Abstract: Immune Genetic Algorithm(IGA) accelerates searching though introducing the immune operator, but it reduces the populations’ variety. This paper proposes an improved algorithm with group division and hybridization, which divides the initial populations into groups and hybridizing local best individuals between groups to improve searching speed and stability. Simulation results show that the performance of the algorithm improves about 10% and the near global optimal solution can be easy and quick.

Key words: genetic algorithm, immune operator, group division, hybridization

摘要： 在免疫遗传算法中引入免疫算子可以提高算法的收敛速率，但也会降低种群个体多样性，不利于搜索。该文提出一种基于种群划分和杂交的免疫遗传算法，通过划分种群并对种群间的最优个体进行杂交来提高算法的速率和稳定性。实验表明，该算法在性能上可提高10%左右，收敛速度快、稳定性好、精确度高。

关键词: 遗传算法, 免疫算子, 种群划分, 杂交

CLC Number: 

• TP181