南京大学学报(自然科学版) ›› 2016, Vol. 52 ›› Issue (4): 647–.

• • 上一篇    下一篇

基于基因重组的新颖克隆选择算法

张伟伟,林静静,景红蕾,王 晓*   

  • 出版日期:2016-07-24 发布日期:2016-07-24
  • 作者简介: 郑州轻工业学院,郑州,450000
  • 基金资助:
    基金项目:国家自然科学基金青年科学基金(61403349),2014年度河南省教育厅科学技术研究重点项目基础研究计划(14B520066),2015年度河南省教育厅科学技术研究重点项目基础研究计划(15A520033),郑州轻工业学院博士基金(2013BSJJ044),郑州轻工业学院研究生科技创新基金,郑州轻工业学院大学生科技活动项目收稿日期:2016-06-15 *通讯联系人,E­mail:pandaxiaoxi@163.com

A novel immune clonal selection computation with genetic recombination

Zhang Weiwei,Lin Jingjing,Jing Honglei,Wang Xiao*   

  • Online:2016-07-24 Published:2016-07-24
  • About author:Zhengzhou University of Light Industry,Zhengzhou,450000,China

摘要: 受生物免疫系统免疫机制的启发,提出了一种新的克隆选择算法RBCSA.首先,根据B细胞免疫反应中的基因重组原理引入了一种新的基因重组复合算子来加强种群个体之间的信息交互,进而提高算法全局搜索的能力.然后,对克隆选择算法中的超变异算子进行了改进,进一步加强了算法的局部搜索和寻优的能力.最后,结合新提出的基因重组算子和改进的超变异算子,提出了一种新的基于基因重组的克隆选择算法,并通过求解16个常用的全局最优化问题的经典测试函数进行仿真实验,结果表明RBCSA算法具有很好的平衡全局探索和局部寻优的能力,有效地提升了克隆选择算法的寻优性能,尤其对于高维最优化测试函数.另外,与现有其他进化算法相比,RBCSA算法显示出了很强的竞争力.

Abstract: Inspired by mechanism and principle of biological immune system,a novel recombination based clonal selection algorithm denoted as RBCSA is proposed.Firstly,a genetic recombination operator is put forward according to the principle of genetic recombination,to enhance the information interaction among individuals in the population,and furthermore,to improve the ability of global search.Then,the hypermutation operator in clonal selection algorithm is modified to strengthen the local search ability of the algorithm.Finally,the genetic recombination operator and modified hypermutaiton are combined together in the new proposed clonal selection algorithm.Sixteen common used benchmark functions in global optimization are employed for testing the proposed algorithm.The simulation results show that RBCSA algorithm has good performance in balancing the ability of exploration and exploitation,and is able to effectively improve the performance of clonal selection algorithm,especially for high dimensional optimization test functions.In addition,RBCSA algorithm is quite competitive in comparing with the state­of­the­art evolutionary algorithms.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!