A novel combination of LF-NMR and NIR to intelligent control in pulse-spouted microwave freeze drying of blueberry

Abstract

In order to improve the drying efficiency while maintaining product quality, pulse-spouted microwave freeze-drying (PS-MFD) was applied to the dehydration process of blueberries. The present study was carried out to establish a strategy for intelligently determine sublimation/desorption drying transition point (TPS-A) and drying endpoint of blueberry during PS-MFD process. Back-propagation artificial neural network (BP-ANN) model was used to establish the relationship between low field nuclear magnetic resonance (LF-NMR) spectrum and moisture content (MC) as well as between near infrared (NIR) spectrum and MC. Then the two models were applied to predict the moisture content of blueberry during PS-MFD process. Model fitting results showed that BP-ANN models based on LF-NMR spectrum and NIR spectrum could accurately predict the moisture content of blueberries. BP-ANN model of LF-NMR-MC couldn't capture the signal of free water during the early stage of PS-MFD (about 90 min), while BP-ANN model of NIR-MC had moisture content output values of blueberries during the whole PS-MFD process. The TPS-A and drying endpoint of blueberry during PS-MFD process could be accurately determined by comparative analysis of PS-MFD blueberry samples moisture content output values obtained from two models.