摘要： 以蜂窝纸板的静态压缩实验数据为训练样本，建立了蜂窝纸板静态缓冲特性的神经网络模型，并对网络训练算法进行了一些改进，仿真结果表明蜂窝纸板的神经网络特性模型较好地模拟了实际的蜂窝纸板缓冲特性，训练误差均控制在5％以内，可满足工程设计的要求。

关键词: 神经网络, 蜂窝纸板, 缓冲特性, 建模

Abstract: A model of artificial neural networks of honeycombed paperboard static cushion properties is set up in this paper. The training arithmetic for the networks is also ameliorated. The calculation results show that model of artificial neural networks of honeycombed paperboard simulate actual honeycombed paperboard cushion properties well. The training errors are controlled under 5%, so that the model could be used in engineering.

中图分类号: 

• TP183