RESUMEN
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 microm and 0.19 microm, and the maximal relative error is 14.23%. These analysis results completely meet the practical measurement need.
RESUMEN
Feature variable selection and modeling are two of the most principal research contents in spectral analysis. In the present paper, beginning from the introduction of feature spectrum selection based on Tikhonov regularization and discussion on it's application in multi-component mixed alkane gas analysis, 7 sets of feature spectra were abstracted from the absorption spectra of 7 kinds of alkane gas, including methane, ethane, propane, iso-butane, n-butane, iso-pentane and n-pentane. In order to overcome the problem of over-training of neural network, a method called optimal parameter selection of neural netework (NN) was presented to build analysis model of analyte. Optimal parameters were selected from many trained networks with same architecture based on error process. And analysis models of spectral analysis for 7 kinds of alkane gas were built. Finally, the testing analysis results done with standard gases are given. The results show that the method presented in this paper can be used to reduce the cross--sensitivity between any two kinds of gas. The cross-sensitivity is less than 0.5%. The resolving power is as high as 20 X 10(-6).
RESUMEN
Due to its many advantages, such as miniaturization, high accuracy, high resolution, fast scanning speed, increased robustness and good stability, acousto-optic tunable filter (AOTF)-near infrared (NIR) spectrometer has been successfully applied in many fields. However, up to now, the commercial AOTF-NIR spectrometers can only be used for liquid and solid detection, but not for the detection of gaseous samples. In the present paper, the feasibility of quantitative analysis of gaseous mixtures by using AOTF-NIR spectrometer was investigated. A homemade gas cell was assembled to an AOTF-NIR spectrometer with probe for liquid detection to obtain NIR spectra of detected gas samples. The gas samples were composed of two groups: single-component CH4 and ternary component gaseous mixture of CH4, C2 H6, and C3 H8. The detection ability of fitted AOTF-NIR spectrometer was tested firstly. Comparing the absorption spectra of various concentrations, the absorbance of CH4 in absorption bands obviously increased with concentration increasing when the concentration was over 0.1%. According to the detection results, the lower limit of detection (LLD) of the AOTF-NIR spectrometer with gas cell was estimated to be 898 microL x L(-1). Subsequently, the NIR spectra of ternary mixtures were collected. The kernel partial least squares (KPLS) regression was employed to create the quantitative analysis model of three components gases. To evaluate the analysis ability of KPLS model, the PLS model was also created. The prediction results of the identical testing set show that the root mean square error of prediction (RMSEP) of three components predicted by KPLS model was 1.08%, 0.87%, and 0.79%, respectively, less than the RMSEP by PLS model. The exploratory work indicates that accurate quantitative analysis of ternary component alkane gaseous mixtures can be achieved by fitted AOTF-NIR spectrometer despite of some limitations, and KPLS regression is an excellent approach to NIR spectra analysis.
RESUMEN
In the present paper, the authors present a new quantitative analysis method of mid-infrared spectrum. The method combines the kernel principal component analysis (KPCA) technique with support vector regress machine (SVR) to createa quantitative analysis model of multi-component gas mixtures. Firstly, the spectra of multi-component gas mixtures samples were mapped nonlinearly into a high-dimensional feature space through the use of Gaussian kernels. And then, PCA technique was employed to compute efficiently the principal components in the high-dimensional feature spaces. After determining the optimal numbers of principal components, the extracted features (principal components) were used as the inputs of SVR to create the quantitative analysis model of seven-component gas mixtures. The prediction RMSE (phi x 10(-6))of seven-component gases of prediction set samples by use of KPCA-SVR model were respectively 124.37, 72.44, 136.51, 87.29, 153.01, 57.12, and 81.72, ten times less than that by use of SVR model. The elapsed time of modeling and prediction by using KPCA-SVR were respectively 46.59 (s) and 4.94 (s), which was consumedly less than 752.52 (s) and 26.21 (s) by using only SVR These results show that KPCA has an excellent ability of nonlinear feature extraction. It can make the most of the information of entire spectra range and effectively reduce noise and the dimension of the spectra. The KPCA combined with SVR can improve the model's analysis precision and cut the elapsed time of modeling and analysis. From our research and experiments, we conclude that KPCA-SVR is an effective new method for infrared spectroscopic quantitative analysis.
