摘要
利用神经网络对Cu Cr Zr合金时效温度、时间与硬度和电导率样本集进行学习 ,采用改进的BP网络算法———Levenberg Marquardt算法 ,建立了时效强化工艺BP神经网络模型。预测结果表明 :该BP神经网络可以充分挖掘样本蕴含的领域知识 ,可以对材料性能进行有效预测和分析。
The BP neural network model(ANN) for age hardening process was established by introducing modified Levenberg-Marqurdt algorithm based on the learning of sample collection of the ANN to model the non-linear relation of hardness and electric conductivity of the Cu-Cr-Zr alloy as a function of ageing temperature and time. The results showed that the BP neural network could fully explore the domain knowledge in the sample collection to effectively predict and analyze the properties of the Cu-Cr-Zr alloy.
出处
《特种铸造及有色合金》
CAS
CSCD
北大核心
2003年第5期29-31,共3页
Special Casting & Nonferrous Alloys
基金
国家"8 63"高技术研究发展计划资助项目 (2 0 0 2AA331 1 1 2 )
河南省重大科技攻关项目 (0 1 2 2 0 2 1 30 0 )
