Rapid determination of process parameters during simultaneous saccharification and fermentation (SSF) of cassava based on molecular spectral fusion (MSF) features

In this study, the Raman spectroscopy technique and the near-infrared spectroscopy technique were used to obtain the spectral information of the cassava SSF process samples, respectively. The characteristics of the Raman spectra and the near-infrared spectra after preprocessed were optimized, respectively. SVM quantitative detection models were established based on the fusion optimization features to achieve rapid detection of the cassava SSF process parameters with high precision. [Display omitted] •The SSF of cassava is one of the key steps in the production of fuel ethanol.•This study proposes a new idea for process monitoring of the cassava SSF by the MSF technique.•The CARS algorithm selects the features of Raman spectra and NIR spectra.•The SVM model based on the MSF feature monitors the process of the cassava SSF. Simultaneous saccharification and fermentation (SSF) of cassava is one of the key steps in the production of fuel ethanol. In order to improve the monitoring efficiency of the ethanol production process and the product yield, this study puts forward a new idea for monitoring of the cassava SSF process based on the molecular spectroscopy fusion (MSF) technique. Savisky-Golay (SG) combined with standard normal variable (SNV) was used to preprocess the obtained Raman spectra and near-infrared (NIR) spectra. Competitive adaptive reweighted sampling (CARS) was used to optimize the characteristic wavelengths of the preprocessed Raman spectra and the NIR spectra, and the optimized features were fused in the feature layer. The support vector machine (SVM) model of the process parameters during the cassava SSF based on the MSF features was established. The experimental results showed that compared with the best CARS-SVM model based on the single-molecule spectral features, the performance of the best CARS-SVM model based on fusion features has been significantly improved. For detection of the glucose content, the RMSEP, RP2 and RPD of the best CARS-SVM model were 5.398, 0.957 and 4.922, respectively. For detection of the ethanol content, the RMSEP, RP2 and RPD of the best CARS-SVM model were 4.394, 0.977 and 6.758, respectively. The obtained results reveal that the combination of MSF technique and appropriate chemometric methods can achieve high-precision quantitative detection of the process parameters during the cassava SSF. This study can provide technical basis and experimental reference for the development of portable spectrometer equipment fo