摘要
用探头式近红外光谱仪(NIR)在线测量工业硝基苯的纯度,选择甲苯为模型杂质,基于实验设计,配制纯度99%~100%的硝基苯样品,使用气相色谱法离线标定样品纯度数据,使用偏最小二乘法(PLS)研究了基于NIR的硝基苯样品的纯度预测模型。为克服近红外光谱中的合频、倍频以及基线漂移等干扰信息,研究了五种数据预处理方法对纯度预测模型准确性的影响。选择标准正态变量校正预处理方法(SNV),其中R^(2)_(Cal)为0.9974,R^(2)_(CV)为0.9968,R^(2)_(Pre)为0.9955,校正均方根误差(RMSEC)为0.0002,交叉验证均方根误差(RMSEVCV)为0.0002,预测均方根误差(RMSEP)为0.0002,相对偏差在0.05%以内。实现了对含甲苯杂质的硝基苯样品纯度的准确在线测量和质量控制,相比以往的气相色谱离线检测法,有效提高了检测速度,避免了制样对测量准确性的影响,能帮助实现工业硝基苯输送过程连续化,具有重要的工业应用潜力。
The purity of industrial nitrobenzene was measured online by NIR.The nitrobenzene was selected as the research object,and the concentration online measurement method bases on NIR.The model impurity is methylbenzene.Nitrobenzene solution with purity of 99%~100%is prepared.Gas chromatography is used to develop off-line calibration of sample concentration data.The models for prediction of concentration of nitrobenzene samples based on NIR are built by partial least squares method(PLS).In order to overcome the interference information such as combined-frequency,frequency-doubling and baseline drift in NIR spectra,five data preprocessing methods on the accuracy of the concentration prediction model are researched.The standard normalized variate method(SNV)is selected,and R^(2)_(Cal) is 0.9974,R^(2)_(CV) is 0.9968,R^(2)_(Pre) is 0.9955,root mean square error of calibration(RMSEC)is 0.0002,root mean square error of cross validation(RMSEVCV)is 0.0002,root mean square error of prediction(RMSEP)is 0.0002.The relative deviation is less than 0.05%.Compared with the off-line gas chromatography detection method,NIR method effectively improves the detection speed and avoids the influence of sample preparation on the accuracy of measurement.The results would help realize the continuous transport process of industrial nitrobenzene,which is importance of industrial application potential.
作者
李俊杰
王学重
王焱宇
张扬
LI Jun-jie;WANG Xue-zhong;WANG Yan-yu;ZHANG Yang(School of Chemistry and Chemical Engineering,South China University of Technology,Guangzhou 510640,China;Beijing City Key Laboratory of Enze Biomass Fine Chemicals,College of New Materials and Chemical Engineering,Beijing Institute of Petrochemical Technology,Beijing 102627,China)
出处
《应用化工》
CAS
CSCD
北大核心
2022年第1期277-280,共4页
Applied Chemical Industry
基金
国家自然科学基金重点项目(61633006)
广东省应用型科技研发资金项目(2015B020232007)
广东省自然科学基金项目(2017A030310262)
广东省自然科学基金面上项目(2018A030313263)
中央高校基本科研业务费资助项目(2017MS092)
北京石油化工学院恩泽生物质精细化工北京市重点实验室开放课题
青海省科技厅重点研发与转化计划项目(2021-GX-C10)。
关键词
硝基苯
质量控制
近红外光谱
偏最小二乘法
nitrobenzene
quality control
near infrared spectrum
partial least squares
