基于组合算法的油类污染物三维荧光光谱分析
|
摘要
针对油类污染物成分复杂、光谱重叠难以识别的问题,提出三维荧光光谱结合组合算法(ACM)。将交替三线性分解(ATLD)、自加权交替三线性分解(SWATLD)与平行因子分析(PARAFAC)算法组合,实现3种算法的优势互补。通过配制以四氯化碳为溶剂的不同质量浓度的柴油、汽油和煤油的混合溶液,利用F-7000荧光光谱仪测量混合溶液的三维荧光光谱,采用空白扣除法与缺损数据修复——主成分分析法进行预处理消除散射干扰,对三维光谱数据矩阵进行分解,并与以上3种算法解析结果进行对比。结果表明,ACM对组分数不敏感,且解析结果更准确,样本中对柴油、汽油和煤油的平均回收率分别为96.68%、97.83%、97.11%。实现了混合油类物质的定性、定量分析,具有一定的普适性。
This study proposes a new technique that combines three-dimensional fluorescence spectra with algorithm combination methodology(ACM)to address issues associated with complex components of oil pollutants and the difficulty in identifying their overlapping spectra.By combining alternating trilinear decomposition(ATLD),selfweighted alternating trilinear decomposition(SWATLD),and parallel factor analysis(PARAFAC),ACM realizes the complementary advantages of using three algorithms.First,using carbon tetrachloride as the target contaminant,a three-component mixed solution of diesel,gasoline,and kerosene with different concentrations are prepared.Then,the three-dimensional fluorescence spectra of the mixed solution are measured using a F-7000 fluorescence spectrometer.Blank deduction and missing data recovery-principal component analysis are then employed as pretreatment methods to eliminate the scattering interference.Finally,ACM is used to decompose the three-dimensional spectral data matrix.Results are compared with the three separate algorithms for component analysis,revealing that ACM is insensitive to component concentration.The average recoveries for diesel,gasoline,and kerosene are 96.68%,97.83% and 97.11%,respectively,which further indicated that this method is more universal and can be used for the qualitative and quantitative analyses of contaminants in oil mixtures.
This study proposes a new technique that combines three-dimensional fluorescence spectra with algorithm combination methodology(ACM)to address issues associated with complex components of oil pollutants and the difficulty in identifying their overlapping spectra.By combining alternating trilinear decomposition(ATLD),selfweighted alternating trilinear decomposition(SWATLD),and parallel factor analysis(PARAFAC),ACM realizes the complementary advantages of using three algorithms.First,using carbon tetrachloride as the target contaminant,a three-component mixed solution of diesel,gasoline,and kerosene with different concentrations are prepared.Then,the three-dimensional fluorescence spectra of the mixed solution are measured using a F-7000 fluorescence spectrometer.Blank deduction and missing data recovery-principal component analysis are then employed as pretreatment methods to eliminate the scattering interference.Finally,ACM is used to decompose the three-dimensional spectral data matrix.Results are compared with the three separate algorithms for component analysis,revealing that ACM is insensitive to component concentration.The average recoveries for diesel,gasoline,and kerosene are 96.68%,97.83% and 97.11%,respectively,which further indicated that this method is more universal and can be used for the qualitative and quantitative analyses of contaminants in oil mixtures.
引文
[1]Azcarate S M,Teglia C M,Karp F,et al.A novelfast quality control strategy for monitoring spoilageon mayonnaise based on modeling second-order front-face fluorescence spectroscopy data[J].Microchemical Journal,2017,133:182-187.
[2]Wang J,Zhang F, Wang X P,et al. Three-dimensional fluorescence characteristics by parallelfactor method coupled with self-organizing map andits relationship with water quality[J].Acta OpticaSinica,2017,37(7):0730003.王娟,张飞,王小平,等.平行因子法结合自组织映射神经网络的三维荧光特征及其与水质的关系[J].光学学报,2017,37(7):0730003.
[3]Wang Y T,Liu T T,Liu L F,et al.Determinationof polycyclic aromatic hydrocarbons in water based onthree dimensional fluorescence spectroscopy combinedwith wavelet compression and APTLD[J].Spectroscopy and Spectral Analysis,2018,38(4):1171-1177.王玉田,刘婷婷,刘凌妃,等.基于三维荧光光谱结合小波压缩与APTLD对水中多环芳烃测定[J].光谱学与光谱分析,2018,38(4):1171-1177.
[4]Xu B Y,Ye Y,Ruan R Y,et al.Determination ofillegal drugs in urine using excitation-emission matrixfluorescence coupled with second-order calibration[J].Chemical Journal of Chinese Universities,2015,36(9):1667-1673.徐布一,叶懿,阮若云,等.三维荧光光谱结合二阶校正算法用于尿液中常见毒品的检测[J].高等学校化学学报,2015,36(9):1667-1673.
[5]Wu H L,Li Y,Kang C,et al.Research progress ofthree-dimensional fluorescence coupled with chemicalmultiway calibration[J]. Chinese Journal ofAnalytical Chemistry,2015,43(11):1629-1637.吴海龙,李勇,康超,等.三维荧光化学多维校正方法研究新进展[J].分析化学,2015,43(11):1629-1637.
[6]Tao C X,Ruan J,Shu S P,et al.Detection of friedoil in edible oil based on three-dimensionalfluorescence spetroscopy[J]. Chinese Journal ofLasers,2016,43(1):0115001.陶春先,阮俊,舒顺朋,等.三维荧光光谱对食用植物油中煎炸油的检测[J].中国激光,2016,43(1):0115001.
[7]Kong D M,Zhang C X,Cui Y Y,et al.Detection ofoil species in mixed oil based on alternating penaltytrilinear decomposition[J]. Acta Optica Sinica,2018,38(11):1130005.孔德明,张春祥,崔耀耀,等.基于交替惩罚三线性分解的混合油液油种成分的检测[J].光学学报,2018,38(11):1130005.
[8]Bro R.PARAFAC.Tutorial and applications[J].Chemometrics and Intelligent Laboratory Systems,1997,38(2):149-171.
[9]Wu H L,Shibukawa M,Oguma K.An alternatingtrilinear decomposition algorithm with application tocalibration of HPLC-DAD for simultaneousdetermination of overlapped chlorinated aromatichydrocarbons[J].Journal of Chemometrics,1998,12(1):1-26.
[10]Wang Y T,Liu L F,Wang S T,et al.Fluorescencespectra of aromatic hydrocarbons based on alternatingtrilinear decomposition[J].Laser &OptoelectronicsProgress,2018,55(9):093001.王玉田,刘凌妃,王书涛,等.基于交替三线性分解的芳烃类化合物荧光光谱[J].激光与光电子学进展,2018,55(9):093001.
[11]Chen Z P,Wu H L,Yu R Q.On the self-weightedalternating trilinear decomposition algorithm-theproperty of being insensitive to excess factors used incalculation[J].Journal of Chemometrics,2001,15(5):439-453.
[12]Yu Y J.Researches on theory and application ofsecond-order calibration method and its developmentof application software[D]. Changsha:HunanUniversity,2012.于永杰.化学计量学二阶校正理论及应用基础研究与实用软件开发[D].长沙:湖南大学,2012.
[13]Bro R,Kiers H A L.A new efficient method fordetermining the number of components in PARAFACmodels[J].Journal of Chemometrics,2003,17(5):274-286.
[14]Maggio R M,de la Pea A M, Olivieri A C.Unfolded partial least-squares with residualquadrilinearization:a new multivariate algorithm forprocessing five-way data achieving the second-orderadvantage. Application to fourth-order excitation-emission-kinetic-pH fluorescence analytical data[J].Chemometrics and Intelligent Laboratory Systems,2011,109(2):178-185.
[15]Du W.Missing data recovery and alternating residualtrilinearization applied to multivariate calibration andcalibration transfer[D]. Changsha:HunanUniversity,2016.杜文.多元校正及模型转移中的缺损数据重构和交替残差多线性方法研究[D].长沙:湖南大学,2016.
[2]Wang J,Zhang F, Wang X P,et al. Three-dimensional fluorescence characteristics by parallelfactor method coupled with self-organizing map andits relationship with water quality[J].Acta OpticaSinica,2017,37(7):0730003.王娟,张飞,王小平,等.平行因子法结合自组织映射神经网络的三维荧光特征及其与水质的关系[J].光学学报,2017,37(7):0730003.
[3]Wang Y T,Liu T T,Liu L F,et al.Determinationof polycyclic aromatic hydrocarbons in water based onthree dimensional fluorescence spectroscopy combinedwith wavelet compression and APTLD[J].Spectroscopy and Spectral Analysis,2018,38(4):1171-1177.王玉田,刘婷婷,刘凌妃,等.基于三维荧光光谱结合小波压缩与APTLD对水中多环芳烃测定[J].光谱学与光谱分析,2018,38(4):1171-1177.
[4]Xu B Y,Ye Y,Ruan R Y,et al.Determination ofillegal drugs in urine using excitation-emission matrixfluorescence coupled with second-order calibration[J].Chemical Journal of Chinese Universities,2015,36(9):1667-1673.徐布一,叶懿,阮若云,等.三维荧光光谱结合二阶校正算法用于尿液中常见毒品的检测[J].高等学校化学学报,2015,36(9):1667-1673.
[5]Wu H L,Li Y,Kang C,et al.Research progress ofthree-dimensional fluorescence coupled with chemicalmultiway calibration[J]. Chinese Journal ofAnalytical Chemistry,2015,43(11):1629-1637.吴海龙,李勇,康超,等.三维荧光化学多维校正方法研究新进展[J].分析化学,2015,43(11):1629-1637.
[6]Tao C X,Ruan J,Shu S P,et al.Detection of friedoil in edible oil based on three-dimensionalfluorescence spetroscopy[J]. Chinese Journal ofLasers,2016,43(1):0115001.陶春先,阮俊,舒顺朋,等.三维荧光光谱对食用植物油中煎炸油的检测[J].中国激光,2016,43(1):0115001.
[7]Kong D M,Zhang C X,Cui Y Y,et al.Detection ofoil species in mixed oil based on alternating penaltytrilinear decomposition[J]. Acta Optica Sinica,2018,38(11):1130005.孔德明,张春祥,崔耀耀,等.基于交替惩罚三线性分解的混合油液油种成分的检测[J].光学学报,2018,38(11):1130005.
[8]Bro R.PARAFAC.Tutorial and applications[J].Chemometrics and Intelligent Laboratory Systems,1997,38(2):149-171.
[9]Wu H L,Shibukawa M,Oguma K.An alternatingtrilinear decomposition algorithm with application tocalibration of HPLC-DAD for simultaneousdetermination of overlapped chlorinated aromatichydrocarbons[J].Journal of Chemometrics,1998,12(1):1-26.
[10]Wang Y T,Liu L F,Wang S T,et al.Fluorescencespectra of aromatic hydrocarbons based on alternatingtrilinear decomposition[J].Laser &OptoelectronicsProgress,2018,55(9):093001.王玉田,刘凌妃,王书涛,等.基于交替三线性分解的芳烃类化合物荧光光谱[J].激光与光电子学进展,2018,55(9):093001.
[11]Chen Z P,Wu H L,Yu R Q.On the self-weightedalternating trilinear decomposition algorithm-theproperty of being insensitive to excess factors used incalculation[J].Journal of Chemometrics,2001,15(5):439-453.
[12]Yu Y J.Researches on theory and application ofsecond-order calibration method and its developmentof application software[D]. Changsha:HunanUniversity,2012.于永杰.化学计量学二阶校正理论及应用基础研究与实用软件开发[D].长沙:湖南大学,2012.
[13]Bro R,Kiers H A L.A new efficient method fordetermining the number of components in PARAFACmodels[J].Journal of Chemometrics,2003,17(5):274-286.
[14]Maggio R M,de la Pea A M, Olivieri A C.Unfolded partial least-squares with residualquadrilinearization:a new multivariate algorithm forprocessing five-way data achieving the second-orderadvantage. Application to fourth-order excitation-emission-kinetic-pH fluorescence analytical data[J].Chemometrics and Intelligent Laboratory Systems,2011,109(2):178-185.
[15]Du W.Missing data recovery and alternating residualtrilinearization applied to multivariate calibration andcalibration transfer[D]. Changsha:HunanUniversity,2016.杜文.多元校正及模型转移中的缺损数据重构和交替残差多线性方法研究[D].长沙:湖南大学,2016.