摘要
针对交互式遗传算法中小种群规模导致的算法后期收敛缓慢问题,以及多个用户对同一问题领域进行优化的可能性,本文结合文化算法的双层进化结构,提出了一种基于知识迁移的多用户交互式遗传算法模型。基于服装设计系统对上述模型及相关策略进行了实例分析,实例比较表明该模型能有效提高各用户的进化收敛速度,减轻用户疲劳。
Aiming at the problem in the interactive genetic algorithm （IGA） that the small size of population may result in slow convergence rate, and the possibility of many users taking part in the evolution of the same problem, a knowledgetransform based IGA for multi-user is proposed in combination with a two-layer structure of cultural algorithm. With the application in garment industry, the above model and the relative strategies are analysed. Examples show that this model can improve the convergence speed of each population, and alleviate the user＇s fatigue.
出处
《自动化技术与应用》
2007年第10期14-16,13,共4页
Techniques of Automation and Applications
关键词
知识
知识迁移
交互式遗传算法
文化算法
knowledge
knowledge transform
interactive genetic algorithm
cultural algorithm
