北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2024, Vol. 47 ›› Issue (1): 65-71.

• 论文 • 上一篇    下一篇

基于动态分治的大规模多场站无人机应急救援优化方法

苏立晨1,赵浩然2,郭通2,杜文博2,李宇萌2   

  1. 1. 北京航空航天大学 自动化科学与电气工程学院
    2. 北京航空航天大学 电子信息工程学院
  • 收稿日期:2023-02-22 修回日期:2023-04-20 出版日期:2024-02-26 发布日期:2024-02-26
  • 通讯作者: 李宇萌 E-mail:liyumeng@buaa.edu.cn

Optimization Method for Large-Scale Multi-Site Unmanned Aerial Vehicle mergency Rescue Based on Dynamic Divide-and-Conquer Strategy

SU Lichen1, ZHAO Haoran2, GUO Tong2, DU Wenbo2, LI Yumeng2   

  • Received:2023-02-22 Revised:2023-04-20 Online:2024-02-26 Published:2024-02-26

摘要: 面对应急救援任务时间紧、需求量大、待救援点数量规模较大等特点,提出了基于动态分治的大规模多场站无人机应急救援优化方法。在充分考虑无人机平台约束和应急救援任务约束的基础上,以最小化累计救援时间为目标函数,建立了多场站无人机应急救援模型。基于该模型,提出了基于路径相似度的动态分治策略,根据救援点的耦合关系进行空间聚类,将大规模问题分解为若干个规模较小、且耦合度较低的子问题;提出了自适应扰动邻域的变邻域搜索算法,通过多维邻域的协同搜索和动态交互,实现大规模应急投送方案的高效寻优。以典型样本为例,与先进元启发算法在不同规模的数据集上进行了对比,结果验证了所提方法能够有效地缩短应急投送的时间,为高效的灾后应急救援任务提供技术支撑。

关键词: 无人机任务规划, 车辆路径规划, 混合整数规划, 邻域搜索算法

Abstract:  Since the emergency rescue tasks need tight time, large demand, and large scale of rescue points, a large-scale multi-site unmanned aerial vehicle emergency rescue optimization method based on dynamic divide and conquer is proposed. In particular, a multi-station unmanned aerial vehicle emergency rescue model is established with the goal of minimizing the cumulative resume time. Besides, the model consideres the unmanned aerial vehicle platform constraints and emergency rescue mission constraints. Based on the model, a dynamic divide-and-conquer optimization framework based on path similarity is established and spatial clustering is performed based on the coupling relationship of rescue points. Then, large-scale problems are decomposed into several smaller-scale sub-problems with low coupling degrees. Finally, a variable neighborhood search algorithm for adaptive perturbation neighborhoods is proposed to achieve efficient optimization of large-scale emergency delivery plans through collaborative search and dynamic interaction of multi-dimensional neighborhoods. The simulation takes typical samples as an example and compares the proposed method with the advanced metaheuristic method on samples of different sizes. It is verified that the proposed method can effectively shorten the emergency delivery time and provide technical support for efficient post-disaster emergency rescue missions.

Key words: mission planning of  ummanned aerial vehicle, vehicle routing problem, mixed integer programming, neighborhood search algorithm

中图分类号: