Two-Step Weighted Least-Squares Positioning Algorithm Based on AUV Navigation Velocity in Complex Underwater Acoustic Environments

This paper focuses on mobile positioning for autonomous underwater vehicle (AUV) in complex underwater environments. The positioning algorithm based on time of arrival (TOA) assumes that the underwater acoustic communication between AUV and underwater sensors is real-time and reliable. The distance between the AUV and underwater sensors can be obtained through the time delay of acoustic signal transmission. However, the uncertain and complex nature of the underwater environment can cause underwater sensors to fail to receive acoustic signals from the AUV reliably. This leads to an insufficient effective observation data in the positioning solution process, causing underdetermined problems in the solution equation, and thus making it impossible to solve the AUV’s position information stably in real time. To overcome this problem, we frist established a positioning model considering the navigation velocity of the AUV. Based on this Model, we then proposed a two-step weighted least squares algorithm combined with a fast horizontal distance model for the deep sea (TWLS-HD-V). Furthermore, we derived the Cram´er-Rao Lower Bound (CRLB) of the proposed model and compared it with that of the basic model. Finally, the effectiveness of the proposed algorithm was tested on both numerical simulations and deep-sea data. In conclusion, this algorithm effectively overcomes two key challenges in complex underwater environments: sound line bending and insufficient time delay information. As a result, it improves AUV positioning accuracy and ensures reliable positioning trajectories.