航空学报 > 2026, Vol. 47 Issue (2): 332075-332075   doi: 10.7527/S1000-6893.2025.32075

基于时序耦合分析的无人机集群任务分配方法

王浩宇1, 张泽旭1(), 闻单1, 刘金龙1, 朱倍孝2, 包为民3   

  1. 1.哈尔滨工业大学 航天学院,哈尔滨 150001
    2.上海科技大学 信息科学与技术学院,上海 201210
    3.中国航天科技集团有限公司 科学技术委员会,北京 100048
  • 收稿日期:2025-04-03 修回日期:2025-04-17 接受日期:2025-05-13 出版日期:2025-05-28 发布日期:2025-05-27
  • 通讯作者: 张泽旭 E-mail:zexuzhang@hit.edu.cn
  • 基金资助:
    航空科学基金(2024Z023077001)

Task allocation algorithm for UAV swarm based on temporal coupling analysis

Haoyu WANG1, Zexu ZHANG1(), Shan WEN1, Jinlong LIU1, Beixiao ZHU2, Weimin BAO3   

  1. 1.School of Astronautics,Harbin Institute of Technology,Harbin 150001,China
    2.School of Information Science and Technology,ShanghaiTech University,Shanghai 201210,China
    3.Corporation Science and Technology Committee,China Aerospace Science and Technology,Beijing 100048,China
  • Received:2025-04-03 Revised:2025-04-17 Accepted:2025-05-13 Online:2025-05-28 Published:2025-05-27
  • Contact: Zexu ZHANG E-mail:zexuzhang@hit.edu.cn
  • Supported by:
    Aeronautical Science Foundation of China(2024Z023077001)

摘要:

随着集群规模的增加和任务场景的复杂化,如何设计高效的任务分配算法已是无人机集群应用的巨大挑战。针对无人机集群对时序任务分配中因跨队列影响导致的分配效能失效和冲突问题,提出一种基于时序耦合分析的任务分配方法(TCATA)。首先,建立考虑无人机载荷、任务需求和任务时序约束的任务分配模型,并从效能函数设计和效能有效性2个方面对市场机制下任务时序耦合约束的影响进行分析;其次,构造集群内各无人机局部调整方案集,并通过通讯一致性获取全局调整方案集;随后,各无人机基于各调整方案效能大小和冲突关系,将全局调整方案集的决策问题建模为最大加权团问题,并采用贪婪算法进行求解;最终,确定任务的执行者并更新全局任务分配结果。仿真实验表明,在考虑通信延迟时间的情况下,求解无人机数和任务数均过百的大规模时序任务分配问题时,TCATA在求解效率和指标上显著优于分布式遗传算法,和时序一致性任务包算法和合同网算法相比,求解指标略优,但运行时间和迭代次数低于CBBA-TCC和CNP的50%,验证了TCATA求解大规模时序任务分配问题的有效性。

关键词: 无人机集群, 时序任务耦合约束, 分布式任务分配, 最大加权团, 市场机制

Abstract:

With the increasing scale of UAV swarm and the growing complexity of mission scenarios, designing efficient task allocation algorithms has become a significant challenge for swarm applications. To address the allocation failures and conflicts caused by cross-queue influences in temporal task allocation for UAV swarm, a Temporal Coupling Analysis-based Task Allocation method (TCATA) is proposed. Firstly, a task allocation model considering UAV payloads, task requirements, and task temporal constraints is established. The impact of task temporal coupling constraints under the market mechanism is analyzed from two aspects: performance function design and performance validity. Next, a local adjustment set is constructed for each UAV within the swarm, and a global adjustment set is obtained through communication consensus. Then, based on the performance magnitude and conflict relationships of each adjustment, the global adjustment scheme decision problem is modeled as a maximum weighted clique problem and solved by a greedy algorithm by each UAV. Finally, the executors of the tasks are determined and the global task assignment result is updated. Simulation experiments demonstrate that in solving temporal task assignment problems with hundreds of UAVs and tasks considering communication delays, TCATA significantly outperforms distributed genetic algorithms in both efficiency and performance metrics. Compared with the CBBA-TCC and CNP algorithm, TCATA achieves marginally superior performance while reducing executing time and iteration number by more than 50%, validating its effectiveness in large-scale sequential task allocation.

Key words: unmanned aerial vehicles swarm, temporal task coupling constraint, distributed task allocation, maximum weighted clique, market mechanism

中图分类号: