摘要
用近红外光谱预测汽油辛烷值的方法，将辛烷值机ASTM－CFR测定的汽油辛烷值数据与汽油的近红外光谱数据用多重线性回归（MLR）和偏最小二乘法回归（PLS）等化学计量方法关联，建立了汽油辛烷值的3个预测模型，可用来预测催化裂化汽油、重整汽油（包括烷基化汽油）和二者与直馏汽油的调合油的马达法辛烷值（MON）、研究法辛烷值（RON）和抗爆指数（泵浦法辛烷值PON），预测均方误差为0．7个辛烷值单位。可与ASTM-CFR相应方法的结果比较。将根据预测模型发展的软件安装于北京第二光学仪器厂试制的付立叶变换近红外光谱仪上，首次研制成功了FT-NIR辛烷值仪。测量所需汽油试样为3mL，测定时间少于3min。
Based on more than 170 gasoline samples which MON, RON and PON were evaluat-ed by ASTM-CFR engine in advance, three predictive models for octane numbers of catalytic crackinggasoline, reforming gasoline and their mixtures with straight-run gasoline were developed by using MLRand PLS chemometrics. The mean standard deviations of the forecasted values were all less than 0. 7 unit.Three predictive softwares for MON, RON and PON of gasolines were established and an octane numberanalyzer was developed by using RFX-300N FT-NIR spectrometer manufactured by BSOIF as the hardware. The forecasted RON, MON and PON of the unknown gasolines can be obtained simultaneously byusing 3 ml samples within three minutes with satisfactory accuracy compared with ASTM method.
出处
《石油炼制与化工》
EI
CAS
CSCD
北大核心
1997年第1期22-27,共6页
Petroleum Processing and Petrochemicals
