A behavior-based planning algorithm for tracking of deep-sea hydrothermal buoyant plumes with autonomous underwater vehicles

Deep-sea hydrothermal plume tracing capability is important for autonomous underwater vehicles (AUVs) to efficiently perform hydrothermal vents prospecting missions. This paper addresses the hydrothermal buoyant plume tracing problem. Based on the nature of hydrothermal plumes, a moth-inspired strategy for the tracking of hydrothermal buoyant plumes in three dimensions is investigated. And a behavior-based planning algorithm that implements the strategy is designed, which consists of three behaviors, tracking, reacquiring, and declaring. The tracking behavior is to navigate an AUV to move toward the sea bottom when the plume is detected. The reacquiring behavior maneuvers the AUV in two-dimensional horizontal plane to find the plume if the plume is lost during the tracking. The declaring behavior is to estimate the vent´s location when the AUV has tracked the plume to the vent. In addition, a mapping behavior is designed to collect the buoyant plume data above the vent after the vent has been localized, with which the vent´s location could be accurately estimated and the adjacent vents could be found. The behavior-based planning algorithm is implemented on a developed simulation environment. The simulation results demonstrate that the presented algorithm is effective for navigating an AUV to track a hydrothermal buoyant plume in three dimensions to its vent and finally localize the vent accurately.