论文部分内容阅读
以枸杞干果为研究对象,将三维荧光光谱技术与平行因子分析法、BP神经网络相结合,建立枸杞定性鉴别模型。采用固体样品支架测得枸杞粉末样品的三维荧光光谱,利用平行因子分析方法对预处理后的三维荧光矩阵进行三线性分解得到2个主因子的浓度得分,然后将浓度得分作为BP神经网络的输入向量,建立枸杞的人工神经网络鉴别模型。利用所建模型对待测样品进行预测,预测正确率为100%。结果表明,平行因子分析结合BP神经网络建立的枸杞产地鉴别模型,能够快速准确地鉴别宁夏枸杞。
Taking wolfberry dried fruit as research object, the three-dimensional fluorescence spectroscopy was combined with parallel factor analysis and BP neural network to establish the qualitative identification model of wolfberry. The three-dimensional fluorescence spectra of Lycium barbarum powder samples were measured by solid sample scaffolds. The three-dimensional fluorescence matrixes were pretreated by parallel factor analysis to obtain the concentration scores of the two main factors. Then, the concentration score was used as the input of BP neural network Vector, the establishment of Chinese wolfberry artificial neural network identification model. Using the model to predict the sample, the prediction accuracy is 100%. The results show that parallel factor analysis combined with BP neural network established wolfberry origin identification model, can quickly and accurately identify Ningxia wolfberry.