Prediction model of permeability index based on Xgboost
ZHAO Jun1,2, LI Hong-wei3, LIU Xiao-jie3, LI Xin3, LI Hong-yang3, LÜ Qing1,3
1. College of Metallurgy, Northeastern University, Shenyang 110819, Liaoning, China; 2. Tangshan Branch, HBIS Group Co., Ltd., Tangshan 063020, Hebei, China; 3. College of Metallurgy and Energy, North China University of Technology, Tangshan 063210, Hebei, China
Abstract:Blast furnace permeability index is an important parameter of blast furnace monitoring index. It is necessary to control the change trend of blast furnace permeability index in time and predict it accurately for the operator to keep the blast furnace running smoothly. Based on the actual production data of a blast furnace site, this paper deals with the problems of original data such as outliers and missing values, and standardizes the data. Spearman, MIC and random forest feature elimination method were used to select feature variables of standardized data, and Xgboost model was used for prediction. The results show that Xgboost has more advantages than random forest and linear regression model, the accuracy of the model is 94.27% within the error ±1.5%, Xgboost can accurately predict the next hour permeability index and guide the blast furnace production in time, keep the blast furnace running smoothly and stably.
赵军, 李红玮, 刘小杰, 李欣, 李宏扬, 吕庆. 基于Xgboost的高炉透气性指数预测模型[J]. 中国冶金, 2021, 31(3): 22-29.
ZHAO Jun, LI Hong-wei, LIU Xiao-jie, LI Xin, LI Hong-yang, LÜ Qing. Prediction model of permeability index based on Xgboost[J]. China Metallurgy, 2021, 31(3): 22-29.
Stanislaw Osowski,Konrad Garanty.Forecasting of the daily meteorological pollution using wavelets and support vector machine[J].Engineering Applications of Artificial Intelligence,2006(10):1.
Reshef D N,Reshef Y A,et al.Detecting novel associations in large data sets[J].Science,2011,334(6062):1518.
[13]
Chatterjee S,Dey D,Munshi S.Integration of morphological preprocessing and fractal based feature extraction with recursive feature elimination for skin lesion types classification[J].Computer Methods and Programs in Biomedicine,2019,178:201.