Implementation of hybrid Control Strategy for Production of a Biotherapeutic in Pichia pastoris
12th September 2018, Wednesday, 04:00‐04:30 pm (25+5 mins), ChE Seminar hall
Recombinant protein production by various host organisms is strictly regulated, and this is determined by regulation of substrate and dissolved oxygen in the reactor. Since, the process dynamics unveils nonlinear interaction between these critical process parameters. PID controllers, while being extremely popular and quite effective for most applications, are suboptimal for control of such nonlinearities that exist in most biological systems. This article demonstrates implementation of a hybrid controller that effectively allows simultaneous control of critical process parameters: Substrate and Dissolved oxygen during fed batch fermentation for the production of a biotherapeutic in Pichia pastoris . While the former is achieved using a decoupled adaptive algorithm and decoupled input output control system mitigates the non-linearity in the DO control. The results demonstrate that the performance of DAC and DIOLC could satisfactorily regulate the substrate and DO, respectively to the desired set point in the present hybrid controller. However, when the individual control algorithms were used in isolation, effective control of the desired parameters were not achieved. The results shows that the deviation from the substrate set point is less than 0.029 g/L in the DAC compared to a deviation of about 0.09 g/L obtained using the PID controller. Though using DIOLC controller individually, deviation from set point was 0.031 g/L but due to frequent large oscillations the system never attained substrate set point, hence the process outcome was poor compared to DAC. In case of DO control, the deviation in DO from the set point was 2.58% using the DAC alone compared to 11.86% deviation obtained using the PID controller while in case of DIOLC the maximum deviation seen was 1.94%. Since substrate control was adequate in DAC, to further improve the performance for DO control, it was augmented with decoupled input-output linearizing controller (DIOLC), where adaptive part cater for substrate control and DIOLC algorithm works for DO control. The resulting hybrid DAC-DIOLC controller yielded more optimal performance than the discrete control schemes with deviation values for substrate loop 0.021g/L and for DO loop 1.12%.
Keywords: Pichia pastoris, Recombinant Human Serum Albumin (rHSA), Bioprocess Control, Adaptive Control.