Blended coal’s property prediction model based on PCA and SVM

Abstract

In order to predict blended coal’s property accurately, a new kind of hybrid prediction model based on principal component analysis (PCA) and support vector machine (SVM) was established. PCA was used to transform the high-dimensional and correlative influencing factors data to low-dimensional principal component subspace. Well-trained SVM was used to extract influencing factors as input to predict blended coal’s property. Then experiments were made by using the real data, and the results were compared with weighted averaging method (WAM) and BP neural network. The results show that PCA-SVM has higher prediction accuracy in the condition of few data, thus the hybrid model is of great use in the domain of power coal blending.

Keywords

prediction model / blended coal’s property / support vector machine / principal component analysis

References

[1] ChengY.-q., YuZ.-feng.. Clean coal combustion technology is the main domain of clean energy technology in China [J]. Clean Coal Combustion Technology, 2000, 6: 5-15

[2] FengJ.. Achievements in pollutant emission control from coal combustion in China [J]. Fuel and Energy Abstracts, 1997, 38: 438

[3] ChenX.-ling.. Research on linear programming model of coal blending [J]. Coal Processing & Comprehensive Utilization, 2003, 7(4): 34-38

[4] QuJ.-k., HeJ.-dong.. The study of the additivity of main properties of blending steam coal [J]. Journal of Heilongjiang Mining Institute, 2000, 10(3): 17-19

[5] LiuZ.-c., LuZ.-hua.. Coal quality index of blended steam coal and structure relationship with each single coal blended rate [J]. Coal Science Technology, 2004, 10(8): 62-64

[6] LiY., ZhouJ.-hu.. Research on BP neural network in the prediction model of optimized coal blending [J]. Coal Conversion, 2002, 25(11): 79-85

[7] LiY., ZhouJ.-hu.. Research on BP neural network for prediction of blended coal’s properties [J]. Power System Engineering, 2002, 18(2): 1-5

[8] YinC.-g., ZhouJ.-hu.. Research on application of artificial neural networks method in the optimum steam coal blending [J]. Journal of China Coal Society, 1997, 24(9): 343-348

[9] LiuZ.-c., LuZ.-hua.. Neural networks technology of the structural relation between quality indexes of blending coals and proportion of single coal in power blending coal [J]. Journal of China Coal Society, 2004, 29(5): 117-120

[10] GuiX.-y., MaY.-dong.. Application in power coal-mining with nonlinear optimize theory [J]. China Mining Magazine, 2005, 14(8): 49-51

[11] StampfliJ., GoodmanV.The mathematics of finance [M], 2003, New York, Springer Publishing Company

[12] ShiJ.-y., ChaY.-y., WangZ.-ying.. Application of principal components analysis in enterprises economical achievements [J]. Journal of Harbin Engineering University, 2005, 26(14): 693-696

[13] VapnikV.The nature of statisfica1 learning theory [M], 1995, New York, Springer-Verlag

[14] TanD.-ning.. Small-sample machine learning theory-statistical learning theory [J]. Journal of Nanjing University of Science and Technology, 2001, 25(7): 108-112

[15] QiH.-n.. Support vector machines and application research overview [J]. Computer Engineering, 2003, 5(4): 6-9

[16] YeN., LiangZ.-peng.. SVM nonlinear regression algorithm [J]. Computer Engineering, 2005, 31(21): 19-21

[17] ZhangH.-l., ZouZ., LiJ., ChenX.-tao.. Flame image recognition of alumina rotary kiln by artificial neural network and support vector machine methods [J]. Journal of Central South University of Technology, 2008, 15(1): 39-43

[18] LiuC.-b., YuL.-s., JiangX.-liang.. Particle growth mechanism of nanocrystalline zirconia powder during high temperature heat treatment [J]. Trans Nonferrous Met Soc China, 2007, 17(5): 1022-1027

[19] ShenZ.-w., LiQ.-ju.. The researches on the relationship between support vector machine and neural network [J]. Journal of Biomathematics, 2006, 21(7): 204-208

[20] DuS.-x., WuT.-jun.. Support vector machines for regression [J]. Journal of System Simulation, 2003, 15(3): 1580-1585