Model-Based Control for Fed-Batch Enzymatic Hydrolysis Reactor of Lignocellulosic Biomass

Abstract

In this study, model predictive control (MPC) based on a semi-mechanistic model with sugar products as a competitive inhibitor was developed for the fed-batch enzymatic hydrolysis of a dilute acid pretreated of lignocellulosic biomass. A classical proportional integral derivative (PID) controller cannot guarantee a satisfactory response. Therefore, to improve the controller response of more complicated systems such as nonlinear processes that need a more efficient control system, the model-based control strategies are needed. Generic model control (GMC) which is a simple structure as PID controller and MPC controller which is applying the advanced optimization algorithm to minimize the error of input variable deviations and predicted output variables are implemented to control the glucose concentration by the manipulation of the dilution rate. The performance of MPC controller is compared to PID and GMC controller. For the results of controller performance, MPC is the most superior controller assessed by settling time and performance criterions in terms of the Integral of Absolute value of Error (IAE) and Integral of Square of Error (ISE).