基于D-NSGA-Ⅲ算法的无人机群高维多目标任务分配方法

随着人工智能技术的快速发展, 种类繁多的无人机在军事领域得到了广泛应用。受单平台资源配备和执行能力限制, 大多数复杂任务需由多个无人机协同完成, 最优任务分配是其中需解决的重点和难点问题之一。最优任务分配方案求解问题已被证明是一个NP难问题, 针对多无人机系统的组织架构, 将非支配排序遗传算法与岛屿模型、主从模型结合, 构建一种分布式高维多目标演化算法D-NAGA-Ⅲ并对实际应用场景中4个目标进行优化, 并引入迁移策略和贪心算法对任务分配方案进行局部提升, 提高算法寻优能力和解质量。实验结果表明: 该方法在求解高维多目标的分布式无人机任务分配问题方面具有一定的效果。

Many-objective task allocation method based on D-NSGA-Ⅲ algorithm for multi-UAVs

With the rapid development of artificial intelligence technology,a variety of unmanned aerial vehicles(UAVs)have been widely used in the military field.Due to the limitation of resource allocation and executive capability of single platform,most of complex tasks should be accomplished by the cooperation of multiple UAVs.Optimal task allocation is one of the critical and difficult problems to be solved,and it has been proven to be an NP-hard problem.Considering the organization architecture of multi-UAVs system,the nondominated genetic algorithm is combined with island model and master-slave model.A distributed many-objective evolutionary algorithm,named D-NAGA-Ⅲ,is built to optimize four objects in a real application,and a migration strategy and greedy algorithm are proposed to improve the optimization ability and enhance the solution quality.Experimental results show that the proposed method is effective in solving distributed many-objective task allocation problems in UAVs systems.