Dynamic Geometry Fitting for Probabilistic Localization of Auv for Mooring Line Inspection

This paper addresses the challenge of improving Autonomous Underwater Vehicle (AUV) self-localization, providing precise navigation for Floating Offshore Wind Turbine (FOWT) mooring line inspections. Building on previous works proposing a mooring line tracking method for AUVs, we employ a particle filter algorithm that integrates geometry fitting for detected mooring lines with dead-reckoning data, without requiring additional acoustic positioning systems. The method was validated through simulation experiments for both catenary and taut mooring line scenarios under various sensor measurement error conditions. Results demonstrate that the proposed method significantly improves AUV localization and mooring line estimation performance compared to dead-reckoning alone and is expected to provide a robust solution for inspecting underwater structures.