摘要
针对不规则零件排样效率低的问题,提出了基于知识进化与自然进化的优化排样算法。该算法利用知识规则和适应度函数相结合的选择方法,既克服了传统"轮盘赌"等选择方法随机性强的缺点,又可以大大提高零件的排样效率、保证子代群体的高质量。实验结果表明,与自然进化的遗传算法相比,该算法不但提高板材的利用率,而且时间复杂度明显降低。
Aiming at the problem of low efficiency of irregular parts layout, based on knowledge evolution and the evolution of optimization layout algorithm is proposed. The selection method is used combining knowledge rules and fitness function, the strong randomness shortcomings of traditional ＂roulette wheel＂ selection method are overcome, and the layout efficiency is greatly improved, the offspring of the components group of high quality is guaranteed. Experiments show that compared with the natural evolution genetic algorithm, this algorithm not only improve the utilization rate of plates, and the time complexity is decreased obviously.
出处
《计算机工程与设计》
CSCD
北大核心
2014年第2期651-656,共6页
Computer Engineering and Design
基金
广州市科技计划基金项目(12C22111580)
关键词
遗传算法
选择方法
知识进化
自然进化
适应度函数
genetic algorithm
selection methods
knowledge evolution
natural evolution
fitness function
