Electric arc furnace end point carbon content prediction method

Abstract

The invention relates to a method for predicting the end point carbon content of an electric arc furnace, which comprises the following steps of: inversely deducing the weight of molten steel in the electric arc furnace of each heat by a refining furnace, measuring and calculating the yield of each steel material in the electric arc furnace by utilizing multiple linear regression and a loss function, and measuring and calculating the weight of the molten steel and the residual steel amount in the electric arc furnace of each heat by utilizing the yield. The method comprises the following steps: by taking electric arc furnace consumption data including residual steel quantity as an input quantity, reversely deducing the end point carbon content of the electric arc furnace by a refining furnace as an output quantity, preprocessing the data, and training a neural network-based electric arc furnace end point carbon content prediction model; inputting electric arc furnace consumption data to calculate an electric arc furnace end point carbon content prediction model to obtain an electric arc furnace end point carbon content prediction value; influence factors of the electric arc furnace endpoint carbon content are fully considered, the problem that electric arc furnace molten steel weighing, residual steel amount and electric arc furnace smelting sampling endpoint carbon measurement and calculation are not accurate is solved, the method is particularly suitable for the steel-making process of the electric arc furnace with narrow-range carbon content, and the accuracy of electric arc furnace endpoint carbon control is improved.

Description

technical field [0001] The invention belongs to the technical field of iron and steel smelting production control methods, in particular to a method for predicting carbon content at the end point of an electric arc furnace. Background technique [0002] In the process of electric arc furnace steelmaking, the electric arc furnace needs to control the carbon at the end point, and the decarburization speed is faster during smelting, and the range of carbon content required by the steel grade is still very narrow, so the control of the carbon content at the end point is a current research focus . [0003] At present, there are many researches on the control of the carbon content of the end point of the converter by neural network, but there is almost no control scheme of the end point of the electric arc furnace. The main reasons are as follows: (1) Compared with the converter, the electric arc furnace is used in steelmaking , in order to prevent the slag floating on the surfac...

AI Technical Summary

Problems solved by technology

[0003] At present, there are many researches on the control of the carbon content of the end point of the converter by neural network, but there is almost no control scheme of the end point carbon content of the electric arc furnace. The main reasons are as follows: (1) Compared with the converter, when the electric arc furnace is making steel, in order to prevent the slag floating on the surface of the molten steel from mixing into the ladle, when the electric arc furnace pours molten steel into the ladle, the molten steel in the furnace is not emptied at one time , there will be a surplus, so it is impossible to accurately measure the remaining steel in the electric arc furnace. As the raw material for the next furnace to smelt steel, the amount of remaining steel in the electric arc furnace will seriously affect the control of the end point carbon, which will affect the prediction of the carbon content of the end point of the electric arc furnace by the neural network; (2) the thickness of the slag layer is relatively thick during the electric arc furnace smelting , and the molten steel is violently tumbling during smelting, the composition of the molten steel is uneven, the weighing of the molten steel in the electric arc furnace is inaccurate, and the carbon content detected by sampling at the end of the electric arc furnace is seriously inconsistent with the actual carbon content in the molten steel, resulting in the output parameters of the training neural network. Inaccurate

Images