A unifying framework for discrete linear estimation Generalized partitioned algorithms
K. S. Govindaraj, DEMETRIOS G. LAINIOTIS · International Journal of Control · 1978
In this paper the generalized partitioned algorithms for the discrete linear estimation problem are presented. These serve as the unifying framework for linear estimation. The fundamental nature of the generalized partitioned algorithms (GPA) is demonstrated by showing that previous major filtering and smoothing algorithms can be obtained as special cases from the forward and the backward formulation of the GPA. A particularly effective algorithm is given for the steady-state solution of the discrete Riccati equation.