Application of conventional nonlinear model predictive control (NMPC) and economic nonlinear model predictive control (E-NMPC) for technical and economical optimization of biochemical reactor system

Abstract

This paper studies the application of Economic Nonlinear Model Predictive Control (ENMPC) and conventional NMPC to a biochemical reactor system with variation in reaction kinetics and disturbance values. The controlled variable for this study is the biomass concentration of the outlet stream leaving the reactor. To control it, conventional NMPC scheme which minimizes the controlled variable deviation to a desired set point and ENMPC scheme which optimizes the biomass production of the system are simulated against disturbance in reactor feed substrate concentration. Result shows that the conventional NMPC schemes are able to bring or maintain the controlled variable to a desired set point. However, the ENMPC scheme outperform the conventional NMPC in cumulative biomass production along the simulation period of up to 57% at the cost of higher computational time.