摘要
针对克隆选择算法自适应能力较弱的缺陷,给出了一种基于危险理论的自适应克隆选择算法。设计了危险信号操作算子,该算子将种群浓度的变动作为环境因素,以抗体—抗原亲和力为依据计算各个抗体在该环境因素下的危险信号,最终通过危险信号自适应地引导免疫克隆、变异和选择等后续免疫应答。实验结果表明本文算法具有较好的自适应能力和多值搜索能力。
To improve the self-adaptive capacity of clone selection algorithm, this paper proposed a novel self-adaptive clone selection algorithm based on danger theory. It designed the danger signal operation operator, saw the change of population concentration as environmental factors, regarded antibody affinity the basis of computing its danger level, and then guided the im- mune response by danger signal self-adaptive in the end. Simulation results show that the proposed algorithm has good perform- ance of self-adaptive ability and multimodal searching ability.
出处
《计算机应用研究》
CSCD
北大核心
2011年第1期332-334,共3页
Application Research of Computers
基金
湖南省教育厅科研资助项目(08D030
07D018)
关键词
人工免疫系统
危险理论
克隆选择算法
artificial immune system(AIS)
danger theory
clone selection algorithm