改进的人工蜂群算法  

Improved Artificial Bee Colony Algorithm

作　　者：郝继升[1] 井文红 任浩然[1] 

机构地区：[1]延安大学数学与计算机科学学院,陕西延安716000

出　　处：《江西科学》2017年第2期242-246,共5页Jiangxi Science

基　　金：陕西省高水平大学建设专项资金资助项目(2012SXTS06);延安市科技局项目(2014ZC-6)

摘　　要：人工蜂群算法是一种新兴的群智能优化算法,以其独特的寻优机制被广泛应用。然而,该算法存在着"早熟"收敛和进化后期搜索能力较差的缺点,针对这一问题,采用反向学习的种群初始化方法,并引入受差分进化算法思想启发的搜索方程,提出了一个改进的人工蜂群算法(简记为DEABC)。通过5个测试函数的仿真实验并与其他算法比较,结果表明DEABC算法具有更好的优化效率和优化性能。Artificial bee colony algorithm is a new group of intelligent optimization algorithm,with its unique optimization mechanism is widely used. However, this algorithm has the shortcomings of ＂early maturing＂ convergence and poor search ability in late evolution. In view of this problem, we use the method of initial learning of reverse learning and introduce the search equation inspired by differential evolution algorithm, and propose an improvement Artificial Bee Colony Algorithm （abbreviated as DEABC）. Simulation results of five test functions are compared with other algorithms. The results show that the DEABC algorithm has better optimization efficiency and optimization performance.

关 键 词：人工蜂群算法 差分进化算法 种群初始化 搜索方程 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]