New digital services for manufacturing industry using analytics: The case of blast furnace thermal regulation
Authors
U. Iffat, S. Bhatia, A. Tantar, J. Sanz, C. Schockaert, A. Schimtz, F. Giroldini, Y. Reuter, and F. Hansen
Reference
in proceedings of the 20th IEEE International Conference on Business Informatics (CBI 2018), Vienna, Austria, 11-13 July 2018, vol. 2, pp. 89-91, 2018
Description
This summary presents the case of digitization of the thermal regulation process in blast furnace. Data from multiple sensors placed on the furnace is leveraged, along with advanced analytics methods, to forecast the temperature of the hot metal produced at the outlet of the furnace. Regulation of hot metal temperature is key to ensuring the quality of the end product. The paper discusses various aspects of this problem such as the choice of relevant predictors used in the model, identification of optimal time lags, challenges due to the non-uniform nature of the hot metal temperature sampling, modelling temperature evolution process within casts vs across casts and evolving nature of the furnace state. The paper also highlights key findings with respect to the existing literature of work on related topics and outlines the practical issues identified with respect hot metal temperature forecasting.
