考虑执行能力约束的多机协同目标分配AEPSO算法

Multi-UAV Cooperative Target Allocation AEPSO Algorithm Considering Execution Capability Constraints

针对复杂环境中的多无人机协同攻击目标分配算法存在约束条件不充分、解空间的多样性与收敛性难以平衡等问题,提出了一种考虑执行能力约束的多机协同目标分配自适应精英粒子群(AEPSO)算法。首先,在以攻击效益最大化、时间及损毁代价最小化为目标的基础上,将多无人机续航能力、跟踪攻击性能等差异性导致的执行能力受限作为约束条件,构建了多无人机协同目标分配模型。然后,提出了一种基于混沌初始化和多尺度协同变异的自适应精英粒子群策略:通过构建一种改进型Logistic映射,实现粒子混沌初始化,提高了初始粒子群的多样性及搜索遍历性;设计了带有不同方差的自适应高斯变异机制作为精英选择策略,增强了粒子群的全局和局部搜索能力,跳出局部最优;最后,将粒子收敛贡献值作为反馈信息自适应地调整粒子群参数,加快算法收敛速度。仿真结果表明,在不同的无人机与目标的分配关系下,所提算法以更快的收敛速度、更小的适应度值求解出最佳的分配方案。

Aiming at the problems of insufficient constraints and difficulty in balancing the diversity and convergence of the solution space in the multi-UAV cooperative attack target allocation algorithm in the complex environment,a multi-UAV cooperative target allocation based on adaptive elite particle swarm optimization(AEPSO)algorithm considering execution capability constraints is proposed.Firstly,based on the goal of maximizing attack benefits and minimizing time and damage costs,the multi-UAV cooperative target allocation model is constructed by taking the limited execution ability caused by the differences in endurance and tracking attack performance of multiple UAVs as constraints.Then,an adaptive elite particle swarm optimization algorithm based on chaotic initialization and multi-scale collaborative mutation is proposed.By constructing an improved logistic map,particle chaos initialization is achieved,which improves the diversity and search traversal of the initial particle swarm.Meanwhile,an adaptive Gaussian mutation mechanism with different variances is designed as the elite selection strategy,which enhances the global and local search capabilities of the particle swarm and jumps out of local optima.To accelerate the convergence speed of the algorithm,the particle convergence contribution value is used as feedback information to adaptively adjust the particle swarm parameters.The simulation results show that under different allocation relationships between UAVs and targets,the proposed algorithm solves the optimal allocation scheme with faster convergence speed and smaller fitness values.