Ping LI
,
Qing ZHANG
,
Kemin XUE
材料科学技术（英文）
Grain shape of the hot deforming alloy is an important index to character the microstructure and performance of material. The fractal theory was applied to analyze the recrystallized microstructure of Ti-15-3 alloy after hot deformation and solution treatment. The fractal dimensions of recrystallized grains were calculated by slit island method. The influence of processing parameters on fractal dimension and grain size was studied. It has been shown that the shapes of recrystallized grain boundaries are self-similar, and the fractal dimension varies from 1 to 2. With increasing deformation degree and strain rate or decreasing deformation temperature, the fractal dimension of grain boundaries increased and the grain size decreased. So the fractal dimension could characterize the grain shape and size. A neural network model was trained to predict the fractal dimension of recrystallized microstructure and the result is in excellent agreement with the experimental data.
关键词:
Ti-15-3 alloy
,
null
,
null
,
null
,
null
杨榛
,
乔文明
,
梁晓怿
新型炭材料
doi:10.1016/S1872-5805(17)60108-2
如何控制和预测孔结构是炭气凝胶研究的重要课题.然而,由于耗时耗财,导致实验方法研究控制和预测孔结构成为难题.本文提出一种基于神经网络的炭气凝胶孔结构的预测与优化模型,并采用遗传算法设计和优化模型,对六种典型训练算法模型性能进行比较分析.利用该模型对孔径和吸附容量进行预测,两者的预测相关系数分别为0.992和0.981,预测均方根误差分别为0.077和0.054.经测试,该模型与实验研究的结果相符,并有效的应用于预测和控制炭气凝胶实验参数.
关键词:
炭气凝胶
,
孔结构
,
神经网络
,
训练算法
,
模型
