Model Predictive Control for steam distillation essential oil extraction process

Abstract

This paper presents the simulation and assessment of a Model Predictive Control (MPC) for the steam distillation essential oil extraction process. The mathematical model of the plant was obtained beforehand from the system identification technique by using ARX model structure. The model is then applied to the MPC control structure as a predicted model. The objective of the controller design is to keep low overshoot in steam temperature in order to improve the dynamic response of the essential oil extraction process. The conventional feedback controller is good in terms of accuracy and to attain the stability of the controlled variable but it is rather complicated when dealing with the nonlinear characteristic and slow process dynamic. This is an important criterion when we opt for MPC since steam temperature was known to exhibit a very large time delay. The simulation results on both PID and MPC show that the MPC can sustain the steam temperature at required set point efficiently with acceptable value of overshoot and RMSE.