人工神经网络预测马氏体开始转变点Ms

Application of Artificaial Networks in the Prediction of Martensite Start Temperature

人工神经网络技术应用于马氏体开始转变点的研究。利用人工神经网络的反向传播 (BP)算法 ,建立了材料的合金成分与马氏体开始转变点的网络模型 ,从而对马氏体开始转变点温度进行分析和预测 ,并与经验公式相比 ,该网络模型具有更高的精度 ,从而为预测马氏体开始转变点温度提供了新颖的方法。

The newly developed neural network technique was employed to study martensite start temperature.The backpropagating algorithm in artificial neural networks is used to build a model for the relationship between the alloy element and martensite start temperature.The results show that the similar values of martensite start temperature M s predicted by neural network model and measured by experiment can be obtained.The neural network model provides a new method for the prediction of M s.