A multi-output fuzzy model for converter gas holder level prediction in steel industry

Considering that the Linz Donawitz converter gas (LDG) holders in one steel-making zone are correlated with each other, a multi-output fuzzy model with parameter estimation based on Kalman filtering is proposed in this paper. First, a multi-output T-S fuzzy model is established on the basis of a larger number of historical data. Then, the Fuzzy C-means clustering algorithm are used to extract the fuzzy rules and identify the premise parameters, and the parameters of the consequent part can be estimated by an iterative Kalman filter. The effectiveness of the proposed model is demonstrated by the experimental results based on the real data of the gas system of a steel enterprise.