摘要
针对二进制粒子群优化算法(BPSO),采用S形映射函数,将粒子在空间中飞行速度的正负值大小映射为其位置向量取1的概率,易于陷入局部最优解的问题,本文提出了采用弧形映射函数的二进制粒子群优化算法(ABPSO).该算法采用弧形映射函数取代BPSO中的S形映射函数,将速度平方值大小映射为位置向量改变的概率大小,当粒子具有较低的速度平方值时能够维持在原位置,较高的速度平方值时改变位置,从而使算法更好地收敛于全局最优解;同时,采用了无强制性位置更新程序,符合弧形映射函数使用速度平方大小映射为位置改变概率大小的需要.通过六个基准函数的仿真实验发现,ABPSO具有更好的收敛精度和更高的收敛速度;ABPSO采用更加符合粒子运动规律的弧形映射函数,表现出更好的收敛于全局最优解的能力和更高的收敛速度.
Aiming at the problem that binary particle swarm optimization algorithm （ BPSO） ,using S shaped transfer function transfers particle＇s flight speed in space as the probability being one of the position vector, easily fall into the local optimal, this paper proposes a binary particle swarm optimization algorithm with arc shaped transfer function （ ABPSO ）. The algorithm uses are shaped transfer function to replace S shaped transfer function in BPSO,transfering the square value of the velocity as the probability of changing the position vector, so that particles in a lower rate of square value can be maintained in the original position, the higher the value of the square change position, which makes the algorithm better convergence to the global optimal solution;Through the simulation experiment of six benchmark functions showed that ABPSO has better convergence accuracy and higher convergence speed;ABPSO using arc shaped transfer function which is more consistent with the particle motion, shows better ability to converge to the global optimal solution and higher convergence speed.
出处
《小型微型计算机系统》
CSCD
北大核心
2017年第12期2637-2640,共4页
Journal of Chinese Computer Systems
基金
2016年国家社科基金年度项目(16BTQ084)资助
关键词
二进制粒子群优化算法
弧形映射函数
S形映射函数
binary particle swarm optimization algorithm
arc shaped transfer function
S shaped transfer function
