摘要
针对粒子群优化(PSO)算法中存在的多样性差、易早熟收敛和鲁棒性不强等缺点,提出一种基于并行结构的多种群PSO算法。并行多种群算法在基本粒子群算法的基础上,增加了分别能加速和减缓粒子移动的两种更新公式,既改善了粒子群算法种群多样性差的缺陷,又能保证算法的计算精度。采用十种典型的标准测试函数对提出的算法进行了仿真实验。测试结果表明:与基本粒子群优化算法相比,并行多种群粒子群算法收敛速度更快,运算时间更短。
Aiming at the disadvantages of the particle swarm optimization(PSO)algorithm,such as poor diversity,premature convergence and weak robustness,a multi-swarm PSO algorithm based on parallel structure is proposed.On the basis of the basic PSO,the parallel multi-swarm algorithm adapts two updating formulas which can accelerate and slow down the particle movement respectively,which not only improves the poor population diversity of PSO,but also ensures the computational precision of the algorithm.Ten typical benchmark functions are used for simulation of the proposed algorithm.The results show that compared with the basic PSO algorithm,the multiswarm PSO algorithm based on the parallel structure has faster convergence speed and shorter operation time.
作者
张天佳
杨永胜
ZHANG Tianjia;YANG Yongsheng(School of Aeronautics and Astronautics,Shanghai Jiao Tong University,Shanghai 200240,China)
出处
《传感器与微系统》
CSCD
2020年第9期119-121,共3页
Transducer and Microsystem Technologies
基金
航空科学基金资助项目(20165557005)。
关键词
粒子群优化(PSO)算法
群体智能
并行结构
标准测试函数
particle swarm optimization(PSO)algorithm
swarm intelligence
parallel structure
benchmark functions
