润滑油黏度可见/近红外光谱快速无损检测方法

Fast and Nondestructive Detection for Viscosity of Lubricating Oil Using Visible/Near Infrared Spectroscopy

提出了一种应用可见/近红外光谱技术进行汽车润滑油黏度快速无损检测的新方法。对150个润滑油样本进行光谱扫描和平滑、变量标准化等预处理,比较了不同建模方法的检测精度。采用主成分分析法(PCA)和连续投影算法(SPA)2种方法提取的特征变量作为模型输入变量,分别建立了偏最小二乘模型(PLS)、多元线性回归模型(MLR)和反向传输人工神经网络模型(BPNN)。结果表明,PCA-BPNN和SPA-BPNN模型的预测效果远优于其它模型(PCA-PLS、PCA-MLR、SPA-PLS和SPA-MLR),预测相关系数(r)分别为0.971和0.964。表明BPNN模型可以很好地利用光谱数据中的非线性信息,同时也表明SPA是一种有效的特征波长提取方法,选取的特征波长有利于汽车润滑油黏度快速检测仪器的开发。

Visible/near infrared（Vis/NIR） spectroscopy was used for the fast and nondestructive determination of viscosity of lubricating oil.A total of 150 oil samples were scanned,and 125 samples（25 for each brand） were used as calibration set,and the remaining 25 samples（5 for each brand） were applied as validation set.Different calibration models were developed with the pretreatment of smoothing and standard normal variate（SNV）.The input variables of calibration were the principal component（PC） selected by principal component analysis（PCA） and characteristic wavelengths selected by successive projections algorithm（SPA）.The calibration models were developed by using partial least squares（PLS） and multiple linear regression（MLR） for linear models,and back propagation neural networks（BPNN） for nonlinear models.The results indicated that PCA-BPNN and SPA-BPNN models were better than the linear models（PCA-PLS,PCA-MLR,SPA-PLS and SPA-MLR）.The correlation coefficients（r） were 0.971 for PCA-BPNN and 0.964 for SPA-BPNN,which demonstrated that BPNN could make good use of the nonlinear information in spectral data and SPA was a powerful way for the selection of characteristic wavelengths.The selected wavelengths were helpful for the development of portable lubricating oil viscosity detection instrument.