Localization algorithm for a fleet of three AUVs by INS, DVL and range measurements

Abstract

The strong research efforts undergone in Marine Robotics during the last two decades open great possibilities for maritime operations. Autonomous Underwater Vehicles (AUVs) take the crew away from dangerous situations at depths of sea, with available and cost-affordable technology. MDM Lab, the Laboratory of Mechatronics and Dynamic Modelling of the University of Florence, is partner of the Thesaurus project, funded by Regione Toscana, with the aim of designing a moderate-cost AUV, called Tifone, to be used, in swarm, for research and monitoring of archaeological sites. In this paper we propose a Localization Algorithm for a fleet of three vehicles and a surface support ship which employs several sensors. Each AUV is equipped with low-cost sensors such as IMU, magnetometer and depth sensor; two of them have an high accuracy velocity sensor such as the Doppler Velocity Log. Acoustic modems, used for communication between AUVs, can provide the time of flight from an AUV to an another one. The positions estimated with a Range-Based algorithm fuse with the ones estimated by INS&DVL, avoiding unbounded error growth in the position estimate of the AUVs.