Improvements in Filters of Strapdown Inertial Navigation System

In this paper, a strapdown inertial navigation system (SINS) with improved filters has been presented. The performance of the SINS relies on the position and velocity data which is further affected by the type of filter. The classic Kalman Filter (KF) performs much worse in solving the state estimation issue for Gaussian white noise. The Maximum Versoria Criterion (MVC) is used to apply an adaptive filter to impulse noise. It is useful for non-Gaussian noise with a low exponential calculation. The Maximum Versoria Criterion Extended Kalman Filter (MVC-EKF) based estimation model has been proposed to reduce the impact of anomalous data on state estimation with reasonable computation cost. The root mean square error (RMSE) is evaluated under the EKF, MCC-EKF and the proposed algorithms. The MVC-EKF outperforms as compared to others. The mean calculation time improved by 34.9%. It is very beneficial for future navigation systems in the military.