摘要
蚁狮算法作为一种新型的群智能仿生算法,由于其参数设置少,寻优精度高等特点,已被广泛的用于求解各种优化问题,但其关键参数值的设置没有理论指导,盲目设置往往会对算法本身性能造成显著影响。以经典统计学中的单因素方差分析为基础,设计了蚁狮算法的参数对收敛性能影响的效能实验,对算法的种群规模P、步长浮动因子γ、陷阱浮动因子β设置不同的参数水平,通过统计分析得出参数水平对蚁狮算法收敛速度和收敛精度等性能影响的一般规律,获得稳健和高效的优化效果。
As a new swarm intelligence algorithm,the ant lion optimizer(ALO)is widely used to solve various optimization problems because of its features of few parameter settings and high optimization accuracy.However,there is no theoretical guidance for the setting of the key parameters of ALO,and blind setting often has a significant impact on the performance of the algorithm itself.Based on the variance analysis in statistics,parameter efficiency experiments concerning ALO were designed,and the different parameter levels about population size P,step floating factorγand trap floating factorβwere analyzed respectively by this article.The general rule of the influence of parameter level on the convergence speed and accuracy of ALO was found by statistical analysis,and the robust and efficient optimization effect was obtained.
作者
赵转哲
张宇
刘永明
张振
ZHAO Zhuan-zhe;ZHANG Yu;LIU Yong-ming;ZHANG Zhen(School of Mechanical and Automotive Engineering,Anhui Polytechnic University,Wuhu Anhui 241000,China)
出处
《计算机仿真》
北大核心
2022年第3期330-334,共5页
Computer Simulation
基金
安徽省自然科学基金面上项目(1808085ME127)
安徽工程大学引进人才科研启动基金(2019YQQ004)。
关键词
蚁狮算法
参数效能
统计分析
Ant lion optimizer
Parameter efficiency
Statistic analysis
