海岛无人机多模式应急配送的模型与算法研究

Model and algorithm research on multi-mode emergency distribution of drones on an island.

中文关键词: 无人机配送  应急物流  选址-路径优化  旅行商问题  迭代启发式算法

英文关键词: Drone distribution  Emergency logistics  Location-routing optimization  Traveling salesman problem  Iterative heuristic algorithm

中文摘要:

      灾害情况下的海岛配送是应急物流的一种特殊应用场景。无人机作为解决应急物流问题的一种新型配送方式受到广泛关注，并涌现出了一系列配送模式。通过考虑无人机的成本、飞行距离、起降平台和地面配送成本，构建无人机直接配送、“点对点”双阶段配送、“回路”双阶段配送和双层物流配送网络四种模式。针对双阶段配送的多指标耦合特征，提出一种迭代启发式算法进行优化。围绕普陀山海岛场景，构建20-100个客户点和2-4个无人机起降平台的算例进行数值实验，分析并验证模型正确性和算法的性能。实验结果表明，四种模式模型的有效性得以验证；算法平均经过2~5次迭代，能够有效减少“回路”双阶段模式19.08%的无人机飞行距离，减少双层物流配送网络28.63%的地面配送距离。

英文摘要:

      Island distribution in disaster situation is a special application scenario for emergency logistics. Drone as a new distribution method to solve the emergency logistics problem has received wide attention, and a series of distribution modes have emerged. By considering the drone cost, flight distance, drone landing platform and ground distribution cost, four distribution modes are proposed: drone direct distribution, point-to-point bi-stage distribution, cyclic bi-stage distribution and two-level logistics distribution network. An iterative heuristic algorithm is proposed to optimize the multi-metric coupling characteristics of two-stage distribution. Based on the Putuoshan island scenario, 20-100 customers and 2-4 drone landing platforms are constructed for numerical experiment to analyze and verify the model correctness and algorithm performance. The experiment results show that the models of the four modes have validity, and the algorithm can effectively reduce the drone flight distance of the cyclic bi-stage mode by 19.08% and the ground distribution distance of the two-level logistics distribution network by 28.63% after 2-5 iterations on average.