基于三维荧光光谱快速识别冷榨 一级核桃油品牌和产地 Rapid identification of brand and origin of cold pressed first-grade walnut oil based on three-dimensional fluorescence spectroscopy

为寻找一种能快速识别核桃油品牌和产地的方法，以来自4个品牌3个产地的冷榨一级核桃油为研究对象，分批次收集了300个样品，用荧光检测仪进行样品扫描，分两个时间段采集三维荧光光谱，间隔为1个月。对收集的共600组光谱数据运用主成分分析（PCA）进行特征提取，同品牌或同产地的样品分别选取主成分1组成新的数据集，达到数据降维，再结合偏最小二乘判别（PLS-DA）和人工神经网络判别（BP-ANN）化学模式识别方法，对应构建核桃油的品牌识别模型和产地识别模型。结果表明：PLS-DA和BP-ANN对核桃油的品牌和产地的识别率都能达到100%。因此，三维荧光光谱与PLS-DA和BP-ANN方法结合，可用于快速识别核桃油的品牌和产地。In order to find a method to quickly identify brand and origin of walnut oil, cold pressed first-grade walnut oils of four brands from three origins were used as the research object, 300 samples were collected in batches, samples were scanned with a fluorometric detector and three-dimensional fluorescence spectra were collected in two time intervals of one month. A total of 600 sets of spectral data collected were extracted by PCA, and principal component 1 was selected from samples of the same brand or origin to form a new data set to achieve data dimension reduction. Then, combined with PLS-DA and BP-ANN chemical pattern recognition methods, brand recognition model and origin recognition model of walnut oil were constructed. The results showed that PLS-DA and BP-ANN could both accurately identify the corresponding brand and origin of walnut oil with the recognition rate of 100%. So the combination of three-dimensional fluorescence spectra with PLS-DA and BP-ANN can be used as a method for rapid identification of brand and origin of cold pressed first-grade walnut oil.