Design and implementation of an adaptive predictive controller for a nonlinear dynamic industrial plant using Hysys and Matlab simulation packages

Predictive controller based on model has been known as a reliable and robust controller in the last 20 years. This paper presents a new idea of design and implementing an adaptive model predictive controller on an industrial “dynamic” and “nonlinear” plant in an integrated software environment using Hysys and Matlab packages. The model predictive controller formation is based on an adaptive state-space prediction model of the system response to obtain the control action by minimizing an objective function. The designed MPC controller is utilized to regulate a gaseous industrial plant, simulated in Hysys. The objective of controlling the plant is to compensate for the pressure variations in topside output of the vessel in on-line form. In this paper, the opening value percentage (OP) of a valve in the output is randomly excited in a given interval to identify the output pressure in the plant, called as Process Variable (PV). The predicted and desired outputs are then employed in the designed model predictive controller to determine the control actions in the prediction horizon. The simulation results obtained in the developed integrated Hysys-Matlab environment, demonstrate the capability of the proposed approach to efficiently monitor and control an industrial gaseous plant in a real and practical manner.