Bearing-Only Aided Bearing, Elevation, Depth Difference Self-localization for Multiple AUVs

This paper proposes a self-localization method for a fleet of small-sized autonomous underwater vehicles (AUVs) that uses a one-way acoustic positioning via wireless broadcast communication. The one-way acoustic signal transmission allows a more frequent reception of messages and position measurements by each AUV in the fleet. In this method, a state observer inputs the bearing, elevation, and depth difference (BEDD) measurements as a default, but it inputs only the bearing when the elevation angle measurement becomes erroneous. Unlike the one-way travel time range measurement, the BEDD method does not require time synchronization between the AUVs. Monte Carlo Particle filter (PF) is the state observer of the self-localization, and it adopts a relaxed Gaussian likelihood function such that it retains some of the low likelihood particles. In the simulation, the proposed method prevented the position tracking of the PF from converging to a wrong position by rejecting the erroneous elevation angle measurements.