傅立叶变换近红外光谱法快速评价涪陵榨菜品质

摘要

应用傅立叶变换近红外光谱技术,建立了评价涪陵榨菜品质的定量分析模型.测定了58份涪陵榨菜的近红外光谱数据,通过光谱预处理方法消除噪声,以偏最小二乘法(PLS)建立回归模型.最终得到评价其品质的水分、总酸(以乳酸计)和氨基酸含量近红外光谱分析模型的决定系数(R2)依次为0.957 8、0.975 4、0.950 4,交叉验证均方差(RMSECV)依次为0.107、0.017 1、0.036 3.用该模型对18份未知涪陵榨菜样本进行外部验证,其水分、总酸和氨基酸外部验证的决定系数依次为0.956 2、0.953 9、0.946 0,预测集标准偏差(RMSEP)依次为0.256、0.034 7、0.038 8.内部交叉验证和外部验证均证明,近红外定量分析有较高的准确度,能满足生产中对涪陵榨菜品质检测的精度要求.%In this paper, the evaluation of moisture, total acid( lactic acid) and amino acid in Fuling mustard tuber by Fourier transform near-infrared spectroscopy (FT- NIRS)was studied. The original spectra of 58 calibration samples of Fuling mustard tuber were obtained using Fr - NIRS machine, and processed with math analysis such as first derivative, straight line subtraction, multiplicative scattering correction and min -max normalization to decrease the noise. The partial least squares(PLS) calibration models were established under the optimal conditions to predict the content of moisture, total acid ( lactic acid) and amino acid. Determination coefficient ( R2 ), root-meansquare error of cross-validation (RMSECV) and root-mean-square error of prediction (RMSEP) were used to evaluate the quality of the model. The R2 and RMSECV of the model for moisture, total acid and amino acid were O. 957 8, O. 975 4, O. 950 4, and 0. 107, O. 017 1, O. 036 3, respectively,while the R2 and RMSEP for them three were 0. 956 2, O. 953 9, 0. 946 0 and O. 256, 0. 034 7,O. 038 8 in test set validation, respectively. By statistical significance test, the determination results were compared with those of standard methods without significant difference at 0. 05 level. Internal and external cross-certification test showed that near-infrared quantitative analysis could meet the requirement of prediction of moisture, total acid and amino acid content of Fuling mustard tuber in production with high accuracy.