A Robust FOAM Initial Alignment Method for SINS Assisted by DVL

The in-motion initial alignment of the strapdown inertial navigation system (SINS) is crucial for improving the rapid response performance of the autonomous underwater vehicle (AUV). However, the acoustic information of Doppler velocity log (DVL) is susceptible to interference with fauna and flora underwater, and outliers are introduced into velocity measurement by DVL when the AUV is performing underwater operations, which affects the underwater initialization accuracy of the integrated navigation system. In this article, a robust fast optimal attitude matrix (FOAM) initial alignment method based on the independence test of vector error distribution is proposed for moving base. First, an alignment method based on FOAM is introduced, and the error composition and noise distribution of constructing inertial frame vectors are analyzed. Second, the exponential smoothing method is used to predict the trend term and noise variance of vector errors, and a robust FOAM alignment method based on the independence test of vector error distribution is designed. Finally, the robustness and feasibility of the proposed method are verified through Monte Carlo simulation and field tests. The alignment results demonstrate that the proposed alignment method has stronger robustness and practicality compared to the existing current popular method, in which the measurements of DVL are corrupted by the outliers.