张卫正, 张梦华, 张伟伟,等.基于双种群的混合免疫动态优化算法[J].河南理工大学学报（自然科学版）,2020,39(1):116-124.

ZHANG W Z, ZHANG M H, ZHANG W W, et al.Hybrid immune dynamic optimization algorithm based on dual population[J].Journal of Henan Polytechnic University(Natural Science) ,2020,39(1):116-124.

基于双种群的混合免疫动态优化算法

张卫正, 张梦华, 张伟伟, 金保华, 吴怀广, 王华, 李国庆

郑州轻工业学院 计算机与通信工程学院，河南 郑州450000

摘要:针对传统的群智能算法在求解动态优化问题时出现的早熟收敛和多样性缺失等问题，提出一种基于双种群的混合免疫动态优化算法BPAIS。首先，受生物免疫系统中固有免疫反应和自适应免疫反应的启发，将初始种群根据适应度值分为2个种群——固有种群和自适应种群；其次，对固有种群进行固有免疫反应操作，在进行全局性搜索的同时保持种群的多样性，而对自适应种群进行自适应免疫反应操作，采用差分进化算法加强局部搜索能力，通过引入记忆跟踪机制在环境变化时跟踪局部最优解；最后，结合双种群的免疫反应和记忆跟踪机制，提出基于双种群的混合免疫动态优化算法，并在简单测试用例产生器（simple test-case generator， STCG）和移动峰测试函数上进行仿真实验。结果表明，BPAIS具有很好的动态优化能力，能够有效地跟踪和定位全局最优解，与其他算法相比，具有很强的竞争力。

关键词:双种群;固有免疫反应;自适应免疫反应;混合免疫动态优化算法;克隆选择

doi:10.16186/j.cnki.1673-9787.2020.1.15

基金项目:国家自然科学基金资助项目（61403349，61501405 ，41601418 ）；河南省科技攻关项目（182102110399 ）；河南省高等学校重 点科研项目（18A210025 ）；河南省科学技术研究重点项目（14B520066，15A520033 ）；郑州轻工业学院博士基金资助项目 （2013BSJJ044）；郑州轻工业学院研究生科技创新基金资助项目；郑州轻工业学院大学生科技活动项目

收稿日期:2019/04/13

修回日期:2019/05/28

出版日期:2020/01/15

Hybrid immune dynamic optimization algorithm based on dual population

ZHANG Weizheng, ZHANG Menghua, ZHANG Weiwei, JIN Baohua, WU Huaiguang, WANG Hua, LI Guoqing

School of Computer and Communication Engineering， Zhengzhou University of Light Industry， Zhengzhou 450000 ， Henan， China

Abstract:Aiming at the problems of premature convergence and lack of diversity encountered by traditional swarm intelligence algorithms in solving dynamic optimization problems， an immune dynamic optimization algorithm （BPAIS） based on dual population was proposed. Inspired by the innate immune response and adaptive immune response in the biological immune system， the initial population was separated into two populations based on fitness values ： innate population and adaptive population. Then， innate immune response was performed on innate populations for global search and diversity maintenance， while adaptive immune response was applied on adaptive population for local search enhancement by using differential evolution method. In addition ，the memory tracking mechanism was introduced to track the local optimal solution when the environment changed. Finally， based on the immune response of dual populations and memory tracking mechanism， a dual population immune dynamic optimization algorithm was proposed. The simulation experiments were performed on the simple test-case generator and moving peaks benchmarks （ MPB ） . The results showed that BPAIS had superior dynamic optimization ability and could effectively track and locate the global optimal solution. Compared with the other algorithms， it was very competitive.

Key words:dual population;innate immune response;adaptive immune response;hybrid immune dynamic optimization;

基于双种群的混合免疫动态优化算法_张卫正.pdf