基于BP-TGA算法的注塑成型工艺参数优化  

Optimization on Injection Process Parameters Base on BP-TGA Algorithm

作　　者：曹素兵[1] 朱婵[2] 

机构地区：[1]四川工程职业技术学院,四川德阳618000 [2]四川建筑职业技术学院,四川德阳618000

出　　处：《塑料科技》2018年第3期102-106,共5页Plastics Science and Technology

基　　金：德阳市科技支撑支助项目(2014zx070-6)

摘　　要：以鼠标壳为实例,利用正交试验数据作为训练样本,以翘曲变形量为目标,经反复训练找出最佳BP网络模型作为注塑工艺参数组合优劣的评价系统。在BP网络模型的基础上,融入禁忌遗传算法作为优化算法（BPTGA算法）对注塑工艺参数组合进行优化,从而找出翘曲变形量最低的参数组合。仿真实验表明,通过BPTGA算法可高效、准确地在指定范围内找出最佳工艺参数组合,从而大大提高产品设计效率和制品质量。Taking the mouse shell as an example, using orthogonal test data as training samples and taking warpage deformation as the goal, the best BP network model was found out after repeated training as an evaluation system for the combination of quality parameters of injection molding. Base on the BP network model and the tabu genetic algorithm（BP-TGA algorithm） as optimization algorithm to optimize the injection molding process parameters combination and find out the lowest warping deformation parameter combination. Simulation experiments show that BP-TGA algorithm can efficiently and accurately identify the best process parameter combinations in a specified range, thus greatly improving product design efficiency and product quality.

关 键 词：注塑工艺 正交试验 BP神经网络 BP-TGA算法 

分 类 号：TQ320.662[化学工程—合成树脂塑料工业]