The industry's recognized authority for design, engineering and application of equipment and services in the global ocean community
Issue link: http://sea-technology.epubxp.com/i/477967
28 st / March 2015 www.sea-technology.com not been found yet. Thus the deviations that occurred dur- ing the dives were relatively big and well-suited for testing the correction algorithm. In this article, the correction al- gorithm is explained at an exemplary dive (AWI number: PS80/187-2) during Arctic expedition ARK 27/2 in July 2012 (RV Polarstern). It represents the frst mission in which the foat maneuver was conducted below Arctic sea ice. Dur- ing this mission the foat maneuver was executed between 50 and 10 meters depth. Maximum horizontal distance be- tween the vehicle and RV Polarstern was approximately 2 kilometers. An analysis of the data shows the high consistency of the AUV data. In terms of vehicle speed and heading, all consecutive positions are realistic. Position updates are calculated in steady intervals without any interruptions. In this case the vehicle updated its position every 0.1 seconds. However, after a mission is accomplished and the vehicle is back on the surface, an unrealistic "jump" appears in the navigation data. As soon as the vehicle is on the surface it determines its position via GPS and corrects its INS-based navigation data accordingly. Plotting these raw data on a map, this correction process occurs as an apparent gap, or jump, in the track. baseline (USBL) systems, which can be operated with less effort, can be used alternatively. Since 2011, AWI has used the USBL system "GAPS" from iXBlue (Marly le Roi, France) to track the vehicle. According to iXBlue, this system offers a tracking range of 4,000 meters distance in every direction and an opening angle of 200°. Thus, the system can still track objects close to the surface, although the transducer is positioned underneath a ship´s hull. GAPS provides both the geographical position and the depth of a tracked object. The AUV itself is equipped with a KN-5053 INS manufac- tured by Kearfott (Little Falls, New Jersey). Additional sensors for navigation include a downward-looking Workhorse DVL operating at 300 kilohertz (Teledyne RD Instruments, Poway, California), a 4,000-meter-depth-rated Digiquartz pressure sensor (Paroscientifc, Redmond, Washington), a DG-14 GPS receiver (Thales Navigation, San Dimas, California) and an SBE 49 FastCAT CTD probe to determine sound velocities (Sea-Bird Electronics, Bellevue, Washington). Methodology In summer 2012 there was relatively little experience with the foat maneuver, and the optimal dive settings had (Top) Raw GAPS data over- laid by phase two data (left) and phase three "knots" (right). (Bottom) Reloca- tion of "fxpoints" onto time-synchronous knots during phase four (left) and the fnal result after the entire correction process (right).