Nonlinearand Constrained State Estimation Based on the Cubature Kalman Filter.

Thispaper investigates the use of several nonlinear estimation algorithmssuch as extended Kalman filter (EKF), unscented Kalman filter (UKF),and cubature Kalman filter (CKF) in the problem of state estimationin chemical processes. Three simulation case studies are consideredto evaluate the performance of the proposed method. The second casestudy uses the experimental data to investigate the accuracy of theCKF against the UKF in practical applications. Simulation resultsconfirm the superiority of the CKF to the EKF and UKF. However, allof these approaches fail to handle the constraint issue in state estimationproblems. Subsequently, a modified CKF is introduced to overcome thelinear constraint in nonlinear estimation problems. The final partof the paper shows simulation results that confirm the effectivenessof the proposed constrained CKF (CCKF). Potential profits that canbe achieved while applying the proposed approach in constrained estimationproblems are shown compared to the conventional moving horizon estimation(MHE) algorithm. [ABSTRACT FROM AUTHOR]