摘要
Abstract
Smoke point in kerosene is one of the important indexes to reflect the performance of kerosene,In this paper,a simple,fast and nonde-structive detection of smoke point is mainly achieved by the Near-infrared Spectroscopy,the model can be built under taking kerosene for in-stance,78 samples were collected,and multivariate statistical analysis was carried out based on partial least square method.The calibrations of the smoke point in kerosene were also performed by different pretreatment methods.The results showed that the second derivative produced a bet-ter prediction,The correction coefficient(R),the Ratio of standard deviation and standard error of prediction(RPD),the Root Mean Squared Er-ror of Cross-Validation(RMSECV)were 0.9252,3.33,and 0.093.The results showed that the prediction accuracy of the model was higher,and this model could determine the smoke point in kerosene used as a rapid and lossless method to detect the quality of kerosene.关键词
近红外光谱/航空煤油/烟点/二阶导数/统计方法/偏最小二乘法Key words
Near-infrared Spectroscopy/Kerosene/the Smoke Point/Second Derivative/Statistical Techniques/Partial Least Square Method分类
能源科技
