摘要 |
A method for control of a basic oxygen furnace includes creating a computer-implemented model of the basic oxygen furnace based on support vector regression (SVR) such that the model is configured to produce a prediction of first turndown quality of the basic oxygen furnace. Historical data associated the first turndown quality is received. The model is applied to the historical data to produce a prediction of first turndown quality, which is compared to an actual measurement of first turndown quality. The control recipe for the basic oxygen furnace based at least in part upon a result of the comparison. The system generally includes a preprocessor, a database configured to store historical data associated with first turndown quality of the basic oxygen furnace, a prediction module coupled to the database, and a run-to-run control module. A graphical user interface is provided for visualization and adjustment of data.
|