基于近红外光谱快速检测杏鲍菇多糖含量的研究

Rapid Detection of Polysaccharide Content of Pleurotus eryngii Based on Near Infrared Spectroscopy

为建立杏鲍菇多糖含量的近红外光谱(NIR)快速检测方法,采用蒽酮-硫酸法和分光光度计检测杏鲍菇多糖含量,得到多糖含量的吸光度值。利用傅立叶变换近红外透射光谱技术采集杏鲍菇多糖含量的近红外光谱图,通过平滑处理、导数处理、散射校正和小波转换的方法对近红外光谱进行预处理以消除光谱噪音的影响,结合最小二乘支持向量回归(LSSR)、偏最小二乘回归(PLSR)和人工神经网络模型(ANN)建立多糖的定量校正模型。得出最佳模型是近红外光谱结合小波变化-径向基神经网络(RBF-NN)模型,结果为Rc^2=1,Rv^2=0.890 075,RMSEC=9.02E-10,RMSEP=7.728 933,REC=7.63E-12,REP=0.086 268。研究建立的近红外模型为快速检测杏鲍菇多糖含量提供了初步依据。

In order to establisha rapid detection method of Pleurotus eryngii polysaccharide content with near infrared spectroscopy （NIR） , the chemical content of polysaccharide was first detected by anthrone-sulfuric acid method, the NIRs of Pleurotus eryngii was scanned using fourier transform near infrared spectrometer and the spectra was later pretreated by methods of smoothing, derivative, scatter correction and wavelet transform afterward,finally the mathematical model of chemical content and NIR of polysaccharide was established using Least squares supp（RBF-NM）model. The best model was RBF-NN model, and the parameters were calculated： Rc^2=1, Rv^2=0.890 075, RMSEC = 9.02E-10, RMSEP =7.728 933, REC = 7.63E-12, REP = 0.086 268. The near infrared modele stablished in the present study provides preliminary basis for the rapid detection of Pleurotus eryngii polysaccharide content