STUDY ON BRAND IDENTIFICATION OF LUBRICATING OIL USING SENSITIVE WAVELENGTHS OF VISIBLE AND SHORT-WAVE NEAR INFRARED SPECTROSCOPY
The feasibility of Visible and short-wave near-infrared spectroscopy(VIS/WNIR) techniques as means for the nondestructive and fast brand identification of lubricating oil was evaluated. And selected sensitive bands was found valdated . 90 lubricating oil samples were randomly selected for the calibration set, while the remaining 90 samples for the prediction set. smoothing way of moving average and standard normal variate (SNV) were used to pretreat spectra data. Based on principal components analysis, 2-D principal components plot was clustered well. Least-squares support vector machines (LS-SVM) was applied to brand identification based on absorbance spectra from 421 nm to 1075 nm. Finally, recognition ratio of 100％ was obtained. Sensitive bands, 442 and 926 nm were obtained by loading weights in partial least squares and discrimination power in principal components analysis (PCA). The prediction results of 100％ indicated that the selected wavelengths reflected the main characteristics of lubricating oil of different brands based on SWNIR spectroscopy and LS-SVM model.
Lubricating oil LS-SVM PLS VIS/WNIR Nondestructive
JIA-JIA YU YONG HE HAI-QING YANG
Department of Biosystem Engineering, Zhejiang University, Hangzhou 310029, China Department of Biosystem Engineering, Zhejiang University, Hangzhou 310029, China College of Informat
国际会议
2008 International Conference on Machine Learning and Cybernetics(2008机器学习与控制论国际会议)
昆明
英文
1450-1454
2008-07-12（万方平台首次上网日期，不代表论文的发表时间）