An Underwater Cooperative Localization Framework Based on Multiple Ranging Circle Intersection

Cooperative localization between autonomous surface vehicles and autonomous underwater vehicles (AUVs) represents a highly promising approach for accurate AUV localization. Good estimation accuracy and consistency are crucial to improve operational precision and reliability. However, current mainstream AUV localization methods, which primarily utilize Kalman filter-like techniques, often struggle to achieve estimation consistency due to inappropriate handling of measurement nonlinearity. This paper therefore proposes an underwater cooperative localization framework for AUVs based on multiple ranging circle intersection. The framework begins by generating multiple potential position estimates through the joint solutions of historical ranging equations. Ambiguous solutions are then systematically removed according to the shape of their overlapping uncertainty regions. The remaining estimates are finally fused by the proposed optimal fusion algorithm. Compared to conventional underwater cooperative localization methods relying on Kalman-like filters, the proposed algorithm demonstrates three major benefits in both simulations and lake-water trials: independence of the initial estimates, improved estimation accuracy and consistency, as well as faster convergence.