2020 Volume 60 Issue 5 Pages 898-904
The application of big data in industry can solve industrial problems through data analysis. In order to establish the evaluation and prediction system of comprehensive state in a commercial blast furnace (BF), problems of ironmaking data were observed, standardized processing technologies were employed to process ironmaking data, big data platform of ironmaking was deployed based on the collected data. Hot metal production, hot metal production quality [Si+Ti] and fuel rate were selected as the target parameters to reflect the comprehensive state of BF. 26 key parameters that were most closely related to BF production and target parameters were screened out through Python 3.7. The relationship between 26 key parameters and hot metal production, [Si+Ti], fuel rate was analyzed, respectively. The first five parameters with strong correlation with target parameters were selected as important data for analysis and prediction. The comprehensive state was scored according to scoring rules derived from data analysis, the total score could be obtained to evaluate the comprehensive state of the commercial BF. The prediction of comprehensive state was realized based on a large number of historical data. The optimization of model was completed and the model could be run online, the evaluation and prediction system helped operators optimize BF operation.
