研究目的
To construct a prediction model for moisture content and pH value changes during the solid fermentation of Monascus using near-infrared spectroscopy combined with the interval least squares support vector machine method (siLS-SVM).
研究成果
The LS-SVM model established by siLS-SVM algorithm has good predictability and robustness for the determination of water and pH in the fermentation process of Monascus. Near-infrared spectroscopy can effectively and quickly detect the changes of water content and pH value during the solid fermentation of Monascus.
研究不足
The study focuses on the solid-state fermentation of Monascus under bran substrate, and the applicability of the model to other substrates or fermentation processes is not explored. The spectral preprocessing methods and model optimization techniques may require further validation for broader applications.
1：Experimental Design and Method Selection:
Near-infrared spectroscopy (NIRS) was used to detect the changes of water content and PH during solid-state fermentation of Monascus under bran substrate. The LS-SVM model was established by siLS-SVM algorithm for quantitative analysis.
2：Sample Selection and Data Sources:
Five batches of culture medium were prepared, with the first four batches used to construct the calibration set and the predicted sample of the infrared model, and the 5th batch as the independent sample.
3：List of Experimental Equipment and Materials:
Instruments included a Thermo Scientific Nicolet iS5 FT-IR spectrometer, among others. Reagents included protein Chen, sodium nitrate, magnesium sulfate, potassium dihydrogen phosphate, etc.
4：Experimental Procedures and Operational Workflow:
Spectral sampling was completed on the 5 batches of culture medium. The spectral acquisition method used was the diffuse reflection method. The spectrum for each sample is the average of the spectra of the three positions equidistant from the center of the flask.
5：Data Analysis Methods:
The PLS model was used for quantitative analysis, and the siLS-SVM algorithm was introduced to optimize the prediction model.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
