Knowledge-Based Control and Optimization of Blast Furnace Gas System in Steel Industry

Aiming at the control and optimization problem of blast furnace gas (BFG) systems in the steel industry, a knowledge-based optimal control algorithm combining fuzzy rules extraction with neural networks (NNs) ensemble-based prediction is proposed. On one hand, a fuzzy model is designed to extract the expert control knowledge from the historical data of the industrial process after community detection, and then, a great deal of scheduling knowledge is employed to compose a fuzzy rule base, which can be used for fuzzy inference of control scheme with a new input. On the other hand, data-driven NNs ensemble is built to model the BFG system for prediction. Meanwhile, the prediction results can provide the inputs when using fuzzy rule base for control and optimization. Finally, a BFG system of one steel enterprise is studied in this paper for experiments, which verifies the effectiveness and practicability of the proposed method.