基于疫苗自动获取与更新的免疫遗传算法

收敛速度缓慢已成为遗传算法研究中亟待解决的主要问题之一．为了提高遗传算法的收敛速度，提出了一种基于疫苗自动获取与更新的免疫遗传算法．从各代种群中选出优良个体，然后从这些优良个体中提取免疫疫苗，概率地对后代种群的个体接种疫苗．接种疫苗是利用疫苗确定位上的等位基因替代个体相应位上等位基因的操作．接种疫苗加速了优良模式的繁殖，修复了被交叉、变异破坏的优良模式．种群与疫苗库相互作用、协同进化，极大地提高了算法的收敛速度．基于模式定理分析了算法的计算效率．最后，几个典型函数优化问题的仿真结果表明了算法的可行性和有效性．

An Immune Genetic Algorithm Based on Vaccine Autonomous Obtaining and Updating

Slow convergence to the global optimum has been one of the main problems in genetic algorithm In order to increase the speed of convergence, an immune genetic algorithm based on vaccine autonomous obtaining and updating (IGAVAOU) is proposed Excellent individuals are selected from each generation population and vaccine is obtained from these excellent individuals Then individuals in succeeding population are vaccinated in stochastic way Vaccination is a kind of operation by which allele in vaccine replace allele on individual corresponding locus Vaccination can not only make excellent schemata proliferate, but also repair the schemata destroyed by crossover and mutation operations Population and vaccine repertory influence each other and co-evolve so that they accelerate convergence to the global optimum IGAVAOU's computation efficiency is analyzed based on the schemata theorem IGAVAOU is verified by several typical functions The results show the feasibility and validity of the algorithm