摘要
以某三相铜牌为研究对象,基于正交试验和BP神经网络模型对铜片嵌入面平面度进行预测和分析。采用Moldflow浇口定位器,确定最佳的浇口位置并建立流道系统。基于初始工艺参数进行模流分析,发现三个位置的平面度均不满足要求。通过正交试验的极差与方差分析,得到各工艺参数对平均平面度的影响程度。建立BP神经网络,基于正交试验数据进行训练。结果表明:基于BP神经网络预测和模流分析,得到三个铜片嵌入面的平面度Ⅰ、Ⅱ和Ⅲ分别为1.402、1.488及1.571 mm,相比初始工艺分别降低25.3%、22.6%和27.1%,优化效果明显,且满足设计指标要求。试模样品外观状态良好,平面度满足要求,优化工艺可用于实际生产。
Based on the orthogonal test and BP neural network model,the flatness of copper embedded surface of a threephase copper connector was forecasted and analyzed.The best gate position was determined,and the runner system was established by using Moldflow gate locator.Based on the initial process parameters,the flatness of the three positions did not meet the requirements.The influence degree of each process parameter on the average flatness was obtained by the range and variance analysis of orthogonal test.BP neural network was established and trained based on orthogonal test data.The results show that based on the BP neural network prediction model and Moldflow analysis,the calculated flatness Ⅰ,Ⅱ and Ⅲ of the three copper embedded surfaces are 1.402,1.488 and 1.571 mm respectively,which are reduced by 25.3%,22.6% and 27.1% respectively compared with the initial process.The optimization effect is obvious,and all meet the design index requirements.The appearance of the product is in good condition,and the flatness meets the requirements.The optimized parameters can be used in actual production.
作者
程亚维
苏文芝
王东霞
CHENG Ya-wei;SU Wen-zhi;WANG Dong-xia(Jiyuan Vocational and Technical College,Jiyuan 459000,China)
出处
《塑料科技》
CAS
北大核心
2021年第9期70-74,共5页
Plastics Science and Technology
关键词
三相铜牌
玻纤增强PBT
正交试验
BP神经网络
Three-phase copper connector
Glass fiber reinforced PBT
Orthogonal test
BP neural network
