On covariance propagation using matrix continued fractions

Publisher Website
Google Scholar
Abstract
The use of matrix continued fractions provides new methods to study the covariance matrix of the discrete Kalman filter. This approach can be used to obtain approximations to the covariance matrix using a reduced number of calculations and computer storage. Sequential bounds can also be calculated from this procedure and the bias resulting from the use of these bounds is determined. This approach has application for very large dimensional systems in which it is not possible, in practice, to store large matrices in computer memory or to perform many matrix computations.

This publication has 13 references indexed in Scilit: