A new hybrid modeling methodology based on delayed differential equations: Application to antibody expression by Pichia pastoris

• Type: Article in International Conference Proceedings Book
• Date: 2008

Authors

M. von Stosch (Author) FEUP; R. Oliveira (Author) FEUP; J. Peres (Author) FEUP; S. Feyo de Azevedo (Author) FEUP

Conference proceedings International

Title: Proceedings of the 10th International Chemical and Biological Engineering Conference

10th International Chemical and Biological Engineering Conference, CHEMPOR 2008

Braga, Portugal, 4 a 6 de Setembro

Scientific classification

FOS: Engineering and technology > Chemical engineering

CORDIS: Technological sciences > Engineering > Chemical engineering

Other information

Abstract (EN): In this paper a novel methodology for biosystems dynamic modelling is presented in which discrete time series, namely AutoRegressive (eXogenous) models are incorporated in the traditional parametric/nonparametric hybrid modelling framework. This results in a set of Delay Differential Equations (DDE) which describe the material balances of a bioreactor system in which dynamic kinetics are mimicked by a parametric/nonparametric submodel. The idea is to display better consistency with the nature of biological systems by associating the dynamics of a cellular metabolism to a parametric/nonparametric subsystem. The proposed hybrid structure is evaluated with fed-batch experimental data taken from a process for antibody expression by recombinant Pichia pastoris in addition to two simulation case studies. The first of these assumes a discrete time delay and the second assumes a distributed delay between kinetics. In this paper, it is shown that the proposed hybrid model is capable of modelling discrete and distributed delays between kinetics and outperforms the standard hybrid modelling methods with static kinetic models.

Language: English

Type (Professor's evaluation): Scientific