Research on Adaptive Kalman Filter for INS

Due to sensor accuracy and noise interference and other reasons, the measured data may be inaccurate or even wrong. This will reduce the accuracy of the filter and the reliability and response speed of the Kalman filter, and even make the Kalman filter lose the stability. In this paper, a new INS initial alignment error model and observation model are derived for the errors in INS initial alignment. The adaptive Kalman filter is built to improve the stability and the accuracy of filter. The specific method is to make the adaptive Kalman filter manage to correct the filter online by getting the observed data. The simulation results show the proposed algorithm improves the accuracy of initial alignment in SINS, and prove the adaptive Kalman filter is effective. The main innovation in this paper is to manage to build the adaptive Kalman filter to modify the filter online by using the observed data. The adaptive Kalman filter algorithm improves the accuracy of the filter.