摘要
利用神经元网络建立了转炉终点碳温预报模型。采用相关性分析和分类的方法 ,研究了转炉终点碳温预报模型的主要参数 ,并对神经元网络模型的结构、训练方法、训练误差和测试误差进行了研究和比较。
A model to predict the temperature and carbon content at the end point of LD steelmaking is developed using neutral network. The main parameters of the model are studied by means of correlation analysis and classifying method. In addition,the structure,training methods, training errors and testing errors of the neutral network model are analyzed in the paper.
出处
《钢铁研究》
CAS
2004年第6期43-46,52,共5页
Research on Iron and Steel
关键词
静态模型
碳温预报
神经元网络
static model
carbon content and temperature prediction
neutral network
