利用神经网络预测连铸保护渣性能

Properties prediction of casting powder with neural network

在实验的基础上建立基于神经网络的连铸保护渣性能预测模型,经对比分析表明,用神经网络对连铸保护渣性能进行训练和学习后,神经网络方法能有效和比较准确地对保护渣性能进行预测。

The neural network prediction model of continuous casting powder was established on the basis of experiment. The results show that the neural network, after studied and trained, can predict the property of CC powder effectively and accurately.