The invention discloses a method for predicting a
coke ratio in a
blast furnace smelting process, and relates to the technical field of
blast furnace smelting, and the method comprises the following steps: constructing a calculation
mathematical model of the
coke ratio, calculating a theoretical replacement ratio of a
blast furnace through a theoretical replacement ratio formula, and analyzing a relationship between a blast furnace parameter and a pulverized
coal utilization rate through
machine learning to obtain a predicted pulverized
coal utilization rate; and obtaining a
coke ratio correction coefficient through
machine learning, and substituting the obtained theoretical replacement ratio, the pulverized
coal utilization rate and the coke ratio correction coefficient into a calculation
mathematical model of the coke ratio to obtain a predicted coke ratio. According to the method for predicting the coke ratio in the
blast furnace smelting process, the predicted value and the actual observed value of the coke ratio show good consistency, the relative error is within an acceptable range, the prediction precision of the method meets the
engineering application requirement, the change of the coke ratio in the
blast furnace smelting process can be effectively predicted, and the method is suitable for popularization and application. The practical feasibility is realized.