摘要
为快速测定S Zorb再生剂中非活性含锌尖晶石物相(硅锌矿和锌铝尖晶石)的含量,首次运用偏最小二乘(PLS)化学计量学的方法,将XRD谱图信息与Rietveld物相定量方法获得的物相含量数据相关联,建立了针对SZorb再生剂中含锌尖晶石物相的定量校正模型,优化了影响模型质量的参数如谱图预处理方法和特征衍射峰的选取,并用未知样品验证了模型的准确性。此方法解决了利用Rietveld方法进行S Zorb吸附剂定量谱图解析时耗时较长、且依赖专家经验的问题,有助于及时监测装置运行过程中吸附剂的活性变化,进而保障工业装置平稳运行。
For fast determination of non-activated gahnite （zinc aluminate and Willemite） content in S Zorb regenerated sorbent, quantitative calibration models were first established by partial least square （PLS） regression based on the data of XRD combined with information from Rietveld quantitative phase analysis （RPA） method. Parameters affecting the models were optimized, such as pretreatment methods of XRD patterns and choices of characteristic peaks. The model correlations were analyzed by using validation set samples. With the obtained models, the problems that RPA method took long time to analyse S Zorb sorbent and relied on expert experience could be solved, which will be helpful for monitoring the changes of sorbent activity in online operation and then ensuring an effective and stable running of S Zorb unit.
出处
《石油学报（石油加工）》
EI
CAS
CSCD
北大核心
2013年第3期453-458,共6页
Acta Petrolei Sinica(Petroleum Processing Section)
基金
中国石化科研开发课题项目(112008)资助
