A NOVEL COOPERATIVE HUNTING ALGORITHM FOR MULTI-AUV IN UNDERWATER ENVIRONMENTS

Xiang Cao, Hongbing Sun, and Xinyuan Xu

Keywords

Multi-AUV, cooperative hunting, decentralized search, angle matching, direction optimization strategy

Abstract

As one of the challenges of multi-AUV (multiple autonomous underwater vehicle) systems, the realization of group security defence is of great significance. This article studies the multi-AUV co-operative hunting in two- and three-dimensional underwater environments. A comprehensive hunting algorithm based on besieging circle shrinking is proposed to control multi-AUVs to search and capture invaders. The invader hunting process consists of three steps: first, multi-AUVs search for invaders in unknown environments. To expand the search scope as far as possible and reduce the repeated search rate, a decentralized search strategy is adopted to search for invaders. Once any invader is detected by an AUV, the discoverer acts as the leader to form a dynamic hunting team. Next, the hunting team determines a besieging point based on an angle matching algorithm. AUVs use direction optimization strategies to navigate to the besieging point while avoiding obstacles. Finally, when all the AUVs reach their exact besieging points to satisfy the encirclement contraction condition, the encirclement will shrink to capture the invader. Simulations and experiments prove the feasibility and effectiveness of the proposed algorithm. Compared with the bio-inspired neural network algorithm, the proposed algorithm performs better.

Important Links:



Go Back