Please wait a minute...
浙江大学学报(工学版)
J4  2010, Vol. 44 Issue (4): 696-699    DOI: 10.3785/j.issn.1008-973X.2010.04.013
    
Prediction of silicon content in blast furnace hot metal
based on TGARCH model
PAN Wei, LIU Xiangguan, ZENG Jiusun
Department of Mathematics, Zhejiang University, Hangzhou 310027, China
Download:   PDF(0KB) HTML
Export: BibTeX | EndNote (RIS)      

Abstract  

In order to account for the high volatility of silicon content in blast furnace hot metal, threshold generalized autoregressive conditional heteroskedastic (TGARCH) model was used to construct a predictive model for the time series of silicon content in hot metal. Portmantea Q test, Lagrange multiplier (LM) test and asymmetric test were applied to the silicon content series. It was proved that silicon content series exhibits ARCH and asymmetric effect, which validated the application of TGARCH model. A TGARCH (1,1,1) model was then adopted and the coefficients were estimated by maximum likelihood method. Two criteria, i.e. hit rate and error rate, were used to evaluate the performance of the model. This approach takes the asymmetric effect as well as ARCH effect of silicon content into account and is more appropriate for prediction of silicon content in hot metal. Data collected from No.6 blast furnace of Baotou Iron & Steel Corporation were used to test the identified model and good results were achieved.

Published: 14 May 2010
CLC:  TG250.2  
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
Cite this article:

BO Wei, LIU Xiang-Guan, CENG Jiu-Sun. Prediction of silicon content in blast furnace hot metal
based on TGARCH model. J4, 2010, 44(4): 696-699.

URL:

http://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2010.04.013     OR     http://www.zjujournals.com/eng/Y2010/V44/I4/696


TGARCH模型预测高炉铁水硅质量分数

为了更好地反映高炉铁水硅质量分数序列的高波动特性,利用门限广义自回归条件异方差（TGARCH）模型对硅质量分数序列进行预测.应用Portmantea Q检验、拉格朗日乘子检验以及非对称项系数显著性检验,验证了高炉铁水硅质量分数序列存在异方差性和非对称性.在此基础上将TGARCH模型应用于高炉铁水硅质量分数预测,采用极大似然估计法确定参数,建立TGARCH(1,1,1)预测模型,并采用命中率和误差率2种评价准则对预测结果进行分析.这种方法克服了以往模型没有考虑序列非对称性影响的缺陷,更加适合于高炉铁水硅质量分数的预测.将预测模型应用于包钢6号高炉,取得了较好的预测效果.

1］ 刘祥官,刘芳.高炉炼铁过程优化与智能控制系统［M］.北京：冶金工业出版社,2003: 4360.
［2］ 王玉涛,严其艳,杨钢,等.高炉铁水含硅量的动态神经网络多步预测［J］.仪器仪表学报,2006,27(11): 14481451.
WANG Yutao, YAN Qiyan, YANG Gang, et al. Multistep prediction of molten iron silicon content in blast furnace using dynamic neural network ［J］. Chinese Journal of Scientific Instrument, 2006, 27(11): 14481451.
［3］ ZHAO Min, LIU Xiangguan, LUO Shihua. An evolutionary artificial neural networks approach for BF hot metal silicon content prediction ［C］∥ Proceedings of International Conference on Natural Computation. Berlin: SpringerVerlag, 2005: 433436.
［4］ ZAKOIAN J M. Threshold heteroskedastic models ［J］. Journal of Economic Dynamics and Control, 1994, 18(5): 931944.
［5］ TSAY R S. 金融时间序列分析［M］.潘家柱,译.北京：机械工业出版社,2007: 6197.
［6］ ENGLE R F. Autoregressive conditional heteroskedasticity with estimates of the variance of UK inflation ［J］. Econometrica, 1982, 50(4): 9871007.
［7］ BOLLERSLEY T. Generalized autoregressive conditional heteroskedasticity ［J］. Journal of Econometrics, 1986, 31(2): 307327.
［8］ NELSON D B. Conditional heteroskedasticity in asset returns: a new approach ［J］. Econometrica, 1991, 59(2): 347370.
［9］ 谢赤,禹湘,储慧斌.基于TGARCH模型的证券投资惯性反向交易策略实证研究［J］.管理科学,2007,20(3): 6875.
XIE Chi, YU Xiang, CHU Huibin. An empirical study on momentum and contrarian investment strategy of mutual fund based on TGARCH model ［J］. Journal of Management Sciences, 2007, 20(3): 6875.
［10］ 余帆,沈炯,刘西陲.基于自回归条件异方差反向传播网络模型的日前边际电价预测［J］.电网技术,2008,32(8): 6366.
YU Fan, SHEN Jiong, LIU Xichui. Dayahead marginal price forecasting based on autoregressive conditional heteroskedasticityback propagation network model ［J］. Power System Technology, 2008, 32(8): 6366.
［11］ 王燕.应用时间序列［M］.北京：中国人民大学出版社,2005: 139207.
［12］ 高铁梅.计量经济分析方法与建模：EViews应用及实例［M］.北京：清华大学出版社,2006: 171199.
［13］ ZENG Jiusun, GAO Chuanhou, LIU Xiangguan, et al. Model to predict silicon content in blast furnace hot metal ［J］. Asian Journal of Control, 2008, 10(6): 16.
No related articles found!