Cybernetic model predictive control of a continuous bioreactor with cell recycle

Area of Science:

• Biotechnology and Biochemical Engineering
• Process Control and Systems Biology

Background:

• Poly-beta-hydroxybutyrate (PHB) is a biopolymer with diverse applications.
• Efficient control of PHB synthesis in continuous bioreactors is crucial for industrial production.
• Cell recycle strategies can enhance productivity but complicate process control.

Purpose of the Study:

• To develop and simulate a cybernetic model for PHB synthesis in Alcaligenes eutrophus.
• To interface this model with a multirate model predictive control (MPC) algorithm.
• To evaluate the MPC's ability to control PHB productivity and concentration under various conditions.

Main Methods:

• Development of a cybernetic model for PHB synthesis.
• Identification of model parameters using experimental data.
• Implementation of a multirate nonlinear MPC algorithm for bioreactor control.
• Simulation of bioreactor operation with imposed disturbances and model uncertainties.

Main Results:

• The nonlinear MPC effectively controlled PHB productivity and concentration by manipulating dilution rate and recycle ratio.
• The controller demonstrated robust performance against unmeasured disturbances and changes in PHB productivity gain.
• The MPC successfully tracked maximum achievable productivity even with unreachable setpoints and significant process/model mismatch.
• Controller performance degraded with decreasing biological complexity in the employed models.

Conclusions:

• A cybernetic model coupled with nonlinear MPC provides effective control of PHB production in continuous bioreactors.
• The MPC strategy is robust to process disturbances and model uncertainties, crucial for industrial scalability.
• Model complexity impacts controller performance, highlighting the importance of accurate biological representation for optimal control.