Prediction Model of End-Point Carbon Content for BOF Based on LM BP Neural Network

Abstract

Reaction process of BOF steelmaking is a very complex physical chemistry process which is very difficult to describe linearity. The traditional static model has poor accuracy, and the target hit rate is low. Based on the analysis of the major influential factors, the influential factors of converter smelting on the endpoint control of carbon content are fixed in this paper. A prediction model of end-point carbon content for BOF is established based on Levenberg-Marquardt (LM) algorithm of BP neural network. The simulated results show that the hitting rates of end-point carbon content reached 80% when accuracy of target error is ±0.025%.