抄録
This paper proposes a practical navigation method for an Autonomous underwater vehicle (AUV) to carry out visual observations of seafloors, especially those of shallow vent fields. Underwater vent fields are scientifically important while their hostile environment presents a real challenge for AUVs. The method localizes the vehicle with regard to natural bubble plumes as well as artificial landmarks found by an onboard profiling sonar, based on the idea of Simultaneous localization and mapping (SLAM).The path plan is also updated in real time with regard to the position of the vehicle, the landmarks and the pre-defined waypoints. The method was implemented in the AUV "Tri-Dog 1" and a series of experiments were carried out at "Tagiri" vent area, Kagoshima bay in Japan. The AUV succeeded in observation of interesting features of the field including bubble plumes, bacteria mats and tube-worm colonies through 12 fully autonomous dives with the total time of more than 12 hours.
