基于综合关联度的注射成型工艺参数多目标优化研究  

Optimization Method for Parameters of Plastics Injection Molding Based on Synthesis Relation Degree

作　　者：黄风立[1] 许锦泓[1] 顾金梅[1] 娄勇坚[1] 

机构地区：[1]嘉兴学院机电工程学院,浙江嘉兴314001

出　　处：《中国塑料》2009年第12期60-64,共5页China Plastics

摘　　要：针对注射成型过程中的两个品质指标(翘曲量和收缩率),提出了基于综合关联度、Kriging模型及自适应遗传-蚁群算法的注射成型工艺参数优化方法。首先给出了综合关联度的定义及计算方法,然后给出了Kriging模型的拟合方法和自适应遗传-蚁群的优化算法。在结合带凸沿杯子的注射成型具体实例的研究中,首先给出了注射成型工艺参数的范围,其次给出了基于正交实验的设计变量确定方法,然后基于超拉丁立方实验及Kriging方法进行模型近似,最后利用自适应的遗传-蚁群算法进行优化求解。模流分析及实际注塑实验表明,注射成型工艺参数优化的结果可靠。In order to control two factors in plastics injection molding, warpage and shrinkage, an optimization method for the parameters of injection molding was proposed based on correlative degree, Kriging model, and adaptive genetic-ant algorithm. The definition and calculation method of correlative degree were defined first, and the approximate model based on Kriging model and genetic ant algorithm were introduced. In the case of plastics injection molding of fruit plates, the range of the technical parameters was first determined through orthogonal experiments, and an approximate model was formulated based on Hyper-Latin square experiment and Kriging method, which was optimized via genetic ant algorithm. The obtained parameters could well be applied to practical injection molding.

关 键 词：注射成型 工艺参数 综合关联度 KRIGING模型 遗传-蚁群混合算法 

分 类 号：TQ320.662[化学工程—合成树脂塑料工业]