论文部分内容阅读
文章分别用多元线性回归方法和神经元网络方法对铁水的硅含量进行预测。在此基础上,又将两种方法有机结合起来运用,收到良好的效果,为解决高炉生产中的一个关键问题提供了有效的方法。
The article uses multiple linear regression method and neural network method to predict the silicon content of hot metal. On this basis, the two methods are combined organically and received good results, which provides an effective method for solving a key problem in blast furnace production.