一种基于任务碎片化重组的多无人机能耗公平性算法  

A Fair Energy Consumption Algorithm for Multiple UAVs based on Task Fragmentation and Reorganization

作　　者：赵佳宜 赵文栋 刘存涛 刘泽原 ZHAO Jiayi;ZHAO Wendong;LIU Cuntao;LIU Zeyuan(Army Engineering University of PLA,Nanjing Jiangsu 210000,China)

机构地区：[1]陆军工程大学,江苏南京210000

出　　处：《通信技术》2021年第8期1947-1954,共8页Communications Technology

基　　金：国家自然科学基金项目(No.61702545)。

摘　　要：多架无人机在同时执行多个侦察目标不一致的数据摆渡任务时,由于侦察目标点数目及分布差异,采取无人机与任务间的"一对一"分配方式会导致部分任务完成效果差,存在无人机能耗效率低等问题。因此,提出了一种基于任务碎片化重组的"多对多"任务分配方法。首先,根据目标点之间的位置关系对任务进行碎片化分解,采用聚类算法对不同间距的任务碎片进行重组;其次,将该问题建模为最小最大路径覆盖问题,以多无人机能耗公平性为目标进行优化,提出了一种基于聚类的改进遗传算法。最后,进行了仿真验证。仿真结果表明,与"一对一"的任务分配算法相比,基于任务拆分与重组的"多对多"协同方法,系统最大能耗降低32.0%~37.3%,标准差最多降低90.2%。When multiple UAVs perform multiple data ferry tasks with inconsistent reconnaissance targets at the same time, due to the differences in the number and distribution of reconnaissance targets, the "oneto-one" distribution mode between UAVs and tasks will lead to poor performance of some tasks and low energy consumption efficiency of UAVs. Therefore, this paper proposes a "many to many" task allocation method based on task fragmentation and reorganization. According to the position relationship between the target points, each task are fragmented, and the task fragments with different distances are reorganized by clustering algorithm. The problem is modeled as the Min-max Path Coverage Problem. Aiming at the fair energy consumption of multiple UAVs, an improved genetic algorithm based on clustering is proposed. The simulation results indicate that, compared with the "one-to-one" task allocation algorithm, the "many to many" collaborative method based on task splitting and reorganization can reduce the maximum energy consumption by 32.0% to 37.3% on average and the standard deviation by 90.2% at most.

关 键 词：无人机 协同任务分配 任务碎片化 聚类 能耗 遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]