考虑子系统执行能力的多无人机协同任务规划  

Multi-UAV cooperative mission planning considering subsystem execution capability

作　　者：张鸿运 王磊 张旭[1] 丁宇 吕琛[2,3,4] 王昕炜 ZHANG Hongyun;WANG Lei;ZHANG Xu;DING Yu;LYU Chen;WANG Xinwei(School of Mathematical Science,Dalian University of Technology,Dalian 116024,China;Science and Technology on Reliability and Environmental Engineering Laboratory,Beijing 100191,China;Institute of Reliability Engineering,Beihang University,Beijing 100191,China;School of Reliability and Systems Engineering,Beihang University,Beijing 100191,China;Department of Engineering Mechanics,Dalian University of Technology,Dalian 116023,China;State Key Laboratory of Structural Analysis for Industrial Equipment,Dalian University of Technology,Dalian 116023,China)

机构地区：[1]大连理工大学数学科学学院,辽宁大连116024 [2]可靠性与环境技术重点实验室,北京100191 [3]北京航空航天大学可靠性研究所,北京100191 [4]北京航空航天大学可靠性与系统工程学院,北京100191 [5]大连理工大学工程力学系,辽宁大连116023 [6]大连理工大学工业装备结构分析国家重点实验室,辽宁大连116023

出　　处：《系统工程与电子技术》2023年第1期127-138,共12页Systems Engineering and Electronics

基　　金：国家重点研发计划(2020YFB1709403);国家自然科学基金(12102077,12072059);辽宁省自然科学基金(2019-ZD-0021);中央高校基本科研业务费专项资金(DUT20YG125)资助课题。

摘　　要：任务分配是提高无人机运维效率的关键技术之一。针对子系统执行能力约束条件下的无人机蜂群协同任务分配问题,提出一种融合拍卖机制的改进狼群算法。首先,定义子系统能力矩阵以实现无人机异构性和任务执行能力的统一描述。其次,对个体狼采用矩阵编码,针对违反攻击次数的非可行解,提出基于拍卖机制的修正策略,以进行处理。然后,在个体狼位置更新过程中融入遗传算法思想,在探索阶段和围捕阶段分别进行相邻行交换操作和间隔列交叉操作,以实现快速寻优。最后,将第三优狼引入到狼群更新过程中,从而增强种群的多样性。仿真实验结果表明,所提方法能够有效求解子系统执行能力约束下无人机蜂群协同任务规划问题;且相比于其他改进进化算法,所提方法具有更好的寻优性与收敛速度。Task assignment is one of the key techniques to improve the operation efficiency of unmanned aerial vehicle(UAV).To solve the UAV swarm cooperative task assignment problem under subsystem capability constraints,an improved wolf pack algorithm incorporating auction mechanism is proposed.Firstly,a subsystem capability matrix is defined to facilitate the unified description of UAVs’heterogeneity and task execution capability.Secondly,the individual wolf is encoded by matrix,and a correction strategy based on auction mechanism is proposed to deal with the infeasible solution of the number of violation attacks.Thirdly,the idea of genetic algorithm is incorporated in the process of individual wolf position updating,and the adjacent row exchange operation and interval column crossover operation are carried out in the exploration stage and the round-up stage respectively to achieve rapid optimization.Finally,the third best wolf is introduced into the wolf pack renewal process,thus enhancing the diversity of the population.Simulation experiments show that the proposed method can effectively solve the multi-UAV collaborative task planning problem under subsystem execution capability constraints.Compared with other improved evolutionary algorithms,the proposed method has better optimality and convergence speed.

关 键 词：协同任务分配 狼群算法 矩阵编码 拍卖机制 子系统能力矩阵 

分 类 号：V279[航空宇航科学与技术—飞行器设计]