A bilevel task allocation method for heterogeneous multi-UUV recovery system. (15th April 2023)

• Record Type: Journal Article
• Authors: Zhang, Wei; Li, Zixuan; Wu, Wenhua; Yang, Yiming
• Is Part Of: Ocean engineering. Volume 274(2023)
• Publication Date: 2023-04-15
• Subjects: Unmanned underwater vehicle -- Recovery system -- Task allocation -- Multiple intelligent agents -- Limited communication; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2023.114057
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: This paper presents a novel task allocation method for multiple unmanned underwater vehicles (multi-UUV) recovery systems. Specifically, a bilevel model that enjoys the advantages of centralized and distributed methods is constructed for mothership and cooperative constraints. In the centralized upper-level, an improved Gaussian mixture mode (IGMM) is presented, and an improved discrete crow search algorithm (IDCSA) is proposed. Thus, the leaders are clustered into subgroups, and the matching relationship between motherships and subgroups is established. In the distributed lower-level, the proposed consensus-based bundle algorithm with group constraint (CBBA-GC) addresses the allocation problem between subgroups and heterogeneous follower UUVs. It is found that the proposed bilevel method can achieve conflict-free and load-balanced multi-UUV recovery task allocation when the communication capability of each UUV and motherships recovery ability are heterogeneous. Highlights: In view of the unfavorable factors that restrict the heterogeneous multi-UUVs recovery progress in the real environment, mathematical models for practical task allocation are constructed. Combining the advantages of centralized and distributed methods, a new bilevel allocation method is proposed to generate UUV subgroups and allocate subgroups to suitable motherships. Several algorithms are selected according to the properties of the different levels, and were modified with proposed improved methods to meet the requirements of the multi-UUVs recovery task allocation, which can fully prove the effectiveness and robustness of the bilevel method.