基于计算机辅助技术和BP神经网络的测振表注塑模具优化研究  

Research on Injection Mold Optimization of Vibration Meter Based on Computer Aided Technology and BP Neural Network

作　　者：吴群妹[1] 陈中标[1] 朱耀武[2] WU Qun-mei;CHEN Zhong-biao;ZHU Yao-wu(Wuxi Vocational College of Science and Technology,Wuxi 214028,China;Wuxi Institute of Technology,Wuxi 214000,China)

机构地区：[1]无锡科技职业学院,江苏无锡214028 [2]无锡职业技术学院,江苏无锡214000

出　　处：《塑料科技》2021年第11期76-79,共4页Plastics Science and Technology

基　　金：江苏省高等教育教改研究课题(2019JSJG454)。

摘　　要：以某型号手持测振表为实例,分析其结构特点,选择合适的材料,在Moldflow中建立注塑系统,并对其进行注塑翘曲变形分析。设计正交试验,将塑件的最大翘曲变形量作为优化指标,利用极差和方差分析法优化注塑工艺参数。基于正交试验数据建立BP神经网络预测模型,并对塑件的最大翘曲变形进行预测。结果表明:当模具表面温度为30℃,充填压力为96%,熔体温度为240℃,塑件的最大翘曲变形为0.969 mm,与初始参数相比降低27.09%。建立的BP神经网络预测模型误差范围在0.13%~2.67%之间,具有较高的准确性和可靠性,可应用于注塑成型参数优化,以提高生产效率。Taking a certain type of hand-held vibration meter as an example, the structural characteristics were analyzed,the appropriate materials were selected, the injection molding system was established in Moldflow, and the injection warpage deformation was analyzed. The orthogonal test was designed, and the maximum warpage deformation of the plastic part was used as the optimization index. The injection process parameters were optimized by range analysis and variance analysis. BP neural network prediction model was established based on orthogonal test data, and the maximum warpage deformation of plastic parts was predicted. The results show that when the mold surface temperature is 30 ℃,filling pressure is 96%, melt temperature is 240 ℃, the maximum warpage deformation of plastic parts is 0.969 mm,which is 27.09% lower than the initial parameters. The error range of the established BP neural network prediction model is 0.13%~2.67%, with high accuracy and reliability, which can be applied to the optimization of injection molding parameters to improve production efficiency.

关 键 词：BP神经网络 注塑成型 正交试验 

分 类 号：TQ325.14[化学工程—合成树脂塑料工业] TP391.7[自动化与计算机技术—计算机应用技术]