摘要
造渣是转炉炼钢的主要任务,转炉炼钢的重点就是对造渣过程的控制,特别是造渣料加入量的确定。针对半钢炼钢[C]、[Si]较低的特点,在对转炉炼钢已有造渣料计算模型分析的基础上,结合钢厂实际的造渣经验,建立了由多个BP神经网络模型组成的级联模型,计算造渣料的加入量。使用某钢厂120 t转炉炼钢过程数据,通过与其他方法的离线试验比较,证明了所建模型的优越性。
Slag-making is one of the main tasks in converter steelmaking,and it are the key points of converter steelmaking to control the slag and determine the amount of added slag materials.For the characteristics of low [C], in the semi-steel steelmaking process,based on the existing slagging material model and the practice of slag-making,a new calculated model of slagging material was built and composed of many BP neural networks to analyze the amount of added slag material.Using 120 t converter′s process data in one steel plant,and by comparing with other methods,the superiority of the model was proved.
出处
《钢铁》
CAS
CSCD
北大核心
2010年第12期37-40,共4页
Iron and Steel
关键词
炼钢
半钢
造渣
BP神经网络
级联
steelmaking
semi-steel
slag-making
BP neural network
cascading
