An NMR-Based Evaluation Model for Chinese Liquor Using KPCA and Cloud Model_Vol. 2 No. 4 (JLSA 2025)_Journal of Life Sciences and Agriculture (ISSN: 3005-5709)

Home > Journal of Life Sciences and Agriculture (ISSN: 3005-5709) > Vol. 2 No. 4 (JLSA 2025) >

An NMR-Based Evaluation Model for Chinese Liquor Using KPCA and Cloud Model

DOI: https://doi.org/10.62517/jlsa.202507411

Author(s)

Linlin Cheng1,*, Jia Zheng2, Xingzhong Xiong3, Mingju Chen1,3, Xingyue Zhang1,3

Affiliation(s)

1School of Automation and Information Engineering, Sichuan University of Science & Engineering, Yibin, China 2Technology Research Center, Wuliangye Yibin Co., Ltd., Yibin, China 3Intelligent Perception and Control Key Laboratory of Sichuan Province, Sichuan University of Science & Engineering, Yibin, China *Corresponding Author

Abstract

With the growing market value of Chinese liquor (Baijiu), smart technologies are increasingly being adopted in this field. Intelligent and accurate classification of Baijiu remains at an early stage in the market, representing a practical research direction that can translate time cost into economic benefits. This study employs nuclear magnetic resonance (NMR) spectroscopy combined with machine learning techniques to classify Baijiu and establish an objective evaluation model for its sensory characteristics. First, principal component analysis (PCA) and kernel principal component analysis (KPCA) were applied to reduce the dimensionality of NMR data. Point cloud models reflecting different sensory features of Baijiu were constructed in both PCA and KPCA spaces. It was observed that the point cloud model based on KPCA exhibited better aggregation density, which more effectively represents the classification outcomes of various Baijiu types. Subsequently, the ranges of point cloud regions and characteristic descriptors of Baijiu were analyzed. Using the well-classified KPCA-NMR point cloud model, regional correlations between sensory descriptors and cloud models were established. Finally, regression models-including linear, quadratic, exponential, and polynomial regression-were developed by correlating KPCA-transformed NMR data with sensory evaluation scores. Experimental results demonstrated that the KPCA-based point cloud models of different Baijiu categories are more separable in space, facilitating the distinction of their specific features. Furthermore, the integration of cloud model regions with objective evaluation terms yielded a polynomial regression model that achieved the highest correlation coefficient. This model outperformed the other three by 256.97%, 43.07%, and 238.84% in goodness-of-fit, respectively. In tests involving unknown grades of Baijiu, the proposed model achieved a classification accuracy of 95.31%, which was the closest to the results obtained from sensory evaluation.

Keywords

Chinese Liquor; Kernel Principal Component Analysis; Nuclear Magnetic Resonance Spectroscopy; Cloud Model

References

[1]Gao L, Zhou J, He G. Effect of microbial interaction on flavor quality in Chinese baijiu fermentation. Frontiers in Nutrition. 2022; 9:960712. [2]Hong J, Zhao D, Sun B. Research progress on the profile of trace components in baijiu. Food Reviews International. 2023;39(3):1666-93. [3]N. Shi ZYQ, L. Lei. Study on the Determination of Various Trace Components in Liquor by Gas Chromatography. Analysis and Testing. 2020. [4]Sandhaus D. Drunk in China: Baijiu and the world’s oldest drinking culture: U of Nebraska Press; 2019. [5]Tu W, Cao X, Cheng J, Li L, Zhang T, Wu Q, et al. Chinese Baijiu: The perfect works of microorganisms. Frontiers in Microbiology. 2022; 13:919044. [6]Wang B, Wu Q, Xu Y, Sun B. Multiple sugars promote microbial interactions in Chinese baijiu fermentation. LWT. 2021; 138:110631. [7]Fan W, Zhao X, Du G, Chen J, Li J, Zheng J, et al. Metaproteomic analysis of enzymatic composition in Baobaoqu fermentation starter for Wuliangye baijiu. International Journal of Food Science & Technology. 2021;56(8):4170-81. [8]Song Z, Du H, Zhang M, Nie Y, Xu Y. Schizosaccharomyces pombe can reduce acetic acid produced by Baijiu spontaneous fermentation microbiota. Microorganisms. 2019;7(12):606. [9]Zhao Z, Sugimachi M, Yoshizaki Y, Yin X, Han XL, Okutsu K, et al. Impact of solid‐state saccharification on the flavor of rice‐flavor baijiu. Journal of Food Science. 2021;86(11):4958-68. [10]Li J, Yang P, Zhang J, Shui Z, Huo D, Hou C, et al. A metal ion-regulated colorimetric sensor array: discriminating Chinese Baijiu from other beverages. Analytical methods. 2019;11(41):5274-81. [11]Chen S, Lu J, Qian M, He H, Li A, Zhang J, et al. Untargeted headspace-gas chromatography-ion mobility spectrometry in combination with chemometrics for detecting the age of Chinese liquor (Baijiu). Foods. 2021;10(11):2888. [12]Song X, Jing S, Zhu L, Ma C, Song T, Wu J, et al. Untargeted and targeted metabolomics strategy for the classification of strong aroma-type baijiu (liquor) according to geographical origin using comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry. Food Chemistry. 2020; 314:126098. [13]Wang Z, He M, Chen B, Hu B. Triazine covalent organic polymer coated stir bar sorptive extraction coupled with high performance liquid chromatography for the analysis of trace phthalate esters in mineral water and liquor samples. Journal of Chromatography A. 2021; 1660:462665. [14]Wu C, Xing X, Liu G, Su D, Li A, Gui S, et al. Effects of Nongxiangxing baijiu (Chinese liquor) on mild alcoholic liver injury revealed by non-target metabolomics using ultra-performance liquid chromatography quadrupole-time-of-flight mass spectrometry. Journal of bioscience and bioengineering. 2022;134(1):62-9. [15]Yan Y, Lu J, Nie Y, Li C, Chen S, Xu Y. Characterization of volatile thiols in Chinese liquor (Baijiu) by ultraperformance liquid chromatography–mass spectrometry and ultraperformance liquid chromatography–quadrupole-time-of-flight mass spectrometry. Frontiers in Nutrition. 2022; 9:1022600. [16]Hu S, Wang L. Age discrimination of Chinese Baijiu based on midinfrared spectroscopy and chemometrics. Journal of Food Quality. 2021; 2021:1-14. [17]Zhang X, Wang C, Wang L, Chen S, Xu Y. Optimization and validation of a head space solid-phase microextraction-arrow gas chromatography-mass spectrometry method using central composite design for determination of aroma compounds in Chinese liquor (Baijiu). Journal of Chromatography A. 2020; 1610:460584. [18]Zhou R, Chen X, Xia Y, Chen M, Zhang Y, Li Q, et al. Research on the application of liquid-liquid extraction-gas chromatography-mass spectrometry (LLE-GC-MS) and headspace-gas chromatography-ion mobility spectrometry (HS-GC-IMS) in distinguishing the Baiyunbian aged liquors. International Journal of Food Engineering. 2021;17(2):83-96. [19]Jun Z, Yang Y, Yao Y, Zi-wen L, Jian W, Chang-jun H. Application of mid-infrared spectroscopy in the analysis of key indexes of strong flavour Chinese spirit’s base liquor. Spectroscopy and Spectral Analysis. 2022;42(3):764-8. [20]Yadav PK, Sharma RM. Classification of illicit liquors based on their geographic origin using Attenuated total reflectance (ATR)–Fourier transform infrared (FT-IR) spectroscopy and chemometrics. Forensic science international. 2019;295: e1-e5. [21]Chen D, Tan Z, Huang Z, Lv Y, Li Q. Detection of lethal fake liquors using digitally labelled gas-phase Fourier transform infrared spectroscopy. Spectroscopy Letters. 2019;52(3-4):204-10. [22]Hou H-R, Meng Q-H, Qi P-F, Jing T. A hand-held electronic nose system for rapid identification of Chinese liquors. IEEE Transactions on Instrumentation and Measurement. 2021; 70:1-11. [23]Wu X, Zhu J, Wu B, Zhao C, Sun J, Dai C. Discrimination of Chinese liquors based on electronic nose and fuzzy discriminant principal component analysis. Foods. 2019;8(1):38. [24]Li Y, Fan S, Li A, Liu G, Lu W, Yang B, et al. Vintage analysis of Chinese Baijiu by GC and 1H NMR combined with multivariable analysis. Food Chemistry. 2021; 360:129937. [25]Zhao T, Chen S, Li H, Xu Y. Determination of linoleic acid oxylipins in Chinese Baijiu using ultra-performance liquid chromatography with quadruple-time-of-flight mass spectrometry (UPLC-QTOF-MS) and nuclear magnetic resonance (NMR). Analytical Letters. 2019;52(14):2165-79. [26]Consonni R, Cagliani L. Quality assessment of traditional food by NMR analysis. Food Control. 2022; 142:109226. [27]Consonni R, Cagliani LR. The potentiality of NMR‐based metabolomics in food science and food authentication assessment. Magnetic Resonance in Chemistry. 2019;57(9):558-78. [28]Radziej S, Scherb-Forster J, Schlicht C, Eisenreich W. NMR-based identification of thickeners in membrane-filtered food premixtures. European Food Research and Technology. 2022;248(7):1715-20. [29]Sobolev AP, Ingallina C, Spano M, Di Matteo G, Mannina L. NMR-based approaches in the study of foods. Molecules. 2022;27(22):7906. [30]Hatzakis E. Nuclear magnetic resonance (NMR) spectroscopy in food science: A comprehensive review. Comprehensive reviews in food science and food safety. 2019;18(1):189-220. [31]Sobolev AP, Thomas F, Donarski J, Ingallina C, Circi S, Marincola FC, et al. Use of NMR applications to tackle future food fraud issues. Trends in Food Science & Technology. 2019; 91:347-53. [32]Fels L, Ruf F, Bunzel M. Quantification of isomaltulose in food products by using heteronuclear single quantum coherence NMR-experiments. Frontiers in Nutrition. 2022; 9:928102. [33]He W, Bertram HC. NMR-based metabolomics to decipher the molecular mechanisms in the action of gut-modulating foods. Foods. 2022;11(17):2707. [34]Shumilina E, Johansen T, Glasner C, Robert J, Rustad T, Dikiy A. Application of NMR spectroscopy and conventional analytical methods for the assessment of wastewater from food industry. Waste and biomass valorization. 2020; 11:1349-57. [35]Viskić M, Bandić LM, Korenika A-MJ, Jeromel A. NMR in the service of wine differentiation. Foods. 2021;10(1):120. [36]Gougeon L, Da Costa G, Guyon F, Richard T. 1H NMR metabolomics applied to Bordeaux red wines. Food Chemistry. 2019; 301:125257. [37]Cassino C, Tsolakis C, Gulino F, Vaudano E, Osella D. The effects of sulphur dioxide on wine metabolites: new insights from 1H NMR spectroscopy based in-situ screening, detection, identification and quantification. Lwt. 2021; 145:111296. [38]Luan X, Shen G, Zheng Y, Feng J, Dong J. Compositional Quantitation and Brand Identification of Beer via NMR Approach Combined with Multivariate Statistical Analysis. Liquor-Making Science & Technology. 2016;(05):48-53. doi: 10.13746/j.njkj.2016043. [39]Liu E-m, Li X-y, Zong X-y, Li L, Wang S-f. Detection of alcohol content in Chinese Baijiu based on 1H nuclear magnetic resonance spectrum. 2021. [40]Masetti O, Sorbo A, Nisini L. NMR tracing of food geographical origin: The impact of seasonality, cultivar and production year on data analysis. Separations. 2021;8(12):230. [41]Riley IM, Nivelle MA, Ooms N, Delcour JA. The use of time domain 1H NMR to study proton dynamics in starch‐rich foods: A review. Comprehensive Reviews in Food Science and Food Safety. 2022;21(6):4738-75. [42]Schmitt C, Schneider T, Rumask L, Fischer M, Hackl T. Food profiling: determination of the geographical origin of walnuts by 1H NMR spectroscopy using the polar extract. Journal of agricultural and food chemistry. 2020;68(52):15526-34. [43]Peng H, Li L, Wu J, Sheng X, Rao J, Wang X, et al. Classification method of Strong-flavor Baijiu base liquor based on electronic tongue technique. China Brewing. 2022;41(03):158-62. [44]Qian Y, Hu X, Sun Y, Li J, Huang Z, Li D, et al. Classification of strong-flavor Baijiu based on fingerprint and stoichiometry. China Brewing. 2021;40(06):152-6. [45]Xin L, Qiang H, Yong-shuai Z, Xian-guo T. Research on the application methods of classifying and recognizing Baijiu wine based on deep learning. Food and Machinery. 2021;37(4):68-71, 9.