摘要
LF精炼炉炼钢是先进炼钢工艺流程中的一个环节,精炼期的终点预报是LF精炼炉优化控制的组成部分。提出了一种改进BP算法的LF炉终点温度预估模型,通过建立钢液终点温度与各影响因素之间的联系,对炼钢终点温度进行预报。仿真结果表明该神经网络有较强的自学习能力和收敛特性,预报结果达到一定精度,完全能够满足工业要求,对指导工业生产具有重要的现实意义。
LF refining furnace is an significant step of advanced steelmaking process, in the period of refining, terminal forecast is an important part of LF furnace optimal control. This paper proposes an terminal temperature forecast model of LF furnace based on improved BP algorithm, which can forecast terminal temperature through establishing an connection between terminal temperature of molten steel and multi influencing factors. The simulation results show that the neural network has strong ability of self-learning and convergence characteristics, the prediction results may reach to a certain precision and can satisfy the industrial requirements completely, this has important practical significance to guide industrial production.
出处
《铸造技术》
CAS
北大核心
2013年第8期1092-1095,共4页
Foundry Technology
基金
陕西省教育厅科研计划资助项目(2010KJ664)
2012年西安市产业技术创新计划项目(CX12181②)
关键词
LF炉
BP算法
终点温度
预报模型
LF furnace
BP algorithm
terminal temperature
forecast model