摘要
根据收集的实验数据,建立了预测钢的奥氏体形成湿度(Ac1和Ac3点)的反向传播人工神经网络模型.用散点图和均方误差、相对均方误差和拟台分值二个统汁学指标评价模型的预测性能人工神经网络预测Ac3和Ac1的三个统计学指标分别为23.8℃,14.6℃;2.89%,2.06%和1.8921,1.7011.散点图和统计学指标均显示:人工神经网络的预测性能优于Andrews公式.此外,用人工神经网络分析了C和Mn的含量对Ac1和Ac3温度的定量影响,计算结果显示, C和Mn含量与Ac3和Ac1点间存在非线性关系,这主要是由于钢中合金元素间存在的相互作用造成的.
The back-propagation artificial neural network was established using data collected from domestic and foreign literatures to predict the austenite formation temperatures (A(c3) and A(c1)) of steels. Scatters diagrams and three statistical criteria - mean squared error, mean squared relative error and score of fitness were used to evaluate the prediction performance. The three criteria of predicting A(c3) and A(c1) using neural network are 23.8 degreesC, 14.6 degreesC; 2.89%, 2.06%; and 1.8921, 1.7011 respectively. Scatters diagrams and the statistical criteria showed that the prediction performance of artificial neural network is superior to that of Andrews formulae. Moreover, the quantitative effects of carbon and manganese contents on A(c3) and A(c1) temperatures were analysed using neural network models, the results showed that there exists nonlinear relationship between contents of C and Mn and the A(c3) and A(c1) temperatures which is mainly related to the interaction among the alloying elements in steels.
出处
《金属学报》
SCIE
EI
CAS
CSCD
北大核心
2004年第11期1133-1137,共5页
Acta Metallurgica Sinica
基金
国家高技术研究发展计划资助项目2003AA331040
关键词
钢的奥氏体形成温度
人工神经网络
预测性能
合金元素
定量影响
austenite formation temperatures A(c1)
A(c3)
artificial neural network
performance of prediction
alloying element
quantitative effect