Research on Prediction Model of End Sulfur Content for Converter Smelting

Abstract:

Converter smelting medium-low carbon ferrochrome replaces electronic by oxygen and reduces production energy consumption greatly. Medium-low carbon ferrochrome end sulfur content analysis is necessary to determine the product quality. Now we still use the product sampling, laboratory analysis to determine end sulfur content, which can not realize online monitoring. According to this status, a PLSBP prediction model of end sulfur content has been established based on particle swarm algorithm, which combined partial least squares method and BP neural network. At the same time, A feedback compensation model has been established by dichotomy, which avoided the model failure caused by raw material quality change. The simulation results showed that the hit rate of prediction model was 96.67% within the absolute error, 90% within relative error. High precision was achieved, which provided an important theoretical basis to improve product quality and optimize the production process.

• Periodical: Advanced Materials Research (Volume 1037)
• Pages: 26-29
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.1037.26
• Online since: October 2014
• Authors: Dong Qiu*, Wen Juan Dai, Nan Zhang
• Keywords: Converter, Prediction Model, Smelting, Sulfur Content

References

[1] Zhenhai Jia. The New Craft of AOD in Production of Medium, Low, Simplex Carbon Ferrochrome[J]. Journal of Ferroalloy, 2005, (2), pp.11-16.

[2] Congxian Li. The Influence Factors of Sulfur Content and Control of Ferrochrome by Converter Method[J]. Journal of Ferroalloy, 2003, 06, pp.6-10.

[3] Lingbiao Hu. The Influence Factors and Control Methods of S and P Content of Ferrochrome by Pure Oxygen Smelting[J]. Journal of Ferroalloy, 2002, 01, pp.1-10.

[4] Dagang Li. Analysis of The Best Controlling Factors Influencing the sulfur content in Converter Steelmaking (translation)[J]. Journal of Taigang translations, 2001(4), pp.11-13.

[5] Huishu Zhang, Dongping Zhan, Zhouhua Jiang. Improve Application of Improved BP Neural Network to Final Sulfur Content Prediction of Hot Metal Pre-desulfurization[J]. Journal of Northeastern University (natural science edition), 2007, 28(8), pp.1140-1142.