(202b) Study of Infinite Horizon MPC Implementation With Non-Minimal State Space Feedback in a Propylene Production Unit Using Dynamic Process Simulation

One of the key challenges in the process industry is how to best control and stabilize the plant while looking for the most profitable operating point. Therefore, advanced control (MPC), which is a control standard in the oil refining industry, is frequently implemented as one of the layers of a control structure where a Real Time Optimization (RTO) algorithm – laying in an upper layer of this structure – defines optimal targets for some of the inputs and/or outputs.

This work applies such optimization structure in a Petrobras heat integrated Propylene/Propane splitter that has a vapor recompression column (VRC), which has become the standard heat pump technology in distillation and substantial energy savings. The approach proposed here uses the rigorous dynamic simulation software DYNSIM^® as a virtual plant which is able to communicate with the advanced control algorithm represented by the Infinite Horizon Model Predictive Control (IHMPC) developed in MATLAB^® and the rigorous steady-state optimizer ROMeo^® through an Open Platform Communication (OPC) interface. The advanced controller is based on a state-space model in the incremental form, where the states are the past system outputs and input increments (realignment model) such that no state observer is required, and considers the existence of zone control, optimizing targets and time delays. The simulation results showed that the proposed integration provides convergence and nominal stability to the closed-loop system, and can be implemented in the real system.