期刊论文

Zhang, Chi(carol_zc@126.com)

中文

中南大学学报(自然科学版)

1672-7207

2016

47

6

1940-1946

10.11817/j.issn.1672-7207.2016.06.016

湖南省教育科学“十二五”规划课题（XJK014CGD013）； 益阳市科技计划项目（2014JZ52）~~；

本校为其他机构

根据免疫算法多样性保持能力不足、易陷入局部最优等缺点,提出一种改进的混合免疫算法(improved hybrid immune algorithm,IHIA),将其与函数相结合,用于解决约束函数优化问题。采用个体字符串编码,通过信息熵法计算抗体亲和度,进而得到浓度。在混合免疫算法中提出克隆选择算子、可变阈值选择算子、郭涛精英变异和自适应变异概率算子等。研究结果表明:该算法提高了种群多样性和收敛性,减少了时间复杂度,提高了计算效率。

To improve the efficiency of basic artificial immune algorithm(AIA), an improved hybrid immune algorithmn (IHIA) was presented for constrained optimization. To maintain high diversity, the fitness of each individual and concerntration were taken into accounted to determine reproduction probability. Many benchmark functions were used to demonstrate the validity and the role of each design of IHIA. The results show that IHIA can reduce the complexity of computation, and increase the efficiency with the maintenance of d...