| Anoectochilus roxburghii is a kind of precious traditional Chinese herbal medicinal that is unique to Chinese folk and used for both food and medicine.At present,the provenance of anoectochilus roxburghii on the market is mixed,the quality is uneven,and there are phenomena such as fakes.Near-infrared spectroscopy has the advantages of high efficiency,simplicity and greenness,and has been well used in the quality detection of other Chinese herbal medicines.In this study,the dried anoectochilus roxburghii of different strains and its pseudo-products of Goodyera schlechtendaliana and Ludisia discolor were used as the main research objects,and the authenticity,strain and flavonoid content of anoectochilus roxburghii were studied by near-infrared spectroscopy.On this basis,a qualitative or quantitative analysis model was constructed to realize the accurate of the authenticity and strain of Aanoectochilus roxburghii,and to provide a reference for further research on Aanoectochilus roxburghii medicinal materials.The specific research contents are as follows:(1)Quick identification of the authenticity of Anoectochilus roxburghii.By comparing the influence of different waveband ranges and different preprocessing methods on the overall performance of the established analysis model,540-670 nm and 736-850 nm were selected as the modeling wavebands.The accuracy rate is 100%.(2)Rapid identification of Anoectochilus roxburghii strains.By comparing the effects of different modeling methods,different waveband ranges and different preprocessing methods on the overall performance of the established analysis model,the SIMCA method was used to select the 540nm-670 nm waveband,and the performance of the identification model for Anoectochilus roxburghii strains established after SG9 point smoothing was the best.The classification accuracy rates of the calibration set and validation set are 98.0% and 97.3%,respectively.(3)The contents of four flavonoids,rutin,narcissin,quercetin and isorhamnetin in Anoectochilus roxburghii were simultaneously determined by HPLC,rutin and narcissus were constructed based on near-infrared spectroscopy combined with partial least squares regression.Model for quantitative analysis of glycoside,quercetin and isorhamnetin content.The results show that the four models established by partial least squares regression combined with the full spectrum band have the best prediction effect.Its Rc is 0.9637,0.9704,0.9684 and 0.9683,Rp is 0.9681,0.9825,0.9725 and 0.9749,RMSEC is 1.8024,1.6414,1.7430 and 1.7128,RMSEP is 1.8207,1.8837,1.8651 and 1.9534.(4)A real-time detection system for the authenticity and strain of Anoectochilus roxburghii based on near-infrared spectroscopy is built.The detection system is mainly composed of hardware and software.The hardware part is mainly responsible for collecting the spectral information of the samples to be detected;the software part supports the functions of spectral data reading,real-time display of spectral data,spectral data analysis and analysis result export.The robustness and stability of the system are verified by testing. |