A real-time approach for moving horizon estimation based nonlinear model predictive control of a fluid catalytic cracking unit

Output feedback non-linear model predictive control (NMPC) approach is illustrated on a simulated FCC unit (FCCU). This approach considers the most important features of a real-time control algorithm, which are often overlooked in simulation studies, contouring thus a framework for practical NMPC implementation. The most important features considered in the approach are: state and parameter estimation, efficient solution of the optimization, and computational delay. In the output feedback NMPC approach used, only measurements that are available in practice are considered, whereas the rest of the states are estimated together with uncertain model parameters using a moving horizon estimation (MHE) technique. The approach developed, is based on a state-of-the-art, large-scale SQP-type nonlinear optimization solver, which uses a sparse, interior-point multiple shooting algorithm, that exploits the special structure of optimization problem that arise in NMPC or MHE. The solution of the optimization problem from the NMPC and MHE requires a usually not negligible amount of time, when the system evolves to a different state where the solution is no longer optimal. The importance of taking this computational delay into account is also assessed and a real-time formulation of the control approach is described that includes the computational delay in the NMPC approach. The advantages of the proposed real-time approach are presented through the simulated industrial FCCU application. This is an abstract of a paper presented at the 7th World Congress of Chemical Engineering (Glasgow, Scotland 7/10-14/2005).