摘要
针对多种群遗传算法在处理复杂多峰函数优化问题时效率低下、容易早熟收敛等缺点,提出一种基于自主计算的双种群遗传算法。双种群包括一个主种群和一个协助种群,协助种群通过系统的内、外监视器动态地向主种群传递优良个体和调整迁移间隔,以帮助主种群进化,并改进适应度函数防止迁移者过早死亡以保持种群多样性。实验结果证明,该算法优于标准遗传算法和双种群的多种群遗传算法。
In order to overcome the disadvantage of multi-population Genetic Algorithm（GA） that it has low efficiency and tends to premature in multimodal-function-optimization, this paper proposes a dual population GA based on autonomic computing. It has two distinct population including a main population and a help population. The help population transmits superior individual and adjusts migrate strategy dynamically to help the main population evolution by interior or exterior monitor. The hnproved fitness function makes migrant not be weeded out prematurely which can maintain the diversity. Experimental results show that the algorithm outperforms Single Genetic Algorithm（SGA） and Multi-population Genetic Algorithm with two Populations（2PMGA）.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第24期189-191,共3页
Computer Engineering
关键词
遗传算法
多种群遗传算法
自主计算
监视器
适应度函数
Genetic Algorithm（GA）
multi-population Genetic Algorithm： autonomic computing
monitor
fitness function
