Prediction of Si content in blast furnace based on fuzzy model

According to the characteristics of massive input and output data, complicated production process and the difficulty to acquire accurate mathematical model of the non-linear blast furnace system, this paper proposes a fuzzy system modeling method based on data driving. This paper utilizes the fuzzy clustering algorithms combined nearest neighbor clustering and fuzzy c-means clustering to classify the input space. And then the parameters of model are identified by adopting recursive least square algorithm. Subsequently, the fuzzy rules are extracted to build the T-S fuzzy model of blast furnace system whose output is the Si content in molten iron which has guiding significance on predicting the trend of furnace temperature changing. The furnace data of number 6 furnace in Baotou Steel is acquired to complete MATLAB simulation. The simulation results verify the feasibility of the method.