摘要
随着优化方法在各领域的广泛应用,对优化技术的要求越来越高,为此,基于免疫系统抗体的自适应、自组织等学习机制,提出一种免疫遗传算法用于高维多模态优化问题求解.算法设计中亲和力随抗体群动态自适应学习,采取高斯突变及混合交叉等策略增强群体的多样性,提高算法的搜索和探测能力.将其与遗传算法及克隆选择算法通过数值实验比较,结果表明:所获的算法收敛速度快,勘测能力强,优越于其他比较的优化算法,具有较强的优化能力.
In recent year, optimization technologies are widely applied to various fields, and the challenge is exactly appear, for that, an immune genetic optimization algorithm, based on the principle of self-leaming and self-adaptation of immune system, is designed to solve the multi-model function optimization problem with high dimensions. The affinity is dynamically changed with the evolutionary population, gauss mutation and commix strategies are used to strength the diversity of antibody populations, and improve the ability of exploration and exploitation of algorithm. We compares the algorithm proposed with GA and CLONALG by 5 well known benchmark problems, the results illustrates that the algorithm proposed is effective, and is comparable.
出处
《吉林师范大学学报（自然科学版）》
2010年第3期46-50,共5页
Journal of Jilin Normal University:Natural Science Edition
基金
贵州省自然科学基金资助项目(20090074)
关键词
高维多模态优化
亲和力
免疫遗传算法
收敛性
Multi-model and High Dimensional Optimization
Affinity
Immune Genetic Algorithm
Convergence Property
