A comparative study of continuous covariance propagation algorithms in sequential estimation

Several continuous factored covariance propagation algorithms are compared to determine their suitability for orbital navigation applications. The algorithms are evaluated in terms of the relative computational cost and computation accuracy via numerical simulations. Consideration is also given to the effect on the algorithm performance of the numerical integrators to be used with the navigation algorithms. The comparison indicates that the Tapley-Choe (1976) algorithm combined with a Carlson square-root measurement-update algorithm consistently offered superior performance with respect to computational efficiency and integration accuracy.