A Fast In-Motion Alignment Method for INS/DVL in Rough Sea Conditions

Inertial navigation system/Doppler velocity log (INS/DVL) integrated navigation is a mature underwater navigation method, where the initial alignment is a critical stage in its application. Alignment accuracy and alignment time are the main challenges for improving alignment performance. In this article, the initial attitude estimation errors caused by acoustic outliers and limited observation time will be primarily investigated. To address this problem, a novel robust coarse alignment method is designed based on a globally optimal model, providing sufficient accuracy of initial attitude for fast alignment during the fine alignment stage. In addition, to achieve fast dynamic alignment in a limited observation information environment, a dynamic alignment model based on time sequence reconstruction is proposed. The algorithm has been validated through simulations and field tests, demonstrating comparable accuracy to the current state-of-the-art approach as well as exhibiting better robustness and faster alignment speed under dynamic conditions.