摘要
针对传统免疫算法存在早熟收敛以及多样性不足的问题,提出一种基于知识域的多目标优化免疫算法。通过初始化知识域选择精英解,利用该精英解集自适应更新知识域的边界,从而维持算法收敛性与多样性的平衡。测试结果表明,相比NSGAII、SPEAII算法,该算法在运行时间、多样性以及覆盖性方面具有较大优势。
Aiming at the problem of premature convergence and insufficient diversity in traditional immune algorithm,this paper proposes a multi-objective optimization immune algorithm based on knowledge domain.The algorithm selects the elite solution by initializing knowledge domain,self-adaptive updates knowledge domain border by using this elite solution to maintain the balance between the convergence and diversity.Test results show that the algorithm has great advantage on convergence,diversity and run time.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第20期161-163,共3页
Computer Engineering
基金
南京邮电大学青蓝基金资助项目(NY207081)
关键词
知识域
多目标优化
免疫算法
knowledge domain
multi-objective optimization
immune algorithm
