A NOVEL TIME-VARYING FORMATION OBSTACLE AVOIDANCE ALGORITHM FOR MULTIPLE AUVs

摘要

In this paper, a time-varying formation reconfiguration and obstacle avoidance strategy are investigated for a fleet of autonomous underwater vehicles (AUVs) in the three-dimensional(3-D) ocean environment. The proposed novel controller consists of improved artificial potential field, graph rigidity theory, and affine transformation (IAPF-GR-AT). Information manoeuvring and swarm AUVs are treated as a whole. First, when in an area containing obstacles, the IAPF method is used to plan a safe navigation route that matches the formation size. The planned path is equivalent to a virtual leader AUV, which serves as the leader AUV's tracking target. Second, the leader AUV and follower AUVs form a cubic structure based on the GR theory and then track the planned path. Third, the theory of AT is used for formation reconfiguration. The formation shape can be arbitrarily changed in 3-D space according to the environmental constraints or the needs of the mission, showing a dynamic characteristic and having good adaptability to the environment. In addition, when the formation AUVs pass through an obstacle area, the method is also conducive to avoiding the formation's long-term split. Finally, several simulation results for nine AUVs are offered to demonstrate the formation reconfiguration and obstacle avoidance performance of the designed controller.