Iterative EKF as a controller in novel MPC formulation: First principles model based IEKF-MPC for SISO systems

Abstract

In the present study, novel MPC control strategy, using first principles models and complete state information is formulated using iterative EKF as predictive controller. This is based on extension of iterative EKF control estimation concept to MPC, instead of solution of an optimization problem approach followed in conventional MPC. The proposed approach incorporates all aspects of MPC including feedback correction term to handle process-model mismatch, move-suppression factor for control effort minimization, and constraints handling capability including terminal constraints. The performance of the proposed control strategy is evaluated through simulation case studies of SISO anaerobic digestion process and isothermal methyl methacrylate polymerization reactor with relative degree 1 and 2, respectively. The proposed strategy has shown superior performance over standard MPC in terms of faster response, set point tracking and disturbance rejection in both the case studies. Also, the proposed approach is found to be computationally more efficient than the standard MPC.