多飞行器协同任务分配的改进粒子群优化算法

Improved Particle Swarm Optimization Algorithm for Cooperative Task Assignment of Multiple vehicles

为提升多飞行器编队执行任务的效率和性能,提出一种用于多飞行器协同任务分配的改进粒子群优化算法。考虑飞行器任务能力约束,飞行器执行任务时付出的威胁代价、航程代价以及完成任务的收益,从而构造问题的数学模型。将粒子的位置属性编码为一组任务分配向量,从任务分配向量可解码出对应的任务分配解,实现粒子群优化算法解的离散化。为解决粒子群优化算法容易陷入局部收敛的缺点,提出一种跳出局部收敛的策略。该策略基于模拟退火算法,生成新粒子,以一定概率决定是否保留新粒子,并将跳出局部收敛的策略应用到传统粒子群优化算法中,建立可用于多飞行器协同任务分配的改进粒子群优化算法。数字仿真实验结果表明,与现有算法相比,所提算法能显著提高多飞行器任务分配的收益和效率。

An improved particle swarm optimization(PSO)algorithm for multi-aircraft cooperative task assignment is proposed.The corresponding mathematical model is developed by considering the constraints of the aircraft's capability,the threat cost,the range cost,and the benefits obtained by completing the tasks.The position attribute of particles is encoded as a set of task assignment vectors,from which we can decode the task assignment solution such that the PSO solution is discretized.In order to solve the problem that the PSO algorithm can easily fall into local convergence,a strategy of jumping out of the local convergence is proposed.Based on the simulated annealing algorithm,this strategy first generates new particles,and then decides whether to retain the new particles with a certain probability.Finally,this jumping-out strategy is applied to the conventional PSO algorithm so as to establish an improved one that can be used for multi-aircraft cooperative task assignment.The digital simulation results verify the effectiveness of the proposed algorithm.