Study on balance scheduling based on the gasholder level prediction for blast furnace gas system in steel industry

摘要

A balance scheduling method based on gasholder level prediction is proposed for the byproduct gas system in this study. First, a class of Gaussian process regression is presented to construct the prediction model, in which a conjugate gradient algorithm is then designed to optimize the model parameters. A reasoning calculation is designed to obtain the gas adjustment amount for the system balance that sustains the whole system security. To verify the quality of the proposed method, the practical data coming from Shanghai Baosteel are employed and the experimental results indicate that the prediction for gasholder level provides the scientific guidance for the energy utility.