Study of model-PI predictive controller for coke fractionation tower using Matlab

Abstract

The control of coke fractionation tower has been an issue in both the Chemical engineering and process control field. It requires significant effort in practice to control the parameters in the coke fractionation tower. The performance of the conventional control design and tuning is found to be poor because of the non-linearity in the system. Real-time implementation showed that classical proportional-integral-derivative (PID) control using such models lacks enough robustness. In this study, a Model Predictive Control (MPC) associated with a PI control system is used to control the level process in a tower. The control loop contains both the MPC and PI control. This paper suggests the integration of conventional controller and MPC to improve the response of the system. The response of MPC-P and MPC-PI is compared. Result shows that MPC-PI produces good response than MPC-P.