Study on the Applicability of NIR Combined with LS-SVM Used for the Quality Monitoring of Medicated Leaven Fermentation / 世界科学技术-中医药现代化

ABSTRACT

This article investigated the applicability of the near infrared spectroscopy (NIR) technology, combined with the least squares support vector machines (LS SVM) used for the quality monitoring of medicated leaven fermentation. First, near infrared spectra of 67 medicated leaven samples were obtained by near infrared spectroscopy system in the wavelength range of 400 2 500 nm, and then the protease and amylase activity were measured by Folin phenol method and DNS method. Thereafter, the LS SVM was employed to calibrate models. The Rc and Rp of protease in near infrared model were 0.975 and 0.938, respectively; The RMSEC and RMSEP were 5.297 and 9.795, respectively. The Rc and Rp of amylase in near infrared model were 0.987 and 0.973, respectively; The RMSEC and RMSEP were 7.215 and 6.864, respectively. This model has good prediction ability and is suitable for quality monitoring in medicated leaven fermentation process. The research achievement could lay a certain foundation for the near infrared spectral analysis technology applied in the field of traditional fermentation processing.