| 基于Elman神经网络的EPDM动态粘弹性能预测 |
| Viscoelastic Performance Prediction of EPDM Based on Elman Neural Network |
| 投稿时间：2017-09-20 修订日期：2017-09-20 |
| DOI： |
| 中文关键词: Elman神经网络 EPDM 动态粘弹性能 |
| 英文关键词: Elman neural network viscoelastic properties EPDM performance prediction. |
| 基金项目:山东省自然科学基金资助项目(ZR2014EMM018) |
|
| 摘要点击次数: 827 |
| 全文下载次数: 469 |
| 中文摘要: |
| 本文通过实验研究了EPDM配方因素与EPDM粘弹性能之间的关系,并建立了Elman神经网络预测模型,用于预测胶料在85℃,15%应变下的动态粘弹性能(储能模量、损耗模量、损耗因子)。通过正交试验设计取得16组试验数据,并利用其中的1-14组数据训练Elman神经网络,利用剩余的15-16组数据来检测Elman神经网络的预测能力,另外设计4组实验数据来检测Elman神经网络的预测能力。结果表明：建立的Elman神经网络的对胶料粘弹性能的预测误差在4%以内,模型能够准确的预测EPDM胶料的粘弹性能。 |
| 英文摘要: |
| In this paper, the relationship between the EPDM formula and the viscoelastic properties of EPDM was studied experimentally. The Elman neural network prediction model was established to predict the dynamic viscoelastic properties (Storage modulus, loss modulus, loss factor) of the compound at 85 ℃ and 15% strain. 16 sets of experimental data were obtained by orthogonal experiment design, and Elman neural network was trained by 1-14 data. The remaining 15-16 data were used to detect the prediction ability of Elman neural network. Four sets of experimental data were designed to detect Elman neural network prediction ability. The results show that the elastic error of the Elman neural network is less than 4%, and the model can accurately predict the viscoelastic properties of the EPDM compound. |
|
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
