Modelling and L1 adaptive control of pH in bioethanol enzymatic process

The enzymatic process is a key step in second generation bioethanol production. Pretreated biomass fibers are liquefied with the help of enzymes to facilitate fermentation. Enzymes are very sensitive to pH and temperature and the main control challenge in the nonlinear process is to ensure minimum deviations from the optimal pH level. This article develops a mathematical model for the pH, which has not been reported earlier for this particular process. The new model embeds flow dynamics and pH calculations and serves both for simulation and control design. Two control strategies are then formulated for pH level regulation: one is a classical PI controller; the other an L₁ adaptive output feedback controller. Model-based feed-forward terms are added to the controllers to enhance their performances. A new tuning method of the L₁ adaptive controller is also proposed. Further, a new performance function is formulated and tailored to this type of processes and is used to monitor the performances of the process in closed loop. The L₁ design is found to outperform the PI controller in all tests.