Multivariable control and online state estimation of an FCC unit

Alexandre Teplaira Boum (1) , Mohamed Abderrazak Latifi (2) , Jean-Pierre Corriou (2)

1 Université de Douala
2 LRGP - Laboratoire Réactions et Génie des Procédés

The purpose of this paper is to realize multivariable control , tuning and online state estimation of some parameters of the FCC unit . We implemented two control structures with the manipulated variables being the air inlet flow rate in the regenerator, the regenerated catalyst flow rate and the feed flow rate and, the controlled variable being the temperatures in the riser and in the densed bed of the regenerator. A novel four transfer function is built and used for controllability studies. Hard constraints are imposed with respect to the manipulated variables. Simulation results show that the configuration made of two inputs and two outputs is more easy to tune for control purposes. Althought there are important dynamic interactions between the components of the FCC and important nonlinearities, linear model predictive control is able to maintain a smooth multivariable control of the plant, while taking into account the different constraints. Tuning strategy is implemented to improve the tracking of the set point. Online state estimation is carried out with the use of the extended Kalman filter. The estimation gives results that can be used for monitoring purposes even in the presence of model mismatch

Mots clés: Extended kalm an filter; Multivariable; Identification; Model predictive control; Extended kalman filter

Domaines: Chimie

Alexandre Teplaira Boum, Mohamed Abderrazak Latifi, Jean-Pierre Corriou. Multivariable control and online state estimation of an FCC unit. Journal of Engineering Science and Technology Review, 2015, 8 (3), pp.158-168. ⟨hal-01303576⟩