摘要
关于配送中心选址问题研究,采用免疫算法时单点交叉会产生超级个体以及固定概率会影响搜索能力的情况,分别以均匀交叉和自适应化的方法,在原算法上做出改进。并以实例验证该算法的可行性和有效性,与传统免疫算法比对,能很好避免超级个体的产生同时搜索能力也有增强。该算法自适应的变化更加符合个体在不同阶段演变情况,相比传统免疫算法,达到收敛速度快、鲁棒性高的效果,进而为选址问题研究在原有基础上匹配了一种更好的方法。
To solve selection of logistics distribution center,the standard immune algorithm(SIA)has the problem of"super individual"in the operation of single-point crossover and the poor search ability by using fixed probability of crossover and mutation.We try to apply uniform crossover and self-adaptation to SIA.Then,the feasibility and effectiveness of the adaptive immune algorithm are verified by a calculation example.Compared with SIA,AIA makes further efforts to avoid the generation of"super individual".Meanwhile,the search ability is enhanced in a certain extent.In my opinion,the adaptive changes are better to conform to the real state of individual in different stages.The AIA has fast convergence speed and high robustness.It provides a better approach on location problem in the original basis.
作者
倪卫红
岳晓伟
邵建峰
钱伟民
NI Wei-hong;YUE Xiao-wei;SHAO Jian-feng;QIAN Wei-min(School of Economics and Management,Nanjing Tech University,Nanjing 210009,China;Huzhou Xinkaiyuan Crushed Stones Co.,Ltd.,Huzhou 313022,China)
出处
《价值工程》
2018年第36期96-99,共4页
Value Engineering
基金
国家自然科学基金"人境交互视角下的员工企业社交网络知识互动行为及其驱动机理研究"(71701093)
国家自然科学基金"框架协议采购模式及库存决策研究"(G71701092)
南京工业大学哲学社会科学科研创新团队"中国制造2025战略背景下的质量与服务管理"
关键词
选址
优化
自适应
均匀交叉
免疫算法
location
optimization
adaptation
uniform crossover
immune algorithm
