多种群免疫协同进化算法在物流配送中心选址中的研究  

Multiple Population Immune Co-evolution Algorithm and Application in Location of Logistics Distribution Center

作　　者：吕红[1] 朱作付[1] 

机构地区：[1]徐州工业职业技术学院信息管理技术学院,江苏省徐州市221000

出　　处：《电脑与信息技术》2015年第3期8-11,共4页Computer and Information Technology

基　　金：江苏省教育厅高校科研成果产业化推进项目(项目编号:2011-47);徐州工业职业技术学院科学技术项目(项目编号:XGY201310)

摘　　要：为提高免疫算法在物流配送中心选址问题的效率,文章引入了多种群协同进化的框架模型,在此模型上提出了一种多种群免疫协同进化算法(Multiple Population Immune Co-evolution Algorithm,MPICA)。MPICA通过对若干个抗体子群进行多样性评价,获得各自的记忆库和父代种群;记忆库之间通过移民算子进行联系,增大优秀抗体亲和度成熟的概率;各父代种群通过期望繁殖率进行选择、动态交叉和变异来提高抗体分布的多样性。针对物流配送中心选址实验数据表明,MPICA具有可靠地收敛性和全局寻优能力,能够高效的进行物流配送中心的选址。In order to improve the efficiency of immune algorithm in location of logistics distribution center, this paper introduces the framework model of＂ multiple population co-evolution,and proposes a multiple population immune co-evolutionary algorithm （Multiple Population Immune Co-evolution Algorithm, MPICA）.MPICA obtains memories and parent populations by diversity evaluation of several antibodies subgroup,and memories are linked by the imrmgration operator to increase the probability of excellent antibody maturation and affinity,and the parent populations are improved the diversity of antibody distribution by expectation reproduction selection and crossover and mutation.According to the experiments of location of logistics distribution center, MPICA has reliable convergence and global optimization ability.

关 键 词：多种群 免疫算法 移民算子 物流配送中心 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]