注塑过程熔体充填长度的动态神经网络软测量模型

Modeling of Melt-flow-length During Injection Filling Stage Based on Dynamic Neural Network Soft-sensor

提出用喷嘴压力、注射速度和螺杆位移等在线可测二次变量来预测充填长度的方法 ,并建立了基于动态神经网络的充填长度模型。验证结果表明 。

Quality control is of great importance of the injection molding study.As a key variable to the injection molded part quality,melt front rate in mold cavity is commonly believed that it should be control at a constant value during the tilling stage as to improve the part uniformity.Modeling of melt flow length,therefore,is of great importance for the eventual implementation of the quality control.A model of melt flow length using the other online measurable variables based on dynamic neural network is founded in the paper.Verification results using molds with different shapes prove that the neural network model can accurately predict the melt folw length.