摘要
利用人工神经网络对Cu Ni Si Cr变形合金时效工艺参数与时效后性能样本集进行学习 ,建立了合金时效后的性能关于时效温度、时间和变形量等参数的非线性映射模型。模型采用改进的BP算法%DLevenberg Marquardt算法。结果表明 :所建立的神经网络模型具有较高的精度及良好的泛化能力 ,可对Cu Ni Si Cr合金时效后的性能进行有效的预测和分析。
A model of the non-linear relationship between properties and parameters of aging treatment such as aging temperature, time deformation degree was proposed based on artificial neural network for Cu-Ni-Si-Cr alloy. The improved model was developed by the Levenberg-Marquardt training algorithm. The results show that the model has high precision and good generalization performance, and can be successfully used to predict and analyze the properties of Cu-Ni-Si-Cr alloy.
出处
《材料热处理学报》
EI
CAS
CSCD
北大核心
2005年第1期86-89,共4页
Transactions of Materials and Heat Treatment
基金
国家"8 63"高技术研究发展计划资助 (2 0 0 2AA331 1 1 2 )
河南省杰出人才创新基金 (0 52 1 0 0 1 2 0 0 )
关键词
CU-NI-SI-CR合金
时效
神经网络
L-M算法
Aging of materials
Algorithms
Electric conductivity
Forecasting
Microhardness
Neural networks
