A vision assisted navigation method suitable for polar regions used by autonomous underwater vehicle

Abstract

In polar region, the strapdown inertial navigation system (SINS) and the Doppler velocity log (DVL) integrated navigation system equipped on the autonomous underwater vehicle(AUV) cannot complete the autonomous initial alignment of the SINS for the convergence of geographical meridian and geomagnetic line. To solve the problem, an initial alignment method for vision assisted SINS based on grid coordinate system in polar region was proposed in this paper. Before launching, we fix a camera on the head of the AUV and put the AUV on a moving base. After that, by controlling the movement of the moving base, AUV will swing as well. During the swing, the misalignment angle, velocity error and position error of the SINS mounted on the AUV in the grid coordinate system can be inferred by using the location and posture information relative to feature points, which location is known, obtained by camera. This completes the initial alignment of the SINS in the polar region. After initial alignment, a simulation is used to verify the effectiveness of the proposed algorithm. In the simulation, Kalman filter is used to fuse the output of DVL/SINS to complete underwater navigation and position of AUV.