Novel in-flight alignment based on observability analysis for SINS/GPS

To retard the accuracy decreasing due to the direct feedback of some system states with lower observability, a novel in-flight alignment method based on observability analysis and level-arm error compensation for SINS/GPS was proposed. The estimated value of Kalman filter is adaptive feedback through a factor which is the normalized observability of the state, and the level-arm error of GPS velocity which degrades the precision of GPS velocity is compensated. The semi-physical simulation was carried out, and the simulation results indicate that this new in-flight alignment can improve the accuracy of the SINS/GPS in the case of incomplete observable.