Nonlinear model predictive control with online estimation of the specific growth rate in bioethanol fermentation

The paper addresses the problem of stabilization of a continuous fermentation process for efficient production of the bioethanol. The mathematical model of the process is derived on the assumption that the inhibitory effect of ethanol concentration is delayed in time and including decay of microorganism cells. Stability analysis of the openloop system demonstrates unstable behavior of the fermentation process in a range of dilution rates. The effect of kinetic parameters on the dynamical behavior of the process and steady-state ethanol concentrations is also studied. To stabilize the fermentation process a nonlinear model predictive control (NMPC) technique is applied. To deal with plant–model mismatch, an observer that estimates the microbial specific growth rate is designed. The effectiveness of the proposed NMPC-based strategy is evaluated under model uncertainty and measurement noise, and compared with two alternative NMPC-based strategies.