摘要
基于280余种典型的原油,采用衰减全反射测量方式结合偏最小二乘(PLS)方法建立红外光谱测定原油密度的校正模型。校正结果显示,交互验证标准偏差为0.005 1 g/cm3,预测标准偏差为0.004 4 g/cm3,均接近于标准方法的误差要求。与U形振动管法相比,该方法具有测量速度快、操作简便等优点,其结果满足快速分析和过程控制分析的需要,在一定场合可以替代U形振动管法用于原油密度的快速预测。
The combined use of mid-infrared attenuated total reflection spectroscopy and multivari- ate calibration to determine the density of crude oil was carried out. Calibration models for crude oil density were established by partial least squares（PLS） based on 280 crude oil samples collected worldwide. Results show that the standard deviation of cross-verification is 0. 005 1 g/cm^3 and the predicted standard deviation is 0. 004 4 g/cm^3 ,which are all close to the requirements of standard method,indica- ting that the results predicted by this method are very close to those determined by standard method. As compared with the U-tube vibration method this method is quick and easy to operate, its result could meet the requirements of fast analysis/process control analysis, and in certain cases it could substitute the U-tube method for fast prediction of crude oil density.
出处
《石油炼制与化工》
CAS
CSCD
北大核心
2011年第12期73-77,共5页
Petroleum Processing and Petrochemicals
关键词
原油
密度
偏最小二乘
红外光谱
快速评价
crude oil
density
partial least squares
infrared spectroscopy
fast evaluation
