摘要
采用傅立叶变换近红外光谱技术与偏最小二乘方法结合测定柴油的十六烷值。讨论了训练集样本光谱范围的选择、光谱预处理方法及最佳主因子数的选择。近红外光谱技术测定的结果与标准方法测定的结果有良好的相关性(R2=0.9724)。采用不同的校正模型测定了催化裂化柴油馏分、成品柴油等不同类型样品的十六烷值。与标准方法相比,近红外光谱方法具有重现性好、分析速度快、使用样品量少、分析成本低等优点。
The method for determining cetane number by using FT NIR and partial least square (PLS) regression method was studied.Selection of spectral ranges,pretreatment of spectra,factor number used in PLS and composition of training sample set are discussed.As a result of this study,good correlation with the cetane numbers determined by ASTM method (R 2=0 9724) and good repeatability (standard deviation=0 02) were obtained.However,Different calculation models have to be used for high aromatic samples from FCC and ordinary samples.It is suggested that this presented method with its advantages of better repeatability,shorter time of testing,lower cost and smaller sample size,can be used as an alternative for determining cetane unmber of diesel oils.
出处
《石油学报(石油加工)》
EI
CAS
CSCD
北大核心
1999年第4期62-66,共5页
Acta Petrolei Sinica(Petroleum Processing Section)
关键词
柴油
十六烷值
近红外光谱
偏最小二乘方法
diesel,cetane number,near infrared spectroscopy,partial least square regression