Prediction of Blast Furnace Temperature Based on Evolutionary Optimization.

Blast furnace temperature, generally characterized by silicon content, is an important indicator to characterize the stability of furnace conditions in the blast furnace ironmaking process. To achieve its accurate prediction, a prediction model based on multi-objective evolutionary optimization and non-linear ensemble learning is proposed in this paper. First, the input features of each sub-learner are optimized through a modified discrete multi-objective evolutionary algorithm to obtain a set of highly accurate and diverse sub-learners. Subsequently, a nonlinear ensemble method based on extreme learning machine is employed to ensemble the obtained sub-learners to further improve the accuracy and robustness of the prediction model. Furthermore, the effectiveness of the proposed prediction method is verified by experiments based on actual production data.