High Precision Velocity Estimation of AUV in an Underwater Distributed Sensor Network

Accurate estimating the velocity of a mobile object like AUV has received much attention in ocean exploration. The positioning-velocity estimation method is commonly used, which relies on observation information from multiple sensors (as nodes) over a time interval in underwater acoustics. However, its estimation accuracy heavily depends on the measurement accuracy of the positions. Also, there is a contradiction between estimation accuracy and the update rate. Thus, we propose a high-precision velocity estimation method in this paper. Firstly, we establish a two-node velocity estimation model based on the propagation delay and the Doppler shift and analyze the solution under different conditions. Then, we deduce the error transfer formula of velocity estimation theoretically and analyze the error spatial distribution characteristics. Besides, we introduce the estimation method in an underwater distributed sensor network with two steps to further improve the resolution of the velocity estimation using the redundant information generated by the pairwise combinations of multiple nodes. Meanwhile, we also give the idea of the post-processing method using the Kalman filter combined with the proposed method. Both simulation analysis and the lake test corroborate the feasibility and effectiveness of this method and higher precision than the pve method.