Abstract:

In the present study, the pattern recognition analysis of Chinese rice wine from different rice varieties was based
on diffuse reflectance fourier transform infrared spectroscopy (DR-FTIR) and soft independent modeling of class analogy
(SIMCA). The results showed that the model of SIMCA was established successfully, and the identification rates and
rejection rates of the predicted samples were 100% at the significance level of 0.05, except Chinese rice wine of glutinous
rice with the rejection ratio of 75%. The model was based on the pre-treatments of nine-point smoothing of Savitzky-Golay,
automatic baseline correction and standard normal variate (SNV) normalization in the spectral regions of 975–1165 cm-1
and 1250–1500 cm-1. It suggested that rice varieties can affect the quality of Chinese rice wine and the pattern recognition
of SIMCA was useful to identify Chinese rice wine from different rice varieties, which will provide scientific proofs for a
traceability system from Chinese rice wine to rice.

Key words: Chinese rice wine, rice, diffuse reflectance Fourier transform infrared spectroscopy, soft independent modeling of class analogy, principle component analysis