Application of a new Adaptive Kalman Filitering algorithm in initial alignment of INS

In order to prevent the filtering divergence and improve real — time of the system, Proposing a new type of sage-husa — based adaptive filtering algorithm on initial alignment method of inertial. The regular kalman filter algorithm suitable for using in noise statistical characteristics known circumstances, but most of the noise statistical characteristics unknown. To achieve the best filiter effect, Adaptive Kalman Filitering (AKF) algorithm use observed data and automatic on-line estimation and correction of noise statistical characteristics. Simulation results show that the algorithm for improving alignment accuracy.