Abstract： In order to realize fast initial alignment of the strapdown initial navigation system (SINS) in a complex circumstance, a new strong tracking unscented Kalman filter (UKF) was proposed with strict derivation based on the UKF and the two conditions of the strong tracking filter. As this algorithm took full advantage of characteristics of SINS nonlinear selfalignment, the steps of the strong tracking UKF were simplified so that the calculation was reduced significantly and the realtime performance of the algorithm was improved. The steps of the algorithm were given with proofs. The superiority of the strong tracking UKF was analyzed after the process to calculate the suboptimal fading factor given. The simulation and test of the initial alignment of large azimuth misalignment SINS based on this new algorithm show the validity and superiority of the new strong tracking UKF.
Key words：