约束条件下的多无人机协同任务分配方法  

Cooperative task assignment method of multiple UAVs under multiple constraints

作　　者：盛景泰 杜亚男 SHENG Jingtai;DU Ya’nan(The 51^(th)Research Institute of China Electronics Technology Group Corporation,Shanghai 201800,China;The 38^(th)Research Institute of China Electronics Technology Group Corporation,Hefei 230088,China;College of Information and Communication Engineering,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]中国电子科技集团公司第五十一研究所,上海201800 [2]中国电子科技集团公司第三十八研究所,安徽合肥230088 [3]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《应用科技》2023年第5期46-53,148,共9页Applied Science and Technology

摘　　要：为了在多约束条件下解决多个目标联合优化多无人机协同任务分配问题,本文综合考虑了任务约束、航程约束、攻击约束等约束条件,建立了攻击效益、损耗代价、时间代价联合为目标函数的多无人机协同作战任务分配模型。为简化问题难度,将带有多个约束条件的有约束问题转化为带有惩罚项的无约束问题,并将食肉植物算法与量子演进机制相结合,设计出一种新的量子食肉植物算法来优化模型中的目标函数。该算法的单链量子编码方法提高了量子食肉植物算法的收敛性能,克服了已有算法解决多无人机作战任务分配问题时易陷入局部收敛的弊端。在6种不同的作战规模中,量子食肉植物算法相比于4种对比算法的仿真结果均展现出更加优越的性能,突破了现有任务分配方法的应用局限和应用局限。In order to solve the collaborative task allocation problem of multi-UAV for joint optimization of multiple objectives under multiple constraint conditions,this paper comprehensively considers constraints such as task constraints,range constraints,and attack constraints,a task allocation model of multi-UAV cooperative combat with combined attack benefit,loss cost and time cost as the objective function was established.To reduce the difficulty of the problem,the constraint-included problem with multiple constraints was transformed into a constraint-free problem with a penalty term,and combine the carnivorous plant algorithm with quantum evolution mechanism,a new quantum carnivorous plant algorithm is designed to optimize the objective function in the model.The single-chain quantum coding method of the algorithm can improve the convergence performance of the quantum carnivorous plant algorithm and overcome the disadvantages that the previous algorithms may fall into local convergence when solving the combat task assignment of multi-UAV.In 6 different combat scales,the quantum carnivorous plant algorithm showed superior performance to the simulation results of 4 comparison algorithms,and break through the application limitations and limitations of existing task allocation methods.

关 键 词：无人机控制 任务分配 约束条件 攻击效益 有约束问题 量子计算 食肉植物算法 收敛性能 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] V279[自动化与计算机技术—控制科学与工程]