Prediction of burden descent speed in blast furnace based on extreme learning machine

The burden charging is the most important operation of the upper operations in blast furnace. The position and descent speed of burden layer can reflect the situation of blast furnace and can guide the operators in the next burden charging. In this paper, the descent speed prediction model of burden layer is established by extreme learning machine algorithm. The model can make single-step and multi-step predictions to the burden descent speed using real radar data and status information in blast furnace. In the simulation part, we collected the real production data in iron-making process and obtained the satisfied and accurate simulation results by employing the proposed scheme.