Please cite this article in press as: L. Dewasme, et al., State estimation and predictive control of fed-batch cultures of hybridoma cells, J. Process Control (2015), http://dx.doi.org/10.1016/j.jprocont.2014.12.006 ARTICLE IN PRESS G Model JJPC-1877; No. of Pages 8 Journal of Process Control xxx (2015) xxx–xxx Contents lists available at ScienceDirect Journal of Process Control j ourna l ho me pa ge: www.elsevier.com/locate/jprocont State estimation and predictive control of fed-batch cultures of hybridoma cells L. Dewasme a,∗ , S. Fernandes a , Z. Amribt b , L.O. Santos c , Ph. Bogaerts b , A. Vande Wouwer a a Control Department, BioSys Center, Biosciences Institute, University of Mons, 31, Boulevard Dolez, 7000 Mons, Belgium b 3BIO-BioControl, Brussels School of Engineering, Université Libre de Bruxelles, AV. F.-D. Roosevelt 50 C.P. 165/61, 1050 Brussels, Belgium c CIEPQPF, Department of Chemical Engineering, Faculty of Sciences and Technology, University of Coimbra, Portugal a r t i c l e i n f o Article history: Received 8 May 2014 Received in revised form 1 December 2014 Accepted 18 December 2014 Available online xxx Keywords: Hybridoma cultures Nonlinear model predictive control Unscented Kalman filtering Biotechnology a b s t r a c t Fed-batch cultures of hybridoma cells are commonly used for the production of monoclonal antibodies (MAb). In this study, a simple macroscopic model of the cell culture is used, which is based on the overflow metabolism paradigm. This allows to specify optimal culture conditions, and the natural formulation of a nonlinear model predictive control strategy (NMPC). As not all the component concentrations are available for measurement, system observability is analyzed, and an unscented Kalman filter (UKF) is designed, which provides satisfactory estimates of glucose and glutamine concentrations. Robustness of the NMPC scheme is investigated, as well as the combined UKF+NMPC scheme, through a minimax robust version and the closed-loop system. © 2014 Elsevier Ltd. All rights reserved. 1. Introduction Hybridoma cells are important vectors for the production of monoclonal antibodies, and the pharmaceutical sector is paying more and more attention to Process Analytical Technologies (PAT) and process control for improving bioprocess yield or productiv- ity. Earlier optimization studies, such as in [1,2], were conducted on the basis of simple macroscopic mass balance models estab- lished from experimental data. In this study, we proceed with the same philosophy, but using a slightly more elaborate kinetic model suggested in [3], which takes metabolic changes and especially overflow metabolism into account. This metabolic phenomenon is induced when the rate of glycolysis exceeds the cell respiratory capacity, i.e., the capacity to oxidize substrates. Depending on the substrate concentration, the cell metabolism is divided in two path- ways: the respirative mode below a certain critical substrate level, and the respiro-fermentative mode when substrate is in excess, leading to the formation of inhibitory by-products (lactate and ammonia). ∗ Corresponding author. Tel.: +32 65374135; fax: +32 65374136. E-mail addresses: Laurent.Dewasme@umons.ac.be (L. Dewasme), Sofia.Fernandes@umons.ac.be (S. Fernandes), Zakaria.Amribt@ulb.ac.be (Z. Amribt), lino@eq.uc.pt (L.O. Santos), Philippe.Bogaerts@ulb.ac.be (Ph. Bogaerts), Alain.VandeWouwer@umons.ac.be (A. Vande Wouwer). To drive the bioprocess close to an optimum, while avoiding this undesirable effect, a closed-loop optimizing strategy proposed in [4] is used. Various forms of optimizing control of bioprocesses have been proposed in the literature, including adaptive, probing, robust or predictive control as discussed for instance in [5–8], respectively. In this study, we give preference to nonlinear model predictive control, which offers a natural and widely accepted framework for the formulation of an optimal control under constraints. There is a vast and rich literature with overviews on NMPC develop- ments, research, and applications (e.g., [9,10]). Some of these works address the problem of robust NMPC of fed-batch processes (e.g. [11,12]). An overview of recent developments can be found in [13] and references therein. A simple and efficient approach to maintain hybridoma cultures in the optimal operating conditions is to regulate the glucose and glutamine concentrations at the critical levels [14]. However, reli- able glucose and glutamine probes are currently rare and/or very expensive on the market and an interesting alternative is to design software sensors, which are at the same time cheap and reliable, and can be used for online measurement, as in [15–19]. In this study, an unscented Kalman filter is designed for online estimation of glu- cose and glutamine in hybridoma cell fed-batch cultures based on the considered available measurements, e.g., biomass, lactate and ammonia, and the robustness of the scheme is investigated. This paper is organized as follows. In Section 2, the mathemat- ical model of HB-58 [3] is briefly described, whereas in Section 3 optimal operating conditions are devised. A NMPC algorithm is http://dx.doi.org/10.1016/j.jprocont.2014.12.006 0959-1524/© 2014 Elsevier Ltd. All rights reserved.