This paper proposes the engagement model and optimal task assignment algorithm for small-UAV swarm operations in hostile maritime environments. To alleviate the complexity of a real engagement environment, several assumptions are made: in the proposed engagement model, a vessel can attack the UAV within a certain range with a constant kill probability rate; and the ability of a vessel to attack UAVs is reduced if the multiple UAVs are involved. The objective function for optimal task assignment is constructed from the engagement model which estimates the total damage to vessels as the engagement outcome. Considering computational time and non-convex nature of the optimization problem, a heuristic approach, SL-PSO (social-learning particle swarm optimization), is adopted to maximize the objective function. In particular, a modified SL-PSO approach is introduced to deal with the optimization problem in a discrete domain. Numerical simulation results for two scenarios are presented to analyze the characteristics of the proposed engagement model and the performance of the optimal task assignment algorithm in the given environment.

中文翻译：

---

敌对环境中无人机群作战的最佳任务分配

本文提出了在敌对海洋环境中小型无人机群作战的交战模型和最优任务分配算法。为了减轻真实交战环境的复杂性，做了几个假设：在提出的交战模型中，船只可以在一定范围内以恒定的杀伤概率攻击无人机；如果涉及多架无人机，则船舶攻击无人机的能力会降低。最佳任务分配的目标函数是根据交战模型构建的，该模型估计对船只的总损坏作为交战结果。考虑到优化问题的计算时间和非凸性质，采用启发式方法 SL-PSO（社会学习粒子群优化）来最大化目标函数。特别是，引入了一种改进的 SL-PSO 方法来处理离散域中的优化问题。给出了两种场景的数值模拟结果，以分析所提出的参与模型的特征和给定环境下最优任务分配算法的性能。