文献类型：期刊文献

中文题名：基于神经网络长期时效下GH4169合金动态拉伸行为研究

英文题名：Study on Dynamic Tensile Behavior of GH4169 Alloy at Long Term Aging Based on BP Neural Network

作者：刘洁[1];虞慧群[1]

机构：[1]华东理工大学,上海200237

年份：2013

卷号：42

期号：6

起止页码：182

中文期刊名：热加工工艺

外文期刊名：Hot Working Technology

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

语种：中文

中文关键词：BP神经网络;GH4169合金;长期时效;拉伸行为

外文关键词：BP neural network; GH4169 alloy; long term aging; tensile behavior

摘要：基于BP神经网络的方法,建立了长期时效下GH4169合金动态拉伸行为研究的神经网络模型。实验表明,在应变速率为101、102s-1时,GH4169合金的力学性能受时效时间影响显著,在应变速率为103s-1时,时效时间对其力学性能无明显影响。采用BP神经网络模型的预测结果与真实实验测试结果吻合度很高,绝对误差比大多为1%～2%,采用BP神经网络的方法是切实可行的。
Based on BP neural network method, the dynamic tensile behavior of GH4169 alloy at long term aging was studied. The experiment results show that under the strain rate of 101 s-l and 102 s-1, GH4169 alloy mechanical properties are significantly affected by aging time. And when the strain rate is 103 s -1, aging time effects the mechanical properties significantly. The predictive results and experimental results match very well and most of absolute errors are 1% to 2% between predictive results and experimental results by using BP neural network model. The method for BP neural network is reasonable and feasible.