【摘要】 为给海洋垃圾清理提供准确的海面信息，使用船载无人机对存在海洋漂浮垃圾但位置和数量未知的区域进行识别定位。由于海洋垃圾的位置会受风和洋流的影响而移动，需为研究区域设置时间窗。考虑到无人机续航时间有限，船舶在协同搜索时为无人机提供电池更换服务。基于无人机摄像头拍照范围，引入网格划分的方法处理研究区域，生成航路点。为实现在优化船舶和无人机路径的同时总成本最小化，提出一种混合变邻域搜索算法，在设计4种邻域操作的基础上，根据Metropolis准则对新解进行筛选，并通过邻域搜索操作加速寻优。选择东海附近的一片海域进行实例研究，结果验证了算法的有效性。对无人机续航时间分析可得，使用续航时间更久的无人机更有利于降低总成本。更多还原

【Abstract】 In order to provide accurate sea surface information for marine debris cleanup,a shipborne drone is used to identify and locate areas where marine floating debris is existent,but its location and quantity are unknown. As the location of marine debris is affected by wind and ocean currents,a time window is set for the study area. Considering the limited endurance of drones,ships provide battery replacement service for drones during collaborative search. Based on the imaging range of the drone’s camera,a grid division method is introduced to process the study area and generate waypoints. To achieve the minimization of total cost while optimizing the routes of both ships and drones,a hybrid variable neighborhood search algorithm is proposed,the Metropolis criterion is used to screen the new solutions based on the design of four neighborhood operations,and the neighborhood search operation is carried out to accelerate optimization. A case study conducted in a sea area near the East China Sea validates the effectiveness of the algorithm. The analysis of drone enduration reveals that using drones with longer enduration is more conducive to reducing the total cost.更多还原