改进混合遗传算法在自动配棉上的应用  

Application of improved hybrid genetic algorithm on automatic cotton-blending

作　　者：李成国[1] 范秀娟[1] 

机构地区：[1]北京服装学院信息工程学院,北京100029

出　　处：《纺织学报》2009年第3期28-33,共6页Journal of Textile Research

基　　金：北京市教育委员会科技发展计划面上计划资助项目(KM200610012004)

摘　　要：为解决多约束条件下配棉难的问题,通过研究原棉性能与纱线质量之间的关系以及分析基本遗传算法在解决该问题时的优缺点,设计了新的计算机自动配棉数学模型,并运用群体排序和局部寻优等关键技术,提出一种改进的混合遗传算法。分别运用基本遗传算法和改进的混合遗传算法对自动配棉模型进行实验。结果表明:改进的混合遗传算法给出的配棉方案比较合理,混合棉质量指标精度较高;由于增加了局部寻优算子和最速下降因子,算法的寻优能力和收敛速度得到了加强。Attempting to solve the problems of computer automatic cotton-blending under multi-constrained conditions, the relationship between raw cotton and yarn quality was studied. The advantages and weaknesses of the basic genetic algorithm in solving this problem was analyzed, then a new computer automatic cotton- blending model was designed. It proposes an improved hybrid genetic algorithm by using key techniques such as group ordering and local optimalizing methods. And computer automatic cotton-blending experiments are carried out by using the basic genetic algorithms and the improved hybrid genetic algorithm respectively. The results show that the cotton assorting scheme obtained by the improved hybrid genetic algorithm is more reasonable and the resulting mixed cotton exhibits better quality. The general optimization and convergence speed of the algorithm has been strengthened by adding the local optimization operator and the steepest descent factor at the same time.

关 键 词：混合遗传算法 自动配棉 多约束条件 组合优化 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]