摘要
球团矿作为高炉入炉的必备原料,其性能直接影响铁水质量和高炉顺行,对球团矿进行入炉前的质量检测不可或缺。以球团矿物理性能指标中的抗压强度、制备过程中的粘结率作为球团矿质量指标进行预测,并对球团矿进行配料寻优。首先,对所搜集的数据进行正态性检验和相关性分析;然后,建立多元线性回归分析模型,分别拟合出抗压强度、粘结率与原料之间的多元线性回归方程;最后,建立配料寻优模型,在已知球团矿抗压强度、粘结率的情况下,将抗压强度、粘结率的实测值与预测值之差作为寻优目标,采用遗传算法和熵权法反向优化对模型进行求解。研究发现:原料w(Ca)、二元碱度(R_2=CaO/SiO_2)、三元碱度(R_3=(CaO+MgO)/SiO_2)与球团矿抗压强度、粘结率存在相关性,可以较好地探究原料与指标的关系并实现配料优化。利用多元线性回归分析模型求解的抗压强度预测相对误差为4.826%,粘结率的预测相对误差为4.360%;利用配料寻优模型求解得到最合适的3种配料预测值与实测值之间的预测相对误差分别为1.18%、0.99%、2.63%,该误差相对于回归分析的误差更小,可作为球团厂配料参考的重要依据,进而为改善球团矿质量提供依据。
As a necessary raw material for the blast furnace,the pellets directly affect the quality of the molten iron and the smooth operation of blast furnace,and the quality inspection of the pellets before entering the furnace is indispensable. In this paper,the compressive strength in the physical properties of the pellets and the sticking index in the preparation process are predicted as the pellet quality indicators,and the pellet proportioning is optimized. First,the normalized test and correlation analysis are performed on the collected data. Then,a multivariate linear regression analysis model is established to fit the multiple linear regression equation between compressive strength,sticking index and raw materials,and finally the proportioning optimization model is established. Under the condition of known compressive strength and sticking index of pellets,the difference between the measured value and the predicted value of compressive strength and sticking index is taken as the proportioning optimization target,and the model is solved by using the genetic algorithm and the entropy weight method. It is found that the raw material w(Ca),alkalinity R_2,and alkalinity R_3 have a correlation with the compressive strength and sticking index of pellets,which can better explore the relationship between raw materials and indicators and realize proportioning optimization. The predicted relative error of the compressive strength calculated by the multiple linear regression analysis model is 4.826%,and the predicted relative error of the sticking index is 4.360%. The relative errors between the predicted and measured values of the three most suitable proportioning materials are 1.18%,0.99% and 2.63% respectively,which are smaller than those of the regression analysis and can be used as an important basis for improving the quality of pellets.
作者
陈金鸿
张柳叶
贾学勇
张磊
梁精龙
Chen Jinhong;Zhang Liuye;Jia Xueyong;Zhang Lei;Liang Jinglong(College of Metallurgy and Energy, North China University of Science and Technology ,Tangshan 063210 ,Hebei;College of Mechanical Engineering ,North China University of Science and Technology ,Tangshan 063210 ,Hebei;College of Electrical Engineering, North China University of Technology ,Tangshan 063210, Hebei)
出处
《烧结球团》
北大核心
2019年第5期43-49,共7页
Sintering and Pelletizing
基金
国家自然科学基金(51774143)
国家自然科学基金(51874141)
华北理工大学大学生创新创业训练计划项目(X2018093)
关键词
球团矿
抗压强度
粘结率
多元线性回归分析
遗传算法
熵权法
配料寻优
pellet
compressive strength
sticking index
multiple linear regression analysis
genetic algorithm
entropy weight method
optimization of ingredients
