中药甘青青兰提取过程的近红外及紫外可见双光谱在线监测OA北大核心CSTPCD

An online monitoring system of near infrared spectrum(NIR)and ultraviolet-visible spec-troscopy(UV-Vis)is constructed using small optical fiber spectrometers,and the spectra are online collected with the system.After a spectrum collection,the sample is sent back into the extraction container to avoid the loss of the extraction liquid,and the pipeline is cleaned with the backward flow to prevent the pipeline from clogging.The extraction process is monitored by the dual-spectrum sys-tem to improve the acquisition of spectral information.Sampling error profile analysis(SEPA)and sta-bility competitive adaptive reweighted sampling(SCARS)and other chemometric methods are used for the spectral process and model establishment.The analytical performances of NIR spectroscopy,UV-Vis spectroscopy and dual-spectrum fusion models are compared.The results show that the NIR spectral model of luteolin and the UV-Vis model of chlorogenic acid are better.The predicted root-mean-square error(RMSEP)of luteolin model is 0.0107 µg·mL-1,and that of chlorogenic acid model is 0.0295 mg·mL-1.The system developed in this study is simple,rapid and has high model accuracy,which provides a new strategy for online monitoring of Chinese medicine extraction.