Using Pickard’s method for calculating the covariance matrices in the discrete Kalman filters.

Abstract

In the course of creating Kalman filter it is inevitable that a continuous linear stochastic system must be transformed to its discrete equivalent. In this paper a new method for calculating covariance matrices in the discrete system is developed on the basis of Picard’s iterative process. Substantial calculating advantages of the proposed method compared to routine methods as applied to Kalman algorithm are shown.