1.中国船舶集团有限公司第七一〇研究所;2.华中科技大学人工智能与自动化学院
1.NO.710 R&2.D Institute, CSSC;3.School of Artificial Intelligence and Automation, Huazhong University of Science and Technology
UUV集群在执行大范围搜索任务时分配方案的优劣对于提高任务执行效率至关重要。本文研究解决UUV集群同时进行多区域搜索中的任务分配问题，使得全部区域搜索完成时长最小。针对传统匈牙利算法无法高效解决不平衡任务分配的问题，提出一种改进匈牙利多轮分配算法。该算法通过多轮任务分配，实现空闲UUV高效利用和目标区域合理分配，通过在代价函数中引入边际代价和保守估计时长，大大减小了全部区域搜索完成时长。仿真实验结果表明，本文提出的算法相比传统匈牙利算法能够合理分配空闲UUV提高分配效率。此外，相较于仅采用搜索时长为代价函数，本文在代价函数中引入的边际代价和保守估计时长能够针对耗时长的区域最大限度缩短搜索时长，保证随着UUV数量的增加，全部区域搜索完成时长单调递减。
Task allocation is crucial to improve the efficiency when UUV performs large-scale searching tasks. This paper solves the task allocation problem of UUV swarm in multi-area searching, so that the completion time of area searching is minimized. Aiming at the problem that the traditional Hungarian algorithm cannot solve the problem of unbalanced task allocation, an improved Hungarian multi-round allocation algorithm is proposed. This algorithm achieves efficient utilization of idle UUVs and allocating multiple UUVs to target areas in multiple rounds. By introducing the marginal cost and conservative estimating time in the cost function, the completion time of area searching is greatly reduced. The simulation experiment results show that the algorithm proposed in this article can allocate idle UUVs and improve allocation efficiency compared with the traditional Hungarian algorithm. In addition, compared with the algorithm only using the search time as the cost function, our algorithm can shorten the search time as much as possible for long-time-consuming areas, ensuring that the completion time of area searching decreases monotonically as the number of UUVs increases.
