ABSTRACT
The grey correlation-TOPSIS method was used to evaluate the quality of the origin herbs of Lonicerae Japonicae Flos, and the Fourier transform near-infrared(NIR) and mid-infrared(MIR) spectroscopy was applied to establish the identification model of origin herbs of Lonicerae Japonicae Flos by combining chemometrics and spectral fusion strategies. The content of neochlorogenic acid, chlorogenic acid, cryptochlorogenic acid, caffeic acid, secoxyloganin, isoquercitrin, isochlorogenic acid B, isochlorogenic acid A, and isochlorogenic acid C in six origin herbs of Lonicerae Japonicae Flos was determined by high-performance liquid chromatography(HPLC), and their quality was evaluated by the grey correlation-TOPSIS method. The Fourier transform NIR and MIR spectra of six origin herbs of Lonicerae Japonicae Flos(Lonicera japonica, L. macranthoides, L. hypoglauca, L. fulvotomentosa, L. confuse, and L. similis) were collected. At the same time, principal component analysis(PCA), support vector machine(SVM), and spectral data fusion technology were combined to determine the optimal identification method for the origin herbs of Lonicerae Japonicae Flos. There were differences in the quality of the origin herbs of Lonicerae Japonicae Flos. Specifically, there were significant differences between L. japonica and the other five origin herbs(P<0.01). The quality of L. similis was significantly different from that of L. fulvotomentosa, L. macranthoides, and L. hypoglauca(P=0.008, 0.027, 0.01), and there were also significant differences in the quality of L. hypoglauca and L. confuse(P=0.001). The PCA and SVM 2D models based on a single spectrum could not be used for the effective identification of the origin herbs of Lonicerae Japonicae Flos. The data fusion combined with the SVM model further improved the identification accuracy, and the identification accuracy of the mid-level data fusion reached 100%. Therefore, the grey correlation-TOPSIS method can be used to evaluate the quality of the origin herbs of Lonicerae Japonicae Flos. Based on the infrared spectral data fusion strategy and SVM chemometric model, it can accurately identify the origin herbs of Lonicerae Japonicae Flos, which can provide a new method for the origin identification of medicinal materials of Lonicerae Japonicae Flos.