KF/UKF based SINS initial alignment under large azimuth misalignment angle

In order to solve the problem that the calculation complexity increase sharply with the state dimension increment, the KF/UKF algorithm is proposed by decomposing the initial alignment model into linear part and nonlinear part. In this paper, the filtering process is designed and the algorithm is theoretically proved to be the optimal estimation with the minimum mean square error. At last, the KF/UKF algorithm is compared with the traditional UKF algorithm by use of the simulation experiment. The results show that the KF/UKF algorithm is better both initial alignment accuracy and computational speed.