人工神经网络在注塑参数预测中的应用  

Application of Artificial Neural Network in Prediction of Injection Parameters

作　　者：郑生荣[1] 辛勇[1] 杨国泰[1] 何成宏[1] 

机构地区：[1]南昌大学机电工程学院,江西南昌330029

出　　处：《塑料工业》2003年第10期26-28,33,共4页China Plastics Industry

基　　金：教育部科技研究重点项目;江西省科委科技项目资助 (Z1891)

摘　　要：应用人工神经网络技术建立了注塑工艺参数值快速预测的双层BP网络模型。利用CAE软件进行模拟 ,获取训练样本。用Matlab语言编制了应用程序 ,对网络的训练及参数预测的过程进行求解。将网络预测结果与模拟所得验证数据进行比较和误差分析 ,显示出BP网络的稳定性和可靠性。经实例验证 。In this paper a two layer BP network model for the rapid prediction of injection parametric values is established by using the artificial neural network The training data were acquired through the CAE simulation,and the application program is written in Matlab engineering computing language The comparison and error analysis were carried out between the networks predictive results and the testing data from CAE simulation,showing the stability and reliability of BP network in the prediction This expresses the fact that it is feasible for the artificial neural network to be used in fast prediction of the injection parametric values

关 键 词：人工神经网络 预测 应用 MATLAB 注塑工艺参数 双层BP网络模型 

分 类 号：TQ320.6[化学工程—合成树脂塑料工业]