Abstract
The study presents the development and experimental verification of an automated control system for the two-stage hydrolysis of wheat straw, implemented through the integration of a dynamic kinetic model into a model predictive control framework. The control object is characterized by nonlinear dynamics, the presence of technological constraints, and sensitivity to variations in feedstock composition, including cellulose, hemicellulose, and lignin content, which justifies the application of predictive control methods. The structure of the control system is based on a hierarchical architecture: local PID controllers ensure stabilization of temperature and pH conditions, while model predictive control optimizes control actions according to the criterion of maximizing sugar yield under constraints on temperature, pressure, and processing time. The mathematical model includes the kinetics of acid hydrolysis (hemicellulose-xylosefurfural) and the enzymatic stage based on a modified Michaelis-Menten equation accounting for enzyme inhibition and deactivation. Control is implemented using a discrete state-space model with a time step of 1 minute, taking into account transient processes. The system is implemented on a programmable logic controller platform with SCADA integration for data acquisition and visualization. Experimental validation demonstrated an increase in sugar yield up to 76-85%, a reduction in the coefficient of variation to 4%, a decrease in inhibitor concentration, reduced stabilization time and energy consumption, as well as robustness to disturbances and variations in feedstock composition.