一种改进的多目标人工蜂群算法  

Improved Multi-objective Artificial Bee Colony Algorithm

作　　者：陈伟栋[1] 童华刚 郜振华[1] 张洪亮[1] 

机构地区：[1]安徽工业大学管理科学与工程学院,安徽马鞍山243000

出　　处：《南华大学学报（自然科学版）》2017年第2期56-61,共6页Journal of University of South China：Science and Technology

基　　金：安徽省教育厅人文社科重点项目(SK2016A0168);安徽省社科联社会科学创新发展课题(A2015070)

摘　　要：由于基本多目标人工蜂群算法存在着种群盲目搜索、算法开发能力有限等缺点,在利用人工蜂群算法求解多目标优化问题时,提出了一种改进的人工蜂群算法.通过在算法中引入自适应搜索机制和变异机制等操作,使得种群个体可以有针对性地进行更新,同时也大大提高了种群个体的多样性.最后利用几种多目标的测试函数对改进前后的多目标人工蜂群算法的性能进行测试,结果表明:改进后多目标人工蜂群算法具有良好的算法收敛性和均匀性.In the use of artificial colony algorithm to solve multi-objective optimization problems,the standard multi-objective artificial colony algorithm has some shortcomings, such as population blind search,limited algorithmic development capability and so on. An improved artificial colony algorithm is proposed in this paper,which introduces the adaptive algorithm search mechanism and the mechanism of variation operation to make the individ-ual species be targeted for updates and improve the diversity of population greatly. Finally several kinds of multi-objective test function are used to verify the standard and improved multi-objective artificial colony algorithm performance,the comparison result indicates the improved multi-objective artificial colony algorithm has better convergence and uniformity.

关 键 词：人工蜂群算法 多目标优化问题 自适应搜索 变异操作 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]