免疫克隆选择算法的改进及其应用  

Application and Improvement of Immune Clonally Selection Algorithm

作　　者：邱亚龙 张昕[1] 范妙炳 叶奕纯 陈婷[1] 

机构地区：[1]华南农业大学数学与信息学院,广州510642

出　　处：《湖南理工学院学报（自然科学版）》2016年第1期27-32,74,共7页Journal of Hunan Institute of Science and Technology(Natural Sciences)

基　　金：广东省大学生创新创业项目(201410564201)

摘　　要：基于生物免疫系统原理,提出了改进的免疫克隆选择算法.引入网割预处理,获得进化终止次数并使初始抗体集的多样性得到控制;采取震荡变异法提高算法的局部搜索精度;引入记忆机制,提高二次免疫收敛速度.本文应用此算法针对大气污染损害率普适公式R_i=1/(1+ ae^(-bx_i))~c 进行参数优化,结果显示改进算法在全局与局部范围内搜索更为细腻,提高了求解精度.Based on the principle of biological immune system, an improved immune clonally selection algorithm（ICSA） was proposed. The algorithm introduced the analysis of antigenic determinant, calculated the network cut factor of antibody space and the end times of antibodies evolution, and created environment required to produce antibodies; shock variation method was adopted to make antibodies mutated; Innovative space adaptive mutation was proposed creatively; The improved ICSA was applied to analyze the parameters optimization problem of the atmospheric pollution harm rate universal formula Ri = 1/（1 ＋ ae^-bxi ）^c ; The results show that the algorithm within the scope of the global and local search is more exquisite. Solution accuracy is significantly increased.

关 键 词：免疫克隆选择算法 抗原预处理 震荡变异 大气污染损害率普适公式 参数优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]