摘要
提出3个神经网络分支组合而成的POE交换机参数优化模型,捕获多样化、非线性的翘曲特征信息。经仿真获得较优的POE交换机外壳注塑工艺参数:模具温度为63℃,熔体温度为254℃,注射时间为8 s。采用Pro/E计算机辅助工具开展模流分析,分析结果表明:当注射位置为A时,POE交换机外壳的各项指标较优,不易出现翘曲、变形等缺陷。
A POE switch parameter optimization model composed of three neural network branches is proposed to capture diversified and non-linear warpage characteristic information.The optimal injection molding process parameters of the POE switch shell are obtained through simulation:the mold temperature is 63℃,the melt temperature is 254℃,and the injection time is 8 s.Pro/E computer-aided tools are used to carry out mold flow analysis.The analysis results show that when the injection position is A,the various indicators of the POE switch shell are better,and defects such as warpage and deformation are not easy to appear.
作者
李付
LI Fu(Zhengzhou Preschool Education College,Zhengzhou 450000,China)
出处
《塑料科技》
CAS
北大核心
2020年第11期76-79,共4页
Plastics Science and Technology