the Michaelis-Menten equations. Thus, the objective of this thesis is to investigate the application of stochastic models and evaluate the uncertainty of these systems, as well as to re-estimate parameters for the xylitol production model. It was found in this work that stochastic methods have the same predictive power as deterministic ones for the most common cases of inhibition in the literature (Competitive, Non-competitive, Un-competitive) and the xylitol production, but with the advantage of being able to analyze the inherent uncertainty of biological systems (enzymatic reactions and fermentation). Besides, It was possible to evaluate the evolution of the uncertainty of the stochastic model throughout the simulation and its relationship with the relative uncertainty of the experimental measurements. The sensitivity analysis allowed making a conscious selection of the size of the system and the number of realizations necessary to obtain a practically constant uncertainty value. Moreover, these methods served to propose a translation of the xylitol bioproduction model into stochastic terms, despite the lack of information on the reactions that lead to the formation of biomass and the inclusion of the mass transport model. Besides, the deterministic re-adjustment of the parameters showed a better fit of the model to the experimental data of xylitol (compound of interest). The greatest contribution, in this case, was to illustrate how this method (ABC rejection sampler) offers greater robustness when evaluating the uncertainty of the model and being able to optimize under uncertainty. Therefore, this thesis includes contributions to the state of the art in Process System Engineering that will provide new perspectives for the modeling and analysis of systems under uncertainty (Texto tomado de la fuente)

Texto completo 91 páginas con texto

Leer la tesis completa Ficha en el repositorio

Tesauro de su biblioteca