Thickness Measurement of Insulation Coating by NIR Spectrometry Based on Boosting-KPLS
HAO Hui-min1,2, LI Shi-wei3, ZHANG Wen-dong1, LI Peng-wei1, HAO Jun-yu2, LU Hai-ning2, Ken Jia4, ZHANG Yong5
1. Micro/Nano System Research Center, Taiyuan University of Technology, Taiyuan 030024, China 2. Automation Company of TISCO, Taiyuan 030003, China 3. Northwest Institute of Nuclear Technology, Xi’an 710024, China 4. Department of Research and Development, Brimrose Corporation of America, Baltimore 21152-9201, USA 5. State Key Laboratory of Electrical Insulation for Power Equipment, Xi’an Jiaotong University, Xi’an 710049, China
Thickness Measurement of Insulation Coating by NIR Spectrometry Based on Boosting-KPLS
HAO Hui-min1,2, LI Shi-wei3, ZHANG Wen-dong1, LI Peng-wei1, HAO Jun-yu2, LU Hai-ning2, Ken Jia4, ZHANG Yong5
1. Micro/Nano System Research Center, Taiyuan University of Technology, Taiyuan 030024, China 2. Automation Company of TISCO, Taiyuan 030003, China 3. Northwest Institute of Nuclear Technology, Xi’an 710024, China 4. Department of Research and Development, Brimrose Corporation of America, Baltimore 21152-9201, USA 5. State Key Laboratory of Electrical Insulation for Power Equipment, Xi’an Jiaotong University, Xi’an 710049, China
摘要： A novel thickness measurement method for surface insulation coating of silicon steel based on NIR spectrometry is explored. The NIR spectra of insulation coating of silicon steel were collected by acousto-optic tunable filter (AOTF) NIR spectrometer. To make full use of the effective information of NIR spectral data, discrete binary particle swarm optimization (DBPSO) algorithm was used to select the optimal wavelength variates. The new spectral data, composed of absorbance at selected wavelengths, were used to create the thickness quantitative analysis model by kernel partial least squares (KPLS) algorithm coupled with Boosting. The results of contrast experiments showed that the Boosting-KPLS model could efficiently improve the analysis accuracy and speed. It indicates that Boosting-KPLS is a more accurate and robust analysis method than KPLS for NIR spectral analysis. The maximal and minimal absolute error of 30 testing samples is respectively -0.02 μm and 0.19 μm, and the maximal relative error is 14.23%. These analysis results completely meet the practical measurement need.
Abstract：A novel thickness measurement method for surface insulation coating of silicon steel based on NIR spectrometry is explored. The NIR spectra of insulation coating of silicon steel were collected by acousto-optic tunable filter (AOTF) NIR spectrometer. To make full use of the effective information of NIR spectral data, discrete binary particle swarm optimization (DBPSO) algorithm was used to select the optimal wavelength variates. The new spectral data, composed of absorbance at selected wavelengths, were used to create the thickness quantitative analysis model by kernel partial least squares (KPLS) algorithm coupled with Boosting. The results of contrast experiments showed that the Boosting-KPLS model could efficiently improve the analysis accuracy and speed. It indicates that Boosting-KPLS is a more accurate and robust analysis method than KPLS for NIR spectral analysis. The maximal and minimal absolute error of 30 testing samples is respectively -0.02 μm and 0.19 μm, and the maximal relative error is 14.23%. These analysis results completely meet the practical measurement need.
HAO Hui-min1,2, LI Shi-wei3, ZHANG Wen-dong1, LI Peng-wei1, HAO Jun-yu2, LU Hai-ning2, Ken Jia4, ZHANG Yong5. Thickness Measurement of Insulation Coating by NIR Spectrometry Based on Boosting-KPLS[J]. 光谱学与光谱分析, 2011, 31(08): 2081-2085.
HAO Hui-min1,2, LI Shi-wei3, ZHANG Wen-dong1, LI Peng-wei1, HAO Jun-yu2, LU Hai-ning2, Ken Jia4, ZHANG Yong5. Thickness Measurement of Insulation Coating by NIR Spectrometry Based on Boosting-KPLS. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2011, 31(08): 2081-2085.
[1] Han Xiaoguang, Zhang Dunming. Science & Technology Information, 2008, 25: 376. [2] Peng Xuelian. Surface Technology, 2004, 33(6): 80. [3] Chen Xiuming, Lin Li, Li Ximeng, et al. Journal of Aeronautical Materials, 2009, 1: 87. [4] Han Xiaoyuan, Zhuo Shangjun, Wang Peiling. Spectroscopy and Spectral Analysis, 2006, 26(1): 159. [5] Hao Huimin, Cao Jianan, Yu Zhiqiang, et al. Spectroscopy and Spectral Analysis, 2009, 29(8): 2087. [6] Zhu Dazhou, Ji Baoping, Shi Bolin, et al. Journal of Infrared and Millimeter Waves, 2009, 28(5): 371. [7] Hu Chuanxin, Song Youhui. Principle and Application of Coating Technology. Beijing: Chemical Industry Press, 2003. [8] Pearson R K. Control Systems Technology, 2002, 10(1): 55. [9] Yan Yanlu, Zhao Longlian, Han Donghai, et al. Foundation and Application of NIR Spectral Analysis. Beijing: China Light Industry Press, 2005. [10] Kennedy J, Eberhart R. Proceedings of the Conference on Systems, Man, and Cybernetics，1997. 4104. [11] Duffy N, Helmbold D. Machine Learning, 2002, 47(23): 153. [12] Yoav Freund, Robert E. Schapire. Journal of Japanese Society for Artificial Intelligence, 1999, 14(5): 771. [13] Friedman J H. The Annals of Statistics, 2001, 29: 1189. [14] Taylor J S, Cristianini N. Kernel Methods for Pattern Analysis. Cambridge University Press, Cambridge, England, 2004. [15] Smola A J. Learning with Kernels. Berlin: Technical University of Berlin, 1998. [16] Shawe-Taylor J, Cristianini N. Kernel Methods for Pattern Analysis. Cambridge: Cambridge University Press, 2004. [17] Hao Huimin. Study on Quantitative Analysis Methods for Multi-component Gaseous Mixture by Infrared Spectroscopy Based on Kernel Method. Xi’an：Xi’an Jiaotong University, Doctoral Thesis, 2008. 74.
