Skip to Main Content
Homing and navigation capabilities are essential in many autonomous underwater vehicle (AUV) applications. The paper presents both problems with respect to a single beacon. The difficulties of this approach are due to the fact that a single range measurement does not completely constrain the beacon's position in the vehicle frame. In order to triangulate his position, the AUV needs to maneuver while measuring its displacements between ranges. In addition, range measurements are noisy and sometimes spurious, speed bias and underwater currents affect dead-reckoning measurements. Homing and navigation procedures are the same beginning. In a first time, we have an initialization phase which is necessary to obtain an initial estimate of the vehicle's absolute position and of the disturbances affecting the quality of the dead reckoned displacements. These initials estimates are refined during the actual displacement towards the beacon by means of a Kalman filter. Two kinds of navigation are used so as to maximize the information matrix and to maintain an accurate absolute position. We present also postprocessing results and a comparison with LBL navigation.
Robotics and Automation, 2002. Proceedings. ICRA '02. IEEE International Conference on (Volume:4 )
Date of Conference: 2002