A Robust In-Motion Alignment Method for a DVL-Aided SINS With Latitude Uncertainty

Strapdown inertial navigation system (SINS) is a vital navigation equipment for underwater vehicles. Achieving accurate in-motion alignment of SINS requires the aid of a Doppler velocity log (DVL). Affected by the underwater working environment, the velocity provided by the DVL often contains outliers, which can negatively impact the alignment results of the SINS. In addition, obtaining latitude in the underwater environment from the outside is difficult, further complicating the SINS alignment process. This article proposes a robust in-motion alignment method for the DVL-aided SINS with latitude uncertainty to address these challenges. The alignment process consists of two stages: coarse alignment and fine alignment. In the coarse alignment stage, a DVL-aided in-motion latitude estimation method that incorporates noise suppression capabilities for the inertial measurement unit (IMU) of the SINS is developed. To mitigate the impact of outliers provided by the DVL on latitude estimation, a robust filtering model based on constructing gravity vectors (CGV) is proposed for outlier detection. The coarse attitude and latitude are then obtained using the multivector attitude determination method. In the fine alignment stage, the stored navigation data and Kalman filter (KF) are used to complete the fine alignment through the designed backtracking navigation framework. This not only reduces the alignment time but also directly transfers to the integrated navigation stage with latitude uncertainty, continuing to prevent the influence of outliers. Simulation and field tests verify the effectiveness and advantages of the proposed algorithm.