摘要
采用回归分析法和神经网络法分别建立了3000 mm中厚板轧机轧制力矩模型,将2种模型比较表明,用BP网络建立的模型计算精度好,预报的轧制力矩相对误差小于5%,可以满足生产需要。
Two models of rolling moment for 3 000 medium plate rolling mill were developed by regression analysis and BP Network, respectively. A comparison of the two models shows that the model of BP Network is better than another. The model of BP Network has better computation precision and smaller prediction error（the relative error of rolling moment is less than 5 %）, which can meet the requirements of medium plate rolling production.
出处
《钢铁研究》
CAS
2008年第3期17-20,共4页
Research on Iron and Steel
