摘要
该文提出了一种基于有性繁殖的遗传算法。标准遗传算法(SGA)已经被成功地应用到很多进化优化问题上,但是对于复杂的多模态函数寻优时,会出现早熟收敛现象。为了解决这个问题,该文结合自然界最常见的有性繁殖现象,赋予了遗传个体年龄和性别特征,提出了基于有性繁殖的遗传算法的框架以及实现。通过建立年龄和有性遗传进化算子,对不同年龄和不同性别的个体赋予不同的进化控制参数,克服了早熟收敛问题且保持了群体的多样性。对运动模糊图像复原的测试,也证明了该方法的有效性。
In this paper, a genetic algorithm with age and sexual structure is proposed. Standard Genetic Algorithm has been successfully applied to many evolutionary optimization problems. But there is a 1 problem of premature convergence for complex multi - model functions. To solve it, the frame and realization of this genetic algorithm are presented. Age and sexual structures are given to the individuals referring the sexual reproduction popular in nature. During age and sexual operators, different evolutionary parameters are given to the individuals with different age and sexual structures. As a result, this genetic algorithm can combat premature convergence and keep the diversity of population . The testing for recovering the vague image proves that the genetic algorithm is effective.
出处
《计算机仿真》
CSCD
2006年第11期168-172,共5页
Computer Simulation
基金
江西省跨世纪学科带头人培养计划项目(第三批)
江西省测控研究中心开放基金项目(200104003)
关键词
遗传算法
早熟收敛
有性繁殖
图像复原
Genetic algorithm
Premature convergence
Sexual reproduction
Image recovering
