Using autonomous underwater vehicles (AUVs) for mapping underwater topography of sea-ice and icebergs, or detecting keels of ice ridges, is foreseen as enabling technology in future arctic marine operations. Wind, current, and Coriolis forces affect an iceberg’s trajectory, making automated mapping difficult. This paper presents a method aiming at enabling autonomous iceberg mapping using AUVs equipped with a multibeam echosounder by estimating the position and orientation of the iceberg. The method is based on a bathymetric simultaneous localization and mapping (SLAM) algorithm, namely, the bathymetric distributed particle filter SLAM (BPSLAM) algorithm. The proposed method estimates the AUV’s pose in an iceberg-fixed coordinate system. The relative states can be used for both guiding the vehicle to achieve complete coverage, as well as estimation of a consistent iceberg topography. The algorithm also provides an estimate of the iceberg’s drift velocity – an important parameter for the AUV trajectory planning as well as any related ice management (IM) operations. Two new weighting algorithms for the BPSLAM method are proposed, enabling batch processing of multibeam echosounder (MBE) measurements to ensure real-time operation without discarding information. The proposed method is demonstrated using a real iceberg topography taken from the PERD iceberg sightings database, with simulated AUV and MBE range measurements. The algorithm is also evaluated on a real world bathymetric dataset, collected using the HUGIN HUS AUV.
Note from Journals.Today : This content has been auto-generated from a syndicated feed.