Sequential data reconciliation algorithms have been developed for input-output models in linear dynamic systems. Existing filtering methods do not treat the case where there are measurement errors in the input variables. In our approach, the measurement errors in the input variables are optimally handled by the least squares method. This method shows good performance for input-output models.
Ayral TE, "Criteria for Comparing Data Reconciliation Software Products," IN CONTROL, VII(3), May-June, 1994
Darouach M, Zasadzinki M, AIChE J., 37, 193, 1991
Darouach M, Ragot J, Fayolle J, Maquin D, "Data Validation in Large-Scale Steady-State Linear Systems," World Cong. on Scientific Computation, IMACS-IFAC, Paris, 1988
Kim IW, Liebman MJ, Edgar TF, Comput. Chem. Eng., 15, 663, 1991