An intelligent positioning method for underwater vehicle under hydroacoustic noises

Abstract

Underwater vehicle positioning is the foundation of ocean exploration, target tracking, and multi aircraft collaboration. As the only carrier for long-distance information transmission, underwater acoustic positioning is easily affected by spatiotemporal frequency variations and multipath interference, resulting in large errors. Hence, this paper proposes an intelligent positioning method for underwater vehicle that can resist underwater acoustic noises. Firstly, based on the slant range, azimuth angle, pitch angle, and motion parameters, the measurement equations and state equations are established for underwater vehicle estimation. Secondly, utilizing the cubature Kalman filter to approximate the motion characteristics of positioning system, the Particle filtering is jointly used to reduce the deviation between importance density and posterior distribution. Thirdly, a sequence weighted correction strategy is introduced to dynamically update particle weights, and the hierarchical resampling is performed based on particle quality evaluation indicators. Finally, the multiple parameter simulations combined with platform experiments are conducted for accuracy comparisons and analyses on the lake. The experimental results show that the proposed positioning method has better positioning accuracy and stability compared to the relevant algorithms, and can be used for the positioning and tracking of actual underwater vehicles.