采用近红外光谱技术结合化学计量学方法对菜籽油中多效唑残留进行定性检测。在4000~10000 cm-1光谱范围内采集126个菜籽油样本的近红外透射光谱。对原始光谱进行初步分析后,分别采用线性判别分析（LDA）、簇类独立软模式法（SIMCA）和最小二乘支持向量机（LSSVM）三种不同方法建立菜籽油中多效唑残留的定性检测模型,并对不同多效唑残留的菜籽油样本的分类正确率进行分析。研究结果表明,LDA,SIMCA及LSSVM 3种方法建立的检测模型均具有较高的判别能力,其校正集和预测集的正确率分别为93.33%,91.11%,95.56%和86.11%,88.89%,83.33%。此外,高多效唑残留样本的分类正确率大致趋于100%,而低多效唑残留样本的分类正确率则有一定波动。由此可知,利用近红外光谱技术可对菜籽油中多效唑残留进行快速、无损的定性检测。
Qualitative detection paelobutrazol residue was carried out in the study in rapeseed oil based near infrared spectroscopy combined with chemometrics methods. Near infrared transmittance spectra of 126 rapeseed oil samples were collected in spectral range from 4000 to 10000 cm1. After a preliminary analysis of original spectra, 3 different methods, including linear discriminant analysis （LDA）, soft independent modeling of class analogy （SIMCA） and least squares support vector machines （LSSVM）, were established for qualitative detection models of paclobutrazol residue in rapeseed oil. The classification accuracy of different samples of paclobutrazol residue in rapeseed oil was analyzed. The results showed that there has high distinguished ability in the detection models which used LI）A, SIMCA and LSSVM, respectively. And the correct rate of calibration set and validation set were 93. 33%, 91.11%, 95. 56% and 86. 11%, 88. 89%, 83. 33%, respectively. In addition, the classification accuracy of high paclobutrazol residue samples was approximately tended to 100%. However, the classification accuracy of low paclobutrazol residue samples has certain wave. Therefore, near infrared spectroscopy can be used in qualitative detection of paclobutrazol residue in rapeseed oil rapid and nondestructive.
Chinese Journal of Analysis Laboratory
near infrared spectroscopy