Prediction of hot metal temperature based on data mining

Abstract Accurately and continuously monitoring the hot metal temperature status of blast furnace (BF) is a challenging job. To solve this problem, we propose prediction model based on AdaBoost integrated algorithm using real production data BF. We cleaned raw analysis technology combined with metallurgical process theory, which mainly included integration, outliers elimination, missing value supplement. The redundant features were removed Pearson’s thermodynamic diagram analysis, input parameters preliminarily determined by recursive feature elimination method. built ensemble dataset selected as well derived K-mean clustering tags. results show that performance K-means tags has been further improved, accurate forecast molten iron achieved. can achieve an accuracy more than 90% error ±5°C.