多无人船协同多目标遍历路径规划

Cooperative Multi-USV Traversing Multi-Target Path Planning

针对远洋捕捞渔船数量急剧增加所导致的频繁碰撞事故和无序捕捞行为增多等问题,提出多无人船协同多目标遍历路径规划策略。该策略由任务分解、任务分配和路径规划等三层架构组成。任务分解是基于K-Means聚类算法实现,将待遍历目标分成多个任务,每个任务内含一个或多个目标;任务分配是基于拍卖算法实现,将任务分配给无人船,优化目标是总路径最短;路径规划是基于指针网络和A*算法实现,完成无人船对被分配任务内所有目标最短路径遍历,并避开静态障碍物。通过仿真试验,验证了该策略在多无人船协同的可行性。所提出的“分而治之”的策略在海上复杂障碍物环境下能够有效解决多无人船协同多目标遍历路径规划问题,可实现无人船路径规划的就近分配和任务量分配均衡,能够有效解决频发的渔船碰撞和无序捕捞问题,还能大幅提升远洋渔业的效率。

To address the issues of frequent collision accidents and the increase in unregulated fishing activities caused by the sharp rise in the number of offshore fishing vessels,a multi-unmanned vessel cooperative multi-target traversal path planning strategy is proposed.Propose a strategy for multi-USV(unmanned surface vessels)cooperative multi-target traversal path planning.The strategy consists of three layers:task decomposition,task allocation,and path planning.Task decomposition is achieved using the K-Means clustering algorithm to divide the targets into multiple tasks,each containing one or more targets.Task allocation is accomplished through the auction algorithm,assigning tasks to USVs with the objective of optimizing the total shortest path.Path planning utilizes the pointer network and A*algorithm to determine the shortest path traversal for all targets within the assigned tasks,while avoiding static obstacles.The feasibility of this strategy for multi-USV cooperation has been validated through simulation experiments.The proposed"divide and conquer"strategy effectively solves the problem of multi-UsV cooperative multi-target traversal path planning in complex maritime environments.It enables the near allocation of USV path planning and balanced task allocation and effectively resolve frequent fishing vessel collisions and unregulated fishing issues,and significantly enhance the efficiency of offshore fisheries.