Research on UUV Carrying Forward-Looking Sonar for Target Location Based on Spatial Analysis

Abstract

Forward-looking sonar (FLS) is one of the most commonly utilized sensors for ocean observation. Current research on ocean mapping and localization using FLS primarily focuses on small-scale simultaneous localization and mapping (SLAM) and single-target 3-D dense mapping. However, limited research exists on target localization for unmanned underwater vehicles (UUVs) equipped with underwater navigation system and FLS. This article addresses the problem of FLS-based target localization by establishing plane constraints and spherical constraints through spatial analysis. It was discovered that the plane constraint optimization method suffers from small gradients during depth optimization, while the spherical constraint method encounters multiple extreme point problem. To overcome these limitations, this study introduces an innovative target localization method that combines plane intersection with directional distance constraints. The proposed method formulates constraint equations by randomly selecting two observation results with significant differences. The solution with the greater depth is then selected as the accurate target location. This process is repeated to generate multiple solutions, and their average is computed to determine the final target location. The experimental results demonstrate that the proposed method is both more stable and computationally efficient.