摘要
选取了148个柴油样品,分别按照GB/T 386-2010和SH/T 0694-2000标准方法测定十六烷值和十六烷指数。通过衰减全反射中红外光谱(ATR-MIR)采集上述柴油样品的中红外光谱图。运用Kennard-Stone算法筛选校正集和验证集,结合偏最小二乘法建立中红外校正模型。十六烷值校正模型的相关系数R^(2)为0.9804,标准偏差RMSECV为0.789。十六烷指数校正模型的R2为0.9825,RMSECV为0.746。应用所建立的校正模型对20个柴油样品的十六烷值和十六烷指数进行预测,预测结果与标准方法准确度相近。实验证明,ATR-MIR法具有分析快速、准确度高、分析成本低等优点。
The results of cetane number and cetane index for 148 diesel samples were determined according to the standard methods of GB/T 386-2010 and SH/T 0694-2000,respectively.Meanwhile,the spectra of those samples were collected by ATR-MIR.Those data were devided into calibration group and validation group by Kennard-Stone algorithm.Two calibration models were established by PLS(partial least square).The correlation coefficient of calibration model for cetane number was 0.9804,and the RMSECV was 0.789.While the correlation coefficient of calibration model for cetane index was 0.9825,and the RMSECV was 0.746.The results of cetane number and cetane index for 20 diesel samples were predicted by the two models.It was founed that the accuracy of ATR-MIR method was similar to standard methods.ATR-MIR method has the advantages of fast analysis,high accuracy,and low analysis cost.
作者
吕玉平
程秀
闻环
徐玲
李荣健
邓小媞
Lv Yuping;Cheng Xiu;Wen Huan;Xu Ling;Li Rongjian;Deng Xiaoti(National Quality Supervision&Testing Center for Petroleum and Petrochemical Products(Guangdong),Huizhou 516003;Hubei Water Resources Technical College,Electromechanical Engineering Department,Wuhan 430070,China)
出处
《广东化工》
CAS
2021年第22期90-92,共3页
Guangdong Chemical Industry