Temperature Prediction in Blast Furnaces: A Machine Learning Comparative Study
In the iron and steel industry, the stable operation of blast furnaces with efficient hot metal temperature (HMT) monitoring and control is a very important task in the process to generate high-quality hot metal. In general, the operation of blast furnaces mostly relies on experienced based decisions of human operators, which use the latter measures of hot metal temperature and other operational variables to execute control decisions. However, due to the large number of variables and complex interaction among them, the operation of such equipment is not an easy task. This work proposes a prediction system as the first step of a larger and more complex control system for improving the efficiency of iron production considering the scenario in Brazil. It compares several machine learning models (e.g., KNN, LR, EBM, LGBM, RF, SVM, XGBoost, and MLP) in the task of hot metal temperature prediction. A good temperature prediction system will allow to better plan the control actions ahead in order to stabilize the furnace temperature during hot metal production. The proposed method was evaluated using real data from an steel producing company. Results show that our system can effectively predict the hot metal temperature 2:03 ±0:44,4:51 ± 0:51 and 07:40 ± 00:58 hours ahead (with mean average error of 9.55, 10.00 and 12.31, respectively), when compared to the baselines (with mean average error of 12.61, 14.91 and 19.17, respectively).