Author(s)

Minjie Xia

Corresponding Author

Minjie Xia

Abstract

In the 21st century, extensive human engagement in activities such as marine resource exploration, maritime transportation, and maritime security has led to a deeper understanding of the oceans. This increased interest has spotlighted Autonomous Underwater Vehicles (AUVs), which possess the capability to operate independently and covertly, making them a preferred choice for diverse underwater tasks. However, as underwater missions grow in complexity, the limitations of individual AUVs have spurred the exploration of multi-AUV systems. These systems offer higher efficiency, greater intelligence, and enhanced fault tolerance, rendering them indispensable in both deep-sea search and rescue missions and national maritime security efforts. Task allocation is a central challenge in the realm of multi-AUV systems. Balancing factors like search environments, ocean currents, obstacles, and target locations, various intelligent optimization methods have been employed. This paper references representative works to explore genetic algorithm-based decentralized task allocation and Hungarian algorithm-based decentralized task allocation, examining their potential to enhance cooperative underwater searches. Centralized task allocation faces robustness and scalability concerns, necessitating decentralized alternatives, particularly for mobile multi-agent systems. The presented research underscores the significance of decentralized methodologies in optimizing multi-AUV task assignments. The paper elucidates these two approaches, discusses their implications, and concludes by presenting an analysis of the results. Through this study, the paper contributes to advancing the field of multi-AUV systems and their effective coordination for efficient underwater search operations.

Keywords

Decentralized Task Allocation and Collaboration; Multi-AUV systems；Maritime activities; Intelligent optimization