拉曼光谱结合光谱特征区间筛选算法快速定量鉴别植物调和油品质

Rapid Quantitative Authentication of Blended Vegetable Oil Quality by Raman Spectroscopy Coupled with a Selection Algorithm of Spectral Characteristic Intervals

本研究提出了一种基于拉曼光谱与光谱特征区间筛选算法实现植物调和油中高价值植物油含量快速定量检测的方法。首先,将粒子群优化(particle swarm optimization,PSO)算法与灰狼优化(grey wolf optimization,GWO)算法融合构建混合智能优化算法,即PSOGWO算法。其次,将PSOGWO与组合移动窗口(combined moving window,CMW)策略结合构建新型的拉曼光谱特征区间筛选算法,即PSOGWO-CMW算法。然后,将玉米油(corn oil,CO)和特级初榨橄榄油(extra virgin olive oil,EVOO)以不同比例配制为CO-EVOO植物调和油,并采集其拉曼光谱。将拉曼光谱输入偏最小二乘回归、PSO-CMW、GWO-CMW和PSOGWO-CMW模型预测EVOO含量,并比较建模效果。结果表明,PSOGWO-CMW模型具有最佳的预测性能。采用本方法与气相色谱-质谱法分别检测真实的CO-EVOO植物调和油样本中EVOO含量,结果表明两者的检测性能无显著差异。本方法快速、准确,亦可用于其他植物调和油中高价值植物油含量的快速定量检测。

In this study,a method for the rapid quantitative determination of the content of high-value vegetable oil in blended edible vegetable oils(BEVO)was proposed based on Raman spectroscopy and a selection algorithm of spectral characteristic intervals.First,the particle swarm optimization(PSO)and grey wolf optimization(GWO)algorithms were combined to develop a hybrid intelligent optimization algorithm called PSOGWO.Second,the PSOGWO algorithm and the combined moving window(CMW)strategy were combined to develop a novel spectral characteristic interval selection algorithm named PSOGWO-CMW.Third,blends of corn oil(CO)and extra virgin olive oil(EVOO)at different ratios were prepared,and then their Raman spectra were acquired.Using the Raman spectra as input variables,partial least squares regression(PLSR),PSO-CMW,GWO-CMW,and PSOGWO-CMW models were developed to predict the content of EVOO,and their performance was comparatively studied.The results showed that the PSOGWO-CMW model had the best prediction performance.The results of the proposed method for the content of EVOO in CO-EVOO blends were not significantly different from those of gas chromatography-mass spectrometry.In conclusion,this method is rapid and accurate,and can be used for rapid and quantitative determination of the content of high-value vegetable oil in BEVO.