人工蜂群算法的改进  

Improvement of artificial bee colony algorithm

作　　者：赵红星[1,2] 常小刚 

机构地区：[1]兰州交通大学交通运输学院,甘肃兰州730070 [2]兰州交通大学现代信息技术与教育中心,甘肃兰州730070

出　　处：《计算机工程与设计》2018年第1期260-265,共6页Computer Engineering and Design

基　　金：兰州交通大学青年科学基金项目(2014027)

摘　　要：为改善人工蜂群算法(ABC)的深度搜索能力,提出一种改进的人工蜂群算法(SABC)。借鉴混合蛙跳算法(SFLA)的进化机制,将蜂群划分为多个模因组,使每个新个体与自身所在模因组的最坏个体进行优劣比较,能够更加容易保存群体中的"新生"个体,改善群体的整体质量,增加算法的深度搜索能力。通过7个测试函数进行实验,统计结果表明了SABC算法在求解函数优化问题时具有较好的算法性能。To improve the artificial bee colony algorithm(ABC) search capabilities,an improved artificial bee colony algorithm(SABC) was proposed.The evolutionary mechanism of the shuffled frog leaping algorithm(SFLA) was referenced.The colony was divided into multiple meme groups and each new individual was compared with the worst individual which belonged to its memo group.This method can easily save the new individual and the quality of the colony is improved effectively.Experiments were conducted on a set of 7 benchmark functions.The statistical results demonstrate good performance of SABC in solving numerical optimization problems.

关 键 词：人工蜂群算法 混合蛙跳算法 深度搜索 寻优速度 函数优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]