【摘要】 根据冶金机理和钢液热平衡原理,综合考虑电极供电、合金、渣料添加情况及钢包状况等因素,并结合人工神经网络算法和自学习方法,建立了LF精炼过程控制模型,包括温度预报模型和合金模型。经测试,温度预报模型在线计算值与实测值的绝对偏差在±5℃以内的比例达到了85%以上,合金模型应用后,LF钢水成分合格率从90.1%提升至94.7%,平均合金化时间由8.9 min降低到6.7 min。利用Microsoft Visual.Basic.net程序设计软件以及Microsoft SQL Server 2000数据库软件,开发了LF精炼过程控制模型软件。更多还原

【Abstract】 Based on the metallurgical mechanism and the thermal balance theory of molten steel the control model for steel-refining process in ladle furnace was established after considering these factors such as power-supplying by electrode, addition of alloy and slag charge as well as the condition of the ladle together with using the artificial neural network algorithm and self-learning algorithm. The said control model consists of the temperature forecast model and the alloy model.By testing, the absolute deviation in the range from minus 5 ℃ to plus 5 ℃ between the calculation values by using the temperature forecast model and actual measured values makes up more than 85% while the percent of pass of compositions in molten steel improves to 94.7% from 90.1%and the average alloying time is shortened to 6.7 minutes from 8.9 minutes when the alloy model is put into operation finally the control model software for steel-refining process in ladle furnace was developed based on Microsoft Visual.Basic.net program designing software and Microsoft SQL Server 2000 database software.更多还原