Underwater Terrain Positioning Method Using Maximum a Posteriori Estimation and PCNN Model

Abstract

Conventional underwater navigation and positioning methods for Autonomous Underwater Vehicles (AUVs) either require the installation of acoustic arrays, which make AUVs less independent, or result in cumulative errors. This paper proposes an Underwater Terrain Positioning Method (UTPM) using Maximum a Posteriori (MAP) estimation and a Pulse Coupled Neural Network (PCNN) model for highly accurate navigation by AUVs. The PCNN model is used as a secondary discriminant to effectively identify pseudo-anchor points in flat terrain feature areas and to find the true positioning point, which significantly improves the matching positioning accuracy in these areas. Simulation results show that the proposed method effectively corrects Inertial Navigation System (INS) cumulative errors and has high matching positioning accuracy, which satisfy the requirements of AUV underwater navigation and positioning.