Adaptive Kalman filtering for the integrated SINS/DVL system

The integrated strapdown inertial navigation system/Doppler velocity log (SINS/DVL) navigation system is one of these applications that contain time-varying statistics of the measurement noise. Although the Sage-Husa adaptive Kalman filtering (SHAKF) has been applied in the field where the dynamic system contains unknown noise statistics, the precision and stability of the SHAKF are still the primary matters need to be overcome. In this paper, an improved SHAKF method is proposed to heighten the performance of the filter in the SINS/DVL system, by smoothing the innovation covariance and raising the weights of a priori estimate error covariance. The results demonstrate the proposed method can reach better navigation accuracy than the SHAKF under time-varying measurement noises.