摘要
采用人工神经网络方法开发了首钢合金结构钢40CrA的力学性能预报模型,并采用Matlab和VC++两种软件平台建立了对应的力学性能预报系统。该预报系统投入现场在线应用后,通过预测值与实际抽样检验值的分析比较,表明该系统预报精度较高,屈服强度、伸长率等各项性能预报精度达到97%以上,能够满足生产检验要求。
The model was developed by using neural network method to predict the mechanical properties of alloy structure steel 40CrA. The visualized on-line prediction system was established by applying the software VC＋＋ and Mattab software. From the application in practice, the system has the very high prediction reliability by comparison of predict values and measured ones. The precision of the prediction properties such as the yield strength and elongation achieves above 97%, which can meet the production demand.
出处
《钢铁》
CAS
CSCD
北大核心
2013年第3期46-50,共5页
Iron and Steel
关键词
合金结构钢
40CrA
力学性能
预报模型
神经网络
alloy structure steel
40CrA
mechanical property
prediction model
neural network
