吹氩站钢水终点温度预报模型及其仿真应用
来源期刊:冶金自动化2009年第5期
论文作者:芦永明 田乃媛 徐安军 贺东风
文章页码:18 - 21
关键词:吹氩; 钢水温度; 多元回归; 预报模型; 仿真应用
Key words:argon blowing; molten steel temperature; multivariate regression; prediction model; simulation application
摘 要:根据对210 t钢包吹氩站热状态实验数据和实际生产工艺数据的统计分析可知,吹氩站钢水终点温度的主要影响因素是钢水到站温度和总调温废钢用量以及吹氩精炼时间。本文采用正交设计方法通过对现场实际生产数据进行筛选,选取最具有代表性的数据,进而利用SPSS软件,采用多元回归的方法建立了210 t钢包钢水吹氩站终点温度预报模型。之后通过随机抽取的100组实际生产数据对所建模型进行验证,验证结果表明模型在对吹氩站钢水终点温度预报时,预报误差在±10℃内的正确率达到90%以上,说明此模型具有较好的应用效果。最后,建立吹氩站钢水终点温度预报仿真系统,对模型进行了仿真应用。
Abstract: Based on statistical analysis of hot state test data and practical production process data of 210 t ladle in argon blowing station,main influence factors of temperature drop were obtained and they were inlet temperature,amount of scrap for temperature adjusting,and argon blowing time.Orthogonal design method was adopted and the most representative data was selected from practical production data.The prediction model of molten steel end point temperature for 210 t ladle in argon blowing station was established by multivariate regression method through SPSS software.The model was verified through random selected 100 group data.Verification results show that the model has a good prediction effect,deviation of prediction was controlled within the range of ±10 ℃,and hitting rate comes to 90% in prediction for molten steel end point temperature in argon blowing station.Finally,a simulation system of prediction for molten steel temperature in argon blowing station has been established.
芦永明1,田乃媛1,徐安军1,贺东风1
(1.北京市北京科技大学 冶金与生态工程学院)
摘 要:根据对210 t钢包吹氩站热状态实验数据和实际生产工艺数据的统计分析可知,吹氩站钢水终点温度的主要影响因素是钢水到站温度和总调温废钢用量以及吹氩精炼时间。本文采用正交设计方法通过对现场实际生产数据进行筛选,选取最具有代表性的数据,进而利用SPSS软件,采用多元回归的方法建立了210 t钢包钢水吹氩站终点温度预报模型。之后通过随机抽取的100组实际生产数据对所建模型进行验证,验证结果表明模型在对吹氩站钢水终点温度预报时,预报误差在±10℃内的正确率达到90%以上,说明此模型具有较好的应用效果。最后,建立吹氩站钢水终点温度预报仿真系统,对模型进行了仿真应用。
关键词:吹氩; 钢水温度; 多元回归; 预报模型; 仿真应用