A New Calibration Method Based on Gravity-Assisted Navigation

Gravity-assisted navigation is an important means for underwater vehicles to achieve autonomous, passive, and long-endurance navigation. Among its key procedures, correcting the inertial navigation system (INS) based on gravity matching positions is of critical significance. In this paper, using shipborne measured INS data and gravity data together with a gravity database inverted from satellite altimetry data, the entire process of gravity-assisted navigation was experimentally validated through the normal time-frequency transform (NTFT) and Kalman filtering. NTFT is mainly used for gravity error correction, while Kalman filtering is primarily applied to matching algorithms. The analysis of the results shows that after 68.8 h, gravity-assisted navigation technology (GANT) can maintain navigation accuracy within 2 n mile (1 n mile ≈ 1852 m). Through comparison of different calibration methods, it is concluded that open-loop calibration applied only in the longitude direction is feasible and achieves satisfactory navigation performance. This surface vessel-based gravity-assisted navigation experiment addresses aspects such as experimental conditions and algorithmic techniques, serving as an important foundation for further underwater navigation applications.