Nonlinear stochastic filtration of satellite measurements

Abstract Existing methods for processing satellite measurements are based on the use of either the least squares method in different versions, or with the known model of motion of an object – various modifications of the Kalman filter. At the same time, the Kalman approach is more accurate, since it takes into account the dynamics of the movement of the object and the history of estimates, but its significant drawback is the need for a priori knowledge of the equations of motion of the object. In this regard, a new approach is proposed to assess the navigation parameters of the object by satellite measurements. On the one hand, this approach takes into account the dynamic nature of motion parameters and the history of estimates, and on the other hand, free from restriction in the form of accurate knowledge of the equations of motion of an object. The effectiveness of the considered approach in comparison with the existing traditional methods of processing satellite measurements is confirmed by the results of numerical modeling.