近红外光谱结合小波变换预处理在蜂蜜产地快速检测中的应用

应用近红外光谱结合化学计量学方法对蜂蜜产地进行了判别分析。kennard-Stone法划分训练集和预测集。光谱用一阶导数加自归一化预处理后，再用小波变换（WT）进行压缩和滤噪。结合滤波后光谱信息，分别用径向基神经网络（RBFNN）和偏最小二乘-线性判别分析（PLS-LDA）建立了苹果蜜产地和油菜蜜产地判别模型。对不同小波基和分解尺度进行了详细讨论。对苹果蜜，WT-RBFNN模型和WT-PLS-LDA模型都是小波基为db1、分解尺度为2时的预测精度最好，都为96.2%。对油菜蜜：WT-RBFNN模型在小波基为db4和分解尺度为1时，预测精度最好；WT-PLS-LDA模型在小波基为db9、分解尺度也为1时，预测精度最好，为90.5%；预测精度WT-PLS-LDA模型优于WT-RBFNN模型。研究表明：WT结合线性的PLS-LDA建模比WT结合非线性的RBFNN建模更适于蜂蜜产地鉴别；近红外光谱结合WT-PLS-LDA可实现对蜂蜜产地的快速无损检测，为蜂蜜产地鉴别提供了一种新方法。

Application of Near Infrared Spectroscopy Combined Wavelet Transform Pretreatment in Geographical Classification of Honey

Near infrared spectroscopy combined with chemometrics methods has been used to determine the geographical origin of honey samples. The samples were divided into the training set and test set by kennard-Stone algorithm. After being pre-treated with first derivative and autoscaling, the spectral data were compressed and de-noised using wavelet transform (WT). The radical basis function neural networks (RBFNN) and partial least squares-line discriminant analysis (PLS-LDA) were applied to develop classification models of the geographical origin of honey using the reconstructed signals, respectively. The performances of different wavelet functions and decomposition levels were evaluated in relation to the total prediction accuracy. For apple honey, when wavelet function was db1 and decomposition level was 2, both WT-RBFNN and WT-PLD-LDA produced the best total prediction accuracy of 96.2%. For rape honey, when wavelet function was db4 and decomposition level was 1, WT-RBFNN produced the best total prediction accuracy; while when wavelet function was db9 and decomposition level was also 1, WT-PLD-LDA produced the best total prediction accuracy of 90.5%; WT-PLD-LDA was better than WT-RBFNN for total prediction accuracy. The results indicated that linear WT-PLS-LDA was more suitable for geographical classification of honey than no-linear WT-RBFNN and near infrared spectroscopy combined with WT-PLS-LDA could be applied to quickly detect geographical classification of honey, which developed a new method of geographical classification of honey.