Method and device for evaluating accuracy of model for predicting silicon content in hot metal of blast furnace

Abstract

The invention provides a method and device for evaluating the accuracy of a model for predicting the silicon content in hot metal of a blast furnace. The method comprises the steps of: obtaining working condition parameters influencing the silicon content in the hot metal of the blast furnace in historical data, a silicon content measurement value and a silicon content prediction value of a silicon content prediction model to be evaluated, and, according to a trained accuracy prediction model, classifying the prediction result accuracy of the silicon content prediction value, so that a first evaluation classification result is obtained; classifying the accuracy of the silicon content prediction value according to the silicon content measurement value, so that a second evaluation classification result is obtained; and, obtaining a true positive rate TPR and a false positive rate FPR based on the first evaluation result and the second evaluation result, and evaluating the reliability ofthe silicon content prediction model through a receiver operating characteristic curve ROC. A ROC is drawn through a prediction result; the performance of the prediction model can be overall judged through the indexes, such as an AUC (Area Under Curve); and thus, production can be guided by selection of a proper prediction model for a site.

Description

technical field [0001] The invention relates to the technical field of automatic control of blast furnace smelting, and more specifically, to a method and equipment for evaluating the accuracy of a prediction model of silicon content in blast furnace molten iron. Background technique [0002] The silicon content of molten iron is the key information to characterize the furnace temperature and its changing trend in the blast furnace ironmaking process, and it is also an important physical quantity that reflects the quality of molten iron, energy consumption and other indicators. However, the silicon content of molten iron cannot be directly detected online, resulting in untimely or blind control of the furnace condition. Therefore, timely and accurate prediction of the silicon content of molten iron is a prerequisite for blast furnace process control. So far, there are many kinds of prediction models for silicon content in blast furnace hot metal, but few of these models are ...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a method, equipment, computer program product and storage medium for evaluating the accuracy of a blast furnace silicon content prediction model that overcomes the above problems or at least partially solves the above problems, and solves the problem that the evaluation method of the prediction model in the prior art is too It is one-sided, unable to achieve a comprehensive evaluation of the prediction model, and can only evaluate the prediction results of the model in a given framework

Images