A Novel DVL Aided Inertial Navigation Strategy for Long-Endurance Robust Positioning of AUVs

Shaohua Pan,Xiaosu Xu,Yiqing Yao,Liang Zhang,Maodong Xia

DOI: https://doi.org/10.1109/tiv.2024.3430479

IF: 8.2

2024-01-01

IEEE Transactions on Intelligent Vehicles

Abstract:Stable and reliable high-precision navigation and positioning is regarded as an urgent demand for marine intelligent transportation system. Medium and large autonomous underwater vehicles (AUVs) can be equipped with various navigation sensors to ensure positioning accuracy. In contrast, small AUVs have limited carrying capacity, and how to achieve reliable positioning for long endurance with more basic and necessary sensors is worthy of in-depth investigation. Therefore, a robust Doppler velocity log (DVL) aided inertial navigation strategy is designed in this survey. A tightly integrated system model based on DVL Doppler shift and ranging information aided inertial navigation system is established, which solves the problem of positioning performance degradation caused by error accumulation and beam failure in traditional methods. Based on consistency principle of sensor equipment, the system residual is identified by dead reckoning model and compensated for DVL terrain aided navigation, which solves the problem that systematic errors have great influence on nonlinear terrain and are difficult to identify. At the same time, the terrain entropy is introduced to design the confidence evaluation algorithm, which improves the adaptability of system in different terrain scenes. Simulation and lake test results show that the proposed method has higher positioning accuracy, reliability and stability, and is suitable for AUVs long-endurance navigation tasks.