Information coupling degree based segregation for homogeneous swarm AUVs

Most existing work on the cooperative control of swarm autonomous underwater vehicles (AUVs) focuses on the consensus related issues that aim to make the swarm move as a cohesive whole. This paper, on the contrary, studies the self-organized segregation problem for homogeneous swarm AUVs. Motivated by the underlying mechanisms of animal segregation behaviors, an information coupling degree based approach is utilized to represent the interaction strength between AUVs. Then, by integrating the motion information of the most correlated AUV into the cooperative control law, a self-organized segregation algorithm is constructed. Moreover, theoretical analysis proves that the velocity of AUV will converge to that of its most correlated neighbor and the swarm AUV system will gradually segregated into separated clusters in a distributed fashion. Simulation results also demonstrate the effectiveness of the proposed algorithm.