基于多元线性回归的转炉石灰加入量计算模型
Calculation Model of Converter Lime Addition Based on Multiple Linear Regression
陈君 1苏春阳 2姜亚清1
作者信息
- 1. 大冶特殊钢有限公司工艺研究所,黄石 435001
- 2. 靖江特殊钢有限公司,靖江 214500
- 折叠
摘要
根据铁水条件和钢水要求准确计算石灰加入量,避免熔剂浪费是转炉炼钢的重要要求之一.然而,由于转炉生产节奏快,影响因素多,很难精准地确定石灰加入量.为了准确计算转炉石灰加入量,避免石灰浪费,建立了机理模型、分类模型和多元线性回归模型,用于计算转炉石灰加入量.三种模型相比,纯粹的机理模型计算精度很低,难以直接应用到现场生产中.基于机理模型优化的分类模型,仅有60%的炉次差值在500 kg/炉以内.利用SPSS建立了多元线性回归模型,得到了石灰加入量和其他参数之间的关系WLime=4 147.24×wTSi+529.67×wTC+60.20×Wqf+0.405×VO-1 833.95,验证结果表明,93%的炉次计算差值在400 kg/炉以内,77%的炉次计算差值在300 kg/炉以内,基本能够满足现场使用的要求,模型应用在现场后,石灰加入量得到有效的降低.
Abstract
One of the important requirements for converter steelmaking is to accurately calculate the amount of lime added based on the conditions of hot metal and the requirements of molten steel,in order to avoid waste of flux.However,due to the fast pace of converter production and multiple influencing factors,it is difficult to accurately determine the amount of lime added.In order to accurately calculate the amount of converter lime,avoid lime waste and reduce unnecessary steel-making cost,mechanism model,classification model and multiple linear regression model are established to calculate the amount of converter lime.Compared with the three models,the pure mechanism model has low calculation accuracy and is difficult to be directly applied to field production.According to the classification model optimized by mechanism model,only 60%of the furnace times difference is less than 500 kg/furnace.A multiple linear regression model is established us-ing SPSS,and the relationship between lime addition amount and other parameters is obtained WLime=4 147.24×wTSi+ 529.67×wTC+60.20×Wqf+0.405×VO-1 833.95.The verification results show that,93%of the calculated difference of the furnace is less than 400 kg/furnace,77%of the calculated difference of the furnace is less than 300 kg/furnace,which can basically meet the requirements of field use.After the model is applied in the field,the amount of lime added is effec-tively reduced.
关键词
转炉/石灰/机理模型/多元线性回归模型/精度Key words
Converter/Lime/Mechanism Model/Multiple Linear Regression Model/Accuracy引用本文复制引用
出版年
2024
NETL
NSTL
维普
万方数据