Filtered Smith Predictor to control pH during enzymatic hydrolysis of microalgae to produce l-aminoacids concentrates

The main step of the production of l-aminoacids concentrates from microalgae is the enzymatic hydrolysis. This process has to be done at the optimum temperature and pH of the enzyme in order to obtain the maximum yield. In order to keep a constant pH, NaOH (1 M) or H2SO4 (1 M) are used, depending on the reaction step. When working with pH it is not possible to use the governing equations due to the complexity of the enzymatic hydrolysis. In this work, the modelling of pH is experimentally performed on a laboratory scale plant. Since the obtained model presents delay and uncertainty in the parameters, a Filtered Smith Predictor is proposed as control strategy. This control scheme has been tested in the real system. The Filtered Smith Predictor has reduced the Integral Absolute Error and the time for the solution addition in more than 25% and it has increased 5% the production of l-aminoacids compared to an on-off control, which is the controller most used in these processes.

► An experimental pH model with delay and uncertainty in the parameters is obtained.
► A Filtered Smith Predictor (FSP) for a pH robust control is designed.
► Control system is verified in simulation and implemented in the real system.
► Real results show a reduction of a 25% of the IAE and the injection time.
► Production of aminoacids increases nearly 5% with respect to on–off control.