Using A Posteriori Moments for Estimating Orientations of Antennas of Radio Engineering Systems

Based on the approximation of the a posteriori probability density of a multidimensional Markov process by a system of a posteriori moments, an algorithm for nonlinear stochastic filtering of the antenna state vector of a radio-engineering system has been developed. The proposed suboptimal filter makes it possible to estimate the angles of head, roll, and pitch according to the indications of a global satellite navigation system using three spaced antennas. The advantage of using the proposed method compared with the widely used Kalman filter is shown in the case of non-Gaussian presentation of the state vector.