Blast furnace molten iron silicon content prediction method based on two-dimensional self-attention enhancement GRU model

Abstract

The invention discloses a blast furnace molten iron silicon content prediction method based on a two-dimensional self-attention enhancement GRU model, and belongs to the field of industrial process monitoring, modeling and simulation. Effective information is obtained from real blast furnace production data to establish a model, the silicon content of molten iron is predicted in advance, and subsequent production operation is guided. In consideration of different influences of parameter variables on the silicon content of molten iron of a product in the blast furnace production process and dynamic changes along with time, it is proposed that self-attention is increased in the feature dimension of the GRU model, and the dynamic weight of each parameter variable is obtained; meanwhile, by considering system dynamics and large time lag problems, a time dimension self-attention mechanism based on causal convolution is disclosed, thereby realizing enhanced perception of local dynamic characteristics of blast furnace operation parameters and soft benchmarking of the operation parameters and process indexes. The method has a good fitting effect on a blast furnace system with large time lag and strong dynamic property, and can realize accurate prediction on the silicon content of the molten iron of the blast furnace.

Description

Technical field [0001] The present invention belongs to the field of industrial process monitoring, modeling, and simulation, and, in particular, the present invention relates to a prediction method of ferroelectric silicon content based on two-dimensional self-focus reinforced GRU model blast furnace. Background technique [0002] Blast furnace iron is an important process of industrial production, and improving the impact of blast furnace iron making will bring huge economic benefits. However, blast furnace iron is an extremely complex high-temperature high pressure, strong coupling, strong interference industry process, which makes it difficult for people to measure the temperature pressure in the blast furnace, it is difficult to accurately assess the operating state of the blast furnace. Research scholars have used the iron silicon content as an important reference index to evaluate the temperature of the blast furnace. Under a stable condition, when the silicon content is b...

AI Technical Summary

Problems solved by technology

However, the traditional detection method of silicon content is obtained by testing the molten iron. The results obtained in this way have a serious lag, and the production reference is greatly reduced, and the operation status of the blast furnace cannot be adjusted as soon as possible.

[0003] Existing data-driven models for silicon content modeling and prediction mainly include: regression analysis, BP neural network, support vector machine, etc. Although such models can realize the nonlinear modeling of the blast furnace system, they cannot fully extract the dynamics of the system. For the problem of variable time lag, the time lag of each variable can only be analyzed manually through correlation. These deficiencies affect the prediction accuracy of the model to a certain extent.

Images