2013 International Conference on Computing, Networking and Communications (ICNC) (2009)
Aug. 14, 2009 to Aug. 16, 2009
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ICNC.2009.668
A new approach for forecasting silicon content in blast furnace hot metal is presented based on the principal component regression. Firstly, with the pre-processed data selected from Laiwu Iron and Steel Group Co., the eigenvalues and eigenvectors of the data correlation matrix are calculated. Then the eigenvectors are used for calculation of the principal components and four of them are selected to represent all the information about blast furnace ironmaking process. Finally, compared with the conventional autoregressive method, our approach is more accurate to predict the silicon content. The main benefit of the approach is that it can reduce the number of factors affecting silicon content and eliminate the multicollinearity between them.
blast furnace, silicon content, principal component analysis, prediction
Juner Ma, Wenhui Wang, "Principal Component Regression Approach for Forecasting Silicon Content in Hot Metal", 2013 International Conference on Computing, Networking and Communications (ICNC), vol. 02, no. , pp. 590-593, 2009, doi:10.1109/ICNC.2009.668