Classification of Lubricating Oil Types Using Mid-Infrared Spectroscopy Combined with Linear Discriminant Analysis–Support Vector Machine Algorithm

To realize the classification of lubricating oil types using mid-infrared (MIR) spectroscopy, linear discriminant analysis (LDA) was used for dimensionality reduction spectrum data, and model established based on support vector machine (SVM). The spectra samples were pre-processed by interval selection, Savitzky–Golay smoothing, multiple scattering correction, normalization. Kennard–Stone algorithm (K/S) to construct calibration validation sets. percentage correct (%CC) evaluate model. This study compared results obtained with several chemometric methods: PLS-DA, LDA, principal component (PCA)-SVM, LDA-SVM in MIR spectroscopy applications. In both verification sets, achieved 100% favorable results. PLS-DA performed poorly. cyclic resistance ratio (CRR) set classified via LDA PCA-SVM as 100%, but CRR not good. superior other three models; it exhibited good robustness strong generalization ability, providing a new method spectroscopy.