一种参数动态调整的自适应微粒群优化算法

Adaptive Particle Swarm Optimization Algorithm with Dynamically Adjusting Parameters

提出了一种参数动态调整的自适应微粒群优化算法.针对微粒群算法中不同适应度值的微粒所需要的搜索能力不同,引入微粒相对优秀度概念,通过相对优秀度来动态调整惯性权重和加速因子,有效地调节算法的全局和局部搜索能力,保持了微粒的个性.利用三个Benchmark函数进行数值试验,仿真结果表明,算法稳定,具有较好的收敛性能.

Aiming at particles with different fitness value needing for different search capabilities, this paper presents an adaptive particle swarm optimization algorithm with dynamically adjusting parameters. The concept of particle relative excellent degree is introduced to dynamically adjust inertia weight and accelera- tion coefficient, which improves the performance of the globe search and local search and maintains particle individuality. Experiment simulations of three benchmark functions show that the proposed algorithm has powerful convergence ability and good stability.