摘要
注塑成型中,工艺参数直接影响到模具内熔体的流动状态和最终制品的质量,而工艺参数与制品质量之间的关系非常复杂,因此如何建立制品质量与工艺参数之间的关系模型并获得优化的工艺参数是改善制品质量的关键。收缩是衡量制品质量的一个重要指标,制品在型腔中的非均匀收缩是引起制品翘曲的主要原因。文中基于成型过程的数值模拟,采用人工神经网络与混合遗传算法结合优化注塑成型工艺,以改善制品质量。对一工业产品进行分析,以制品内的体收缩率差值为质量指标优化工艺,改善了制品内的体收缩率分布,获得了满意效果。
Process parameters have direct influence on the flow behavior of melt and final part quality for injection molding, and the relationship between part quality and process conditions is very complex. So how to build the relationship model and get optimum process conditions is the key to improving part quality. Shrinkage is one of several important indexes determining the quality of injection molded parts, and non-uniform volumetric shrinkage in the cavity is an indication of potential warpage. A combination method of artificial neural network and hybrid genetic algorithm is proposed in this paper to optimize the process parameters and improve part quality. This ANN-HGA method is applied in an industrial product, and the volumetric shrinkage variation in the part is selected as quality index to optimize the process parameters. The results show that volumetric shrinkage variation in the part is minimized under the optimum process and the results are satisfied.
出处
《高分子材料科学与工程》
EI
CAS
CSCD
北大核心
2005年第5期23-27,共5页
Polymer Materials Science & Engineering
基金
国家"十五"863计划(2002AA336120)
郑州大学骨干教师资助项目
关键词
注射成型
工艺优化
神经网络
遗传算法
injection molding
process optimization
neural network
genetic algorithm
