面向多模态函数优化的回溯克隆选择算法  

Backtracking clonal selection algorithm for multi-modal function optimization

作　　者：张英杰[1] 毛赐平[1] 

机构地区：[1]湖南大学信息科学与工程学院,长沙410082

出　　处：《计算机应用》2012年第7期1947-1950,共4页journal of Computer Applications

基　　金：国家自然科学基金资助项目(61174140);湖南省科技计划重点资助项目(2010GK2022)

摘　　要：针对多模态函数优化问题,提出了一种基于回溯机制的改进克隆选择算法——回溯克隆选择算法(BCSA),采用改进回溯机制和记忆库抗体抑制策略,保持了抗体的多样性,以增强算法的全局搜索能力;通过改进动态变异、选择与交叉操作提高算法收敛速度。典型的多模态函数测试结果表明:回溯克隆选择算法具有优良的全局搜索能力和搜索效率。To solve some existing problems in multi-modal function optimization,an improved Clonal Selection Algorithm（CSA） based on the backtracking mechanism,Backtracking Clonal Selection Algorithm（BCSA）,was proposed in this paper.The global search capability could be enhanced by using the improved backtracking mechanism and the restraining operation of memory antibodies,which maintained the diversity of antibodies.In addition,in order to improve the convergence speed,the improved dynamic mutation,selection and crossover operation were adopted.The results tested on typical multi-modal functions show that BCSA has a powerful performance in global search.

关 键 词：函数优化 多模态函数 克隆选择算法 回溯机制 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]