Predictive modeling and loose-loop control for perfusion bioreactors

Perfusion bioreactors are widely used to produce recombinant proteins and m onoclonal antibodies for therapeutic and diagnostic use. Better control of the cellular environment can lead to higher volumetric productivity, ensure product consistency and optimize medium utilization. The objective was to manipulate and control substrate concentrations in the perfusion bioprocess using predictive modeling and control. The goal of the predictive controll er was to minimize future deviations from the set point concentration, by s tructuring the controller output. The appropriate structure for the future manipulated variable was specified using the selected model of glucose upta ke rates (GUR). When there was a deviation from the set point value, the fl ow rates were adjusted to drive the process close to the set point value in a defined first order manner. The shape of the first order process respons e depended on the magnitude of the deviation from the set point value. With daily sampling and glucose measurement, a feed rate profile (eight flow ra tes per day) was specified to control the bioprocess. Despite the infrequen t sampling, the predictive control protocols demonstrated glucose variation of less than 0.4 mM in transient conditions, and less than 0.2 mM in pseud o-steady-state conditions. The non-linear controller allowed for rapid chan ges in set point concentrations (6-9 h) or a reference trajectory to be fol lowed. Set point changes and reference trajectories were simulated and test ed with real process data. Modeling error and measurement bias were simulat ed to have the greatest potential effect during exponential growth. With go od model estimation of the GUR, predictive control was able to maintain the process at the set point with a level of variability approaching that of t he glucose assay. (C) 2001 Elsevier Science B.V All rights reserved.