Method for predicting quality of yellow wine

Abstract

The invention relates to the technical field of data models, in particular to a yellow rice wine quality prediction method, which comprises the following steps: after establishing a yellow rice wine sensory index evaluation table, respectively carrying out sensory analysis on Fujian yellow rice wine in different years to obtain a comprehensive taste actual score; carrying out correlation and significance analysis on physical and chemical components and sensory indexes of the Fujian yellow wine; performing correlation and significance analysis on the color and sensory indexes of the Fujian yellow wine; performing correlation and significance analysis on volatile matter indexes and sensory indexes of the Fujian yellow wine; selecting parameters with significant levels of less than 0.05 in physical and chemical components, color and volatile matters as independent variables, and establishing a sensory comprehensive evaluation and prediction model by taking the comprehensive taste actual score as a dependent variable and adopting stepwise regression analysis; according to the method, the Fujian yellow rice wine in different years is researched, the quality of the Fujian yellow rice wine is evaluated through sensory analysis, electronic nose analysis and physical and chemical analysis, sensory indexes are quantified, datamation indexes for directly evaluating the quality of the Fujian yellow rice wine are obtained, and management and grading are facilitated.

Description

Technical Field

[0001] This invention relates to the field of data modeling technology, specifically to a method for predicting the quality of rice wine. Background Technology

[0002] Yellow wine is one of the oldest alcoholic beverages in the world, ranking alongside beer and wine as one of the three ancient wines. Fujian-style yellow wine is a significant representative of Chinese yellow wine. Pingnan yellow wine from Fujian Province is a prime example of this style and was designated a Chinese geographical indication product in 2008. Fujian-style yellow wine is rich in amino acids, sugars, vitamins, minerals, polyphenols, and other nutrients and active ingredients, giving it certain nutritional and health benefits. Studies have shown that moderate consumption of yellow wine has a significant effect on maintaining health and prolonging life. The sensory indicators of yellow wine include flavor quality, texture quality, physicochemical quality, and appearance quality, among which physicochemical quality and sensory indicators are crucial. However, the current quality evaluation system for Fujian-style yellow wine is relatively weak. Therefore, it is necessary to select quality evaluation indicators and establish evaluation methods to facilitate the selection of high-quality Fujian-style yellow wine.

[0003] Due to its unique raw materials and complex brewing process, Fujian-style rice wine possesses a variety of tastes, including sweet, bitter, umami, astringent, and sour, as well as rich aromas such as mellow, fruity, fermented, and caramel notes. However, the current evaluation of the quality of Fujian-style rice wine mainly relies on sensory analysis, which is significantly influenced by subjective factors. Therefore, instrumental testing technology is the primary means of quantifying the flavor and quality of rice wine, offering higher sensitivity and objectivity compared to sensory analysis. Electronic nose technology, an emerging technology based on bionics that mimics human olfaction, primarily analyzes the concentration and composition of volatile compounds in the air through a sensor network. Widely applicable to food quality assessment, electronic noses have become an important tool for testing the quality of alcoholic beverages due to their speed, accuracy, and ease of operation. Furthermore, the physicochemical content of Fujian-style rice wine is also a crucial indicator of its quality. Studies have found a correlation between total sugar, total acid, alcohol content, pH, and color and the flavor and quality of Fujian-style rice wine. The concentration of most aroma compounds increases with storage time, resulting in better sensory scores. Therefore, exploring the relationship between sensory indicators, electronic nose indicators, and physicochemical components helps to determine and present the sensory indicators of Fujian-style rice wine through electronic nose and physicochemical component data, which is crucial for the establishment of a sensory quality evaluation data system for Fujian-style rice wine.

[0004] This application establishes a comprehensive sensory evaluation and prediction model based on electronic nose indicators and physicochemical components, and then proposes a method for accurately evaluating and judging the sensory characteristics of Fujian-style rice wine through data detection, quantifying sensory indicators, and facilitating commercial management and grading of rice wine quality prediction. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method for predicting the quality of Fujian-style rice wine by accurately evaluating and judging the sensory characteristics through data detection, quantifying sensory indicators, and thus facilitating commercial management and grading.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for predicting the quality of rice wine includes: After establishing a sensory evaluation table for rice wine, sensory analysis was conducted on Fujian-style rice wines from different years to obtain a comprehensive taste score. The sensory indicators include appearance, aroma, taste, texture, and style. Correlation and significance analysis were conducted on the physicochemical components and sensory indicators of Fujian-style rice wine. The physicochemical components included pH, total acid, total sugar, non-sugar solids and alcohol content. Correlation and significance analysis were conducted on the color and sensory indicators of Fujian-style rice wine. Color included red-green hue (a), yellow-blue value (b), and brightness. Saturation C; Correlation and significance analysis were conducted on the volatile matter and sensory indicators of Fujian-style rice wine. The volatile matter indicators included aromatic components (W1C); nitrogen oxides (W5S); aromatic components and ammonia (W3C); hydrogen-containing compounds (W6S); short-chain alkanes and aromatic components (W5C); alkane compounds (W1S); sulfur-containing organic compounds (W1W); alcohols, aldehydes and ketones (W2S); sulfides (W2W); and aliphatic hydrocarbons (W3S). Parameters with a significance level <0.05 among physicochemical components, color, and all volatiles were selected as independent variables, and the actual score of comprehensive taste was used as the dependent variable. Stepwise regression analysis was used to establish a predictive model for comprehensive sensory evaluation. Y = -170.625 + 0.025 × a + 0.195 × +0.179×non-sugar solids +52.122×pH -0.169×W5S; Input the information for the Fujian-style Shaoxing wine whose quality needs to be predicted, including red-green hue (a) and brightness. Non-sugar solids and nitrogen oxides (W5S) are predicted by the sensory comprehensive evaluation prediction model, and the predicted result is the comprehensive taste prediction score Y. The Fujian-style rice wine is managed and graded based on the score Y.

[0007] Preferably, the volatile matter index is measured using an electronic nose, including: Accurately pipette 0.5 mL of Fujian-style rice wine sample into a 20 mL sample bottle at 25±2℃, wait 15 min, then insert an electronic nose to aspirate air from the tip to determine the type of sensitive substance.

[0008] Preferably, the electronic nose parameters are set as follows: sample interval time 60s, automatic cleaning time 180s, zeroing time 5s, insertion time 5s, measurement time 180s, inhalation flow rate 180 mL / min, and injection flow rate 180 mL / min. The electronic nose responds after the electrical signal stabilizes and records the response value.

[0009] Preferably, after recording the response value, the experiment is repeated three times in parallel and the average value is calculated.

[0010] Preferably, Excel 2019 software is used for data processing, and the average value of the data obtained from the three parallel experiments is calculated. The result is expressed as mean ± standard deviation.

[0011] Preferably, the electronic nose has 10 sensors that detect 10 volatile substances in the volatile index.

[0012] Preferably, SPSS 17.0 software is used for stepwise regression analysis, significance analysis, and correlation analysis of the data.

[0013] Preferably, to verify the effectiveness of the regression model, statistical analysis is performed on the actual comprehensive taste scores in sensory analysis of different Fujian-style rice wines and the predicted comprehensive taste scores output by the prediction model, and the goodness of fit is calculated.

[0014] Preferably, the sensory analysis consists of 15 evaluators who place wines of different vintages in different glasses and number them. After tasting each wine sample, they rinse their mouths with water, wait 20-30 seconds, and then taste the next wine sample.

[0015] The beneficial effects of this invention are as follows: By studying Fujian-style rice wine materials from different years as research objects, the quality of Fujian-style rice wine is comprehensively evaluated through sensory analysis, electronic nose analysis, and physicochemical analysis. Correlation analysis is conducted on sensory indicators, electronic nose indicators, and physicochemical component content to establish a comprehensive sensory evaluation prediction model based on electronic nose indicators and physicochemical components. This aims to accurately evaluate and determine the sensory characteristics of Fujian-style rice wine through data detection, quantify sensory indicators, and identify data-driven indicators that can directly evaluate the quality of Fujian-style rice wine, namely, red-green color, brightness, non-sugar solids, and nitrogen oxides. This reduces the required parameters and simplifies the prediction model, eliminating the need for complex calculations. It provides a scientific basis for improving the quality identification and evaluation methods of Fujian-style rice wine and enhancing its quality. The prediction results facilitate commercial management and classification of Fujian-style rice wine, enabling grading. Furthermore, the prediction model is not designed based on simple positive or negative correlations; instead, the correlation data is used to design the prediction model through stepwise regression statistical analysis, ensuring more accurate predictions. Attached Figure Description

[0016] Figure 1 Correlation analysis of physicochemical components and sensory indicators in a method for predicting the quality of rice wine according to a specific embodiment of the present invention. Figure 2 Correlation analysis of color and sensory indicators in a method for predicting the quality of rice wine according to a specific embodiment of the present invention. Figure 3 An electronic nose radar image of Fujian-style rice wine, which is a specific embodiment of the present invention for predicting the quality of rice wine. Figure 4 This invention provides a specific embodiment of a rice wine quality prediction method, which includes an electronic nose sensor and sensory indicators correlation analysis. Detailed Implementation

[0017] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0018] Please refer to Figures 1 to 4 A method for predicting the quality of rice wine, comprising: After establishing a sensory evaluation table for rice wine, sensory analysis was conducted on Fujian-style rice wines from different years to obtain a comprehensive taste score. The sensory indicators include appearance, aroma, taste, texture, and style. Correlation and significance analysis were conducted on the physicochemical components and sensory indicators of Fujian-style rice wine. The physicochemical components included pH, total acid, total sugar, non-sugar solids and alcohol content. Correlation and significance analysis were conducted on the color and sensory indicators of Fujian-style rice wine. Color included red-green hue (a), yellow-blue value (b), and brightness. Saturation C; Correlation and significance analysis were conducted on the volatile matter and sensory indicators of Fujian-style rice wine. The volatile matter indicators included aromatic components (W1C); nitrogen oxides (W5S); aromatic components and ammonia (W3C); hydrogen-containing compounds (W6S); short-chain alkanes and aromatic components (W5C); alkane compounds (W1S); sulfur-containing organic compounds (W1W); alcohols, aldehydes and ketones (W2S); sulfides (W2W); and aliphatic hydrocarbons (W3S). Parameters with a significance level <0.05 among physicochemical components, color, and all volatiles were selected as independent variables, and the actual score of comprehensive taste was used as the dependent variable. Stepwise regression analysis was used to establish a predictive model for comprehensive sensory evaluation. Y = -170.625 + 0.025 × a + 0.195 × +0.179×non-sugar solids +52.122×pH -0.169×W5S; Input the information for the Fujian-style Shaoxing wine whose quality needs to be predicted, including red-green hue (a) and brightness. Non-sugar solids and nitrogen oxides (W5S) are predicted by the sensory comprehensive evaluation prediction model, and the predicted result is the comprehensive taste prediction score Y. The Fujian-style rice wine is managed and graded based on the score Y.

[0019] As described above, this study uses materials from different years of Fujian-style rice wine as research objects, comprehensively evaluating the quality of Fujian-style rice wine through sensory analysis, electronic nose analysis, and physicochemical analysis. Correlation analysis is conducted on sensory indicators, electronic nose indicators, and physicochemical component content to establish a comprehensive sensory evaluation prediction model based on electronic nose indicators and physicochemical components. The aim is to accurately evaluate and determine the sensory characteristics of Fujian-style rice wine through data detection, quantify sensory indicators, and identify data-driven indicators that can directly evaluate the quality of Fujian-style rice wine, namely red-green color, brightness, non-sugar solids, and nitrogen oxides. This reduces the required parameters and simplifies the prediction model, eliminating the need for complex calculations. This provides a scientific basis for improving the quality identification and evaluation methods of Fujian-style rice wine and enhancing its quality. The prediction results can facilitate commercial management and classification of Fujian-style rice wine, enabling grading. Furthermore, the prediction model is not designed based on simple positive or negative correlations; instead, the correlation data is used to design the prediction model through stepwise regression statistical analysis, ensuring more accurate predictions.

[0020] Furthermore, volatile organic compound (VOC) indicators are measured using an electronic nose, including: Accurately pipette 0.5 mL of Fujian-style rice wine sample into a 20 mL sample bottle at 25±2℃, wait 15 min, then insert an electronic nose to aspirate air from the tip to determine the type of sensitive substance.

[0021] Further, the electronic nose parameter settings are as follows: sample interval time 60s, automatic cleaning time 180s, zeroing time 5s, insertion time 5s, measurement time 180s, inhalation flow rate 180 mL / min, and injection flow rate 180 mL / min. The electronic nose responds after the electrical signal stabilizes and records the response value.

[0022] Furthermore, after recording the response values, the experiment was repeated three times in parallel and the average value was calculated.

[0023] As can be seen from the above description, by taking multiple samples and calculating the average value, errors can be reduced or avoided, and the accuracy of predictions can be improved.

[0024] Furthermore, the data was processed using Excel 2019 software, and the average value of the data obtained from the three parallel experiments was calculated. The result is expressed as mean ± standard deviation.

[0025] Furthermore, the electronic nose has 10 sensors that detect 10 volatile substances in the volatile index.

[0026] Furthermore, stepwise regression analysis, significance analysis, and correlation analysis were performed using SPSS 17.0 software.

[0027] Furthermore, to verify the effectiveness of the regression model, statistical analysis was conducted on the actual comprehensive taste scores of different Fujian-style rice wines and the predicted comprehensive taste scores output by the prediction model, and the goodness of fit was calculated.

[0028] Furthermore, the sensory analysis involved 15 evaluators who placed wines of different vintages in different glasses and numbered them. After tasting each wine sample, they rinsed their mouths with water, waited for 20-30 seconds, and then tasted the next wine sample.

[0029] Example 1 A method for predicting the quality of rice wine includes: 1. Materials and Methods 1.1 Materials and Reagents The rice wine samples came from Pingnan, Fujian, and were produced in 2023, 2017, 2015, 2014, and 2010.

[0030] Copper sulfate, methylene blue, potassium sodium tartrate, sodium hydroxide, zinc acetate, glacial acetic acid, hydrochloric acid, phenolphthalein, ethanol, and glucose were all of analytical grade.

[0031] 1.2 Main Instruments and Equipment BSA224S Analytical Balance: Sartorius Scientific Instruments Co., Ltd.; DHG-9203A Electric Drying Oven: Shanghai Jinghong Experimental Equipment Co., Ltd.; PHS-3E pH Meter: Shanghai Instrument & Electronics Scientific Instruments Co., Ltd.; CS-200 Precision Colorimeter: Hangzhou Caipu Technology Co., Ltd.; Laboratory Electronic Multi-Fuel Furnace: Tianjin Tester Instrument Co., Ltd.; HH-6 Digital Display Constant Temperature Water Bath: Jiangsu Changzhou Guohua Electric Co., Ltd.; PEN3 Portable Odor Analyzer: Airsense GmbH, Germany.

[0032] 1.3 Test Methods 1.3.1 Sensory Analysis of Shaoxing Wine The sensory analysis team consisted of 15 evaluators. Five different vintages of rice wine were placed in separate cups and numbered. After tasting each sample, the evaluators rinsed their mouths with water, waited 20-30 seconds, and then tasted the next sample. A sensory evaluation form for rice wine was designed based on the sensory requirements of GB / T 13662-2018 "Rice Wine". Specific sensory analysis standards are shown in Table 1.

[0033] Table 1 Sensory Analysis Standards for Shaoxing Wine 1.3.2 Detection of Physicochemical Components of Shaoxing Wine The pH value, total sugar, non-sugar solids, total acid, alcohol content, and color mass fraction of rice wine were determined according to the acidity meter method of GB / T13662-2018 "Rice Wine", GB 5009.7-2016 "Determination of Reducing Sugars in Food", GB / T 13662-2018 "Rice Wine" gravimetric method, GB 12456-2021 "National Food Safety Standard - Determination of Total Acids in Food", GB / T13662-2018 "Rice Wine" instrumental method, and with reference to the methods of MCDERMOTT A, etc.

[0034] 1.4 Determination of the flavor of Shaoxing wine Accurately pipette 0.5 mL of rice wine sample into a 20 mL sample bottle at room temperature, wait 15 min, then insert the electronic nose probe to aspirate the air at the tip to determine the type of sensitive substance.

[0035] Electronic nose parameter settings: sample interval 60s, automatic cleaning time 180s, zeroing time 5s, insertion time 5s, measurement time 180s, inhalation flow rate 180 mL / min, sample injection flow rate 180 mL / min. The metal sensor responds after it stabilizes, and the response value is recorded. This process is repeated three times, and the average value is calculated.

[0036] The PEN3 electronic nose's sensor array comprises 10 metal oxide gas sensors sensitive to different types of volatile gases: W1C, W5S, W3C, W6S, W5C, W1S, W1W, W2S, W2W, and W3S. These 10 sensors vary in sensitivity to different volatile compounds, allowing the entire electronic nose system to detect diverse odors. Specific performance descriptions of each sensor are shown in Table 2.

[0037] Table 2. PEN3 Electronic Nose Sensor Array and Corresponding Sensitive Substance Types 1.5 Data Processing and Analysis Data were organized using Excel 2019 software. The average value of the data from the three parallel experiments was calculated, and the results are expressed as mean ± standard deviation. Data analysis, such as stepwise regression analysis, significance analysis, and correlation analysis, were performed using SPSS 17.0 software. Origin 2021 software was used to draw flavor radar charts, correlation analysis heatmaps, and comparison charts of actual and predicted scores for overall taste.

[0038] 2 Results and Analysis 2.1 Sensory Analysis Results of Shaoxing Wine Table 3 shows that the actual scores of the overall taste of rice wine samples from different years differed significantly (P < 0.05). The significance analysis indicated that the 2010 rice wine had the highest overall taste score of 75.67, while the 2023 rice wine had the lowest score of 60.57. Rice wines from 2010, 2014, and 2015, with longer aging periods, had better overall taste, scoring above 70 points, indicating good overall quality. Rice wines from 2017 and 2023, with shorter aging periods, had average overall taste, scoring between 60 and 70 points, indicating average overall quality. This is because as the aging time of rice wine increases, new substances are generated. Rice wine contains various organic acids, and during storage, the alcohols in the rice wine undergo esterification reactions with these organic acids. Since the newly generated esters each have their own unique aroma, rice wine becomes exceptionally mellow after a certain period of storage, thus improving its overall taste. The results showed that the longer the rice wine was aged, the better its overall taste.

[0039] Table 3 Sensory Analysis Scores of Yellow Wine Note: Different lowercase letters in the same column are used as superscripts to indicate significant differences (P < 0.05). Correlation and significance analyses were conducted on different sensory indicators to further explore the indicators affecting the overall taste of Shaoxing wine. The results are shown in Table 4. There were positive correlations between each pair of the five indicators: appearance, aroma, taste, touch (mouthfeel), and style. Specifically, the correlations between taste and appearance and touch were significant (P < 0.05), while the correlations between the other indicators were highly significant (P < 0.01). The overall taste showed a highly significant positive correlation with each of the indicators: appearance, aroma, taste, touch (mouthfeel), and style (P < 0.01). This means that the clearer and more transparent the Shaoxing wine's appearance, the richer and more mellow its aroma, the more mellow and sweet its taste, the more mellow, lingering, and delicate its mouthfeel, and the more harmonious its style, the higher its overall taste evaluation. Based on the correlation coefficients, the contribution to the overall taste of Shaoxing wine can be determined in the following order: touch (mouthfeel) > style > aroma > taste > appearance.

[0040] Table 4. Correlation analysis of different sensory indicators of Shaoxing wine Note: *. Significant correlation (P < 0.05); **. Extremely significant correlation (P < 0.01).

[0041] 2.2 Analysis of the main physicochemical components of rice wine and their correlation The physicochemical analysis of rice wine helps determine its quality and improve brewing processes. Non-sugar solids and pH are crucial quality indicators. pH changes at different stages of fermentation, affecting microenzyme growth and reproduction, enzyme activity, and the decomposition of nutrients in the fermentation mash. This study investigated five physicochemical indicators of rice wine, exploring the differences in these components across different vintages and their correlation with sensory indicators. Table 5 shows significant differences in pH, total acid, total sugar, non-sugar solids, and alcohol content among rice wines from different vintages. The 2023 vintage had the lowest pH, while the 2010 vintage had the highest, indicating a significant difference; that is, the pH of rice wine increases with longer aging. The 2023 vintage also had the highest non-sugar solids content, while the 2010 vintage had the lowest, with the non-sugar solids content decreasing with longer aging.

[0042] Table 5 Physicochemical Components Analysis of Shaoxing Wine Note: Different lowercase letters in the same column are used as superscripts to indicate significant differences (P < 0.05). To investigate the influence of different physicochemical components in rice wine, correlation and significance analyses were performed on various physicochemical indicators, and the results are shown in Table 6. pH value showed a highly significant negative correlation with total acid, total sugar, and non-sugar solids, while non-sugar solids showed a highly significant positive correlation with total acid and total sugar. In conclusion, the higher the pH value of rice wine, the lower the content of total acid and total sugar; conversely, the lower the content of non-sugar solids, the lower the content of total acid and total sugar.

[0043] Table 6 Correlation analysis of different physicochemical components of rice wine Note: *. Significant correlation (P < 0.05); **. Extremely significant correlation (P < 0.01) pH value is a measure of the acidity of rice wine, and non-sugar solids contribute to its mellowness. This experiment analyzed the correlation and significance of the physicochemical components and sensory indicators of rice wine, and the results are as follows: Figure 1 As shown, pH is significantly positively correlated with all sensory indicators of Shaoxing wine, while non-sugar solids are significantly negatively correlated with all sensory indicators. In summary, higher pH values ​​and lower non-sugar solids content result in better overall taste.

[0044] 2.3 Correlation and Significance Analysis of the Color of Shaoxing Wine Appearance and color are important indicators of the quality of Shaoxing wine, and have a certain impact on its overall quality. This experiment conducted correlation and significance analysis on the color of Shaoxing wine from different years, and the results are shown in Table 7. As can be seen from the table, the colors of Shaoxing wine from different years show significant differences. The higher the a value, the redder the color; the higher the b value, the yellower the color; the higher the L* value, the brighter the color; and the higher the C value, the more saturated the color. Looking at the b value, the 2023 Shaoxing wine has the highest b value, indicating the yellowest color, while the 2010 Shaoxing wine has the lowest b value, which can be interpreted as the color of Shaoxing wine gradually fading from yellow with longer aging time. Looking at the L* value, the 2023 Shaoxing wine has the lowest L* value, indicating its darker color; the 2017 Shaoxing wine has the highest L* value, indicating the best color brightness. In terms of C value, the 2023 Shaoxing wine had the highest C value, meaning the color was the most saturated, while the 2015 Shaoxing wine had the lowest C value, meaning the color was the least saturated. The difference between the two is extremely significant.

[0045] Table 7 Color Analysis of Shaoxing Wine Note: Different lowercase letters in the same column are used as superscripts to indicate significant differences (P < 0.05). To investigate the influence between different colors, correlation and significance analyses were performed, and the results are shown in Table 8. The L* value showed a significant positive correlation with the red-green hue (a) value; a showed a highly significant negative correlation with the yellow-blue hue (b) value; a showed a highly significant negative correlation with the C value; and b showed a highly significant positive correlation with the C value. It should be noted that the superscripts for red-green hue (a) and yellow-blue hue (b) differ from those for a, b, c, and d, which indicate significant differences.

[0046] Table 8 Correlation Analysis of Different Colors of Shaoxing Wine Note: *. Significant correlation (P < 0.05); **. Extremely significant correlation (P < 0.01) To further investigate the influence of color on the sensory indicators of Shaoxing wine, this experiment conducted correlation and significance analysis on the color and sensory indicators of Shaoxing wine. The results are as follows: Figure 2 As shown in the figure. The results indicate that the a value is significantly positively correlated with all sensory indicators of Shaoxing wine, the b value is significantly negatively correlated with all sensory indicators of Shaoxing wine, the L* value is significantly positively correlated with the appearance and taste of Shaoxing wine, and the C value is significantly negatively correlated with all sensory indicators of Shaoxing wine. In summary, the larger the a value and L* value of Shaoxing wine, and the smaller the b value and C value, the better the overall taste. That is, the paler and less saturated the color of Shaoxing wine, the better its overall taste.

[0047] 2.4 Evaluation, Correlation and Significance Analysis of Shaoxing Wine Flavor The overall distribution of all volatile compounds in rice wine was obtained using information from different sensors of an electronic nose, and the flavor differences of rice wine from different years were analyzed. Table 9 shows that sensors W3C and W5C increased with increasing age, indicating that the aromatic substances in rice wine show an increasing trend with age, and the differences are significant (P < 0.05). Sensors W5S, W1S, W1W, and W2W showed a decreasing trend with increasing age, and the differences were significant (P < 0.05). In summary, aromatic components, ammonia, hydrogen-containing compounds, sulfides, sulfur-containing organic compounds, and aliphatic hydrocarbons showed significant differences with increasing age of rice wine (P < 0.05), while short-chain alkanes, aromatic components, nitrogen oxides, and alkane compounds showed significant differences with increasing age of rice wine (P < 0.05).

[0048] Table 9. Significance analysis of various sensors in the electronic nose for Shaoxing wine from different years. Note: Different lowercase letters in the same column are used as superscripts to indicate significant differences (P < 0.05). Figure 3 The radar chart shows the response values ​​of the electronic nose's 10 sensors to Fujian-style rice wine, by... Figure 3 It can be seen that the response values ​​of sensors W3S, W1C, W3C, W6S, and W5C to all Fujian-style rice wine samples were very small, with no significant differences. Compared with other sensors, the Fujian-style rice wine samples showed the largest response value to sensor W2S, followed by sensors W1S, W1W, W2W, and W5S. Characteristic aroma compounds affect the overall aroma profile of the wine. Related studies have shown that the wine has a strong response to sensors W1S, W1W, W2S, W2W, and W5S, which is consistent with the results of characteristic aroma compounds in this study. Among them, the response values ​​of Fujian-style rice wine samples from 2023 and 2014 were significantly larger than those from other years, while the response values ​​of Fujian-style rice wine from 2014 and 2015 showed little difference.

[0049] 2.5 Correlation and Significance Analysis of Sensory Scores and Flavor of Shaoxing Wine W1C was positively correlated with W3C and W5C, indicating that the aromatic components in rice wine have a certain homology with ammonia and alkanes. W1C was negatively correlated with most other sensors (such as W5S, W6S, W1W, etc.), with a significant negative correlation with W5S, W6S, and W1W (P<0.05), and a highly significant negative correlation with W2S (P<0.01). This suggests that there may be a difference in metabolic pathways between the accumulation of aromatic components and flavor substances such as nitrogen oxides, sulfides, organic acids, and alcohols. Studies have shown that the formation of esters in rice wine mainly involves two mechanisms: (1) esterase-catalyzed reactions during microbial metabolism; and (2) chemical reactions between organic acids and alcohols.

[0050] W5S showed a particularly strong positive correlation with W1S, W1W, and W2W (P<0.01), while exhibiting a highly significant negative correlation with W3C and W5C (P<0.01). This suggests that alcohols in rice wine may decrease with increasing aging time due to oxidation, esterification, and volatilization. W2S showed a highly significant negative correlation with W1C (P<0.01) and a highly significant positive correlation with W1W (P<0.01). It has been reported that alcohols can form esters during aging through non-enzymatic esterification or esterase-catalyzed reactions.

[0051] Table 10. Correlation and significance analysis among the sensors of the electronic nose. Note: *. Significant correlation (P < 0.05); **. Extremely significant correlation (P < 0.01) In this study, the rice wine samples all showed strong response values ​​on W1S, W1W, W2S, W2W, and W5S sensors. There was a highly significant positive correlation between the sensors (P<0.01), and they generally showed strong responses to sulfides, alcohols, and aromatic compounds.

[0052] To further explore the influence of different aging times on the flavor of Shaoxing wine and its sensory indicators, this study conducted correlation and significance analyses on the flavor and sensory indicators of Shaoxing wine. The results are as follows: Figure 4 As shown, W5S, W1S, W1W, and W2W were all negatively correlated with the six sensory indicators, with W5S showing the strongest correlation (highly significant negative correlation, P<0.01). W1W had a weaker correlation with taste. Both W1W and W2W were sensitive to sulfides in rice wine. It is also a sulfur donor for other volatile sulfur-containing compounds, and as the age of rice wine increases, The content of [unclear] will decrease accordingly, the irritation of the wine will decrease significantly, and off-flavors will gradually weaken and disappear, making the rice wine more mellow, rich, and delicate.

[0053] According to the relevant charts, W3C and W5C are significantly positively correlated with all six indicators. The electronic nose sensors W1C, W3C, and W5C are all sensitive to aromatic compounds in rice wine. Aromatic compounds are important flavor compounds in rice wine, often exhibiting delicate floral, fruity, and sweet aromas. A prominent aroma with a low threshold is key to the formation of elegant and mellow flavors. As the aging time of rice wine increases, the concentration of aromatic compounds increases, resulting in a more mellow and rich aroma, a more harmonious style, and a higher overall taste score.

[0054] In summary, the weaker the signals from sensors W5S, W1S, W1W, and W2W, and the stronger the signals from W3C and W5C, the clearer and more transparent the appearance of the rice wine, the more mellow and rich its aroma, the more mellow and sweet its taste, the more mellow, long-lasting and delicate its mouthfeel, and the more harmonious its style. The higher the overall taste evaluation of the rice wine will be.

[0055] 2.6 Establishment and Validation of a Sensory Evaluation Prediction Model for Shaoxing Rice Wine To establish an evaluation model relating sensory indicators to the content of physicochemical components and the content of electronic nose sensors, physicochemical indicators of rice wine and the content of electronic nose sensors were used as independent variables, and the actual taste score was used as the dependent variable. Stepwise regression analysis was conducted, with the significance level of the selected variables being less than 0.05. There were five indicators in total, including α value, L* value, non-sugar solids, pH, and W5S. The resulting comprehensive sensory evaluation prediction model is as follows: Y = -170.625 + 0.025 * a + 0.195 * L * + 0.179 * non-sugar solids + 52.122 * pH - 0.169 * W5S (Y represents the overall taste prediction score).

[0056] The prediction model shows that the five indicators—a value, L* value, non-sugar solids, pH, and W5S—can be used as the main factors for evaluating the sensory indicators of Shaoxing wine. Furthermore, to verify the effectiveness of the regression model, a statistical analysis was conducted on the actual overall taste score in the sensory analysis and the predicted overall taste score output by the prediction model. The R-squared value is close to 1, indicating a good model fit (see Table 11). Table 11. Analysis of the difference between the overall taste and the predicted values ​​based on the regression model. Note: Different lowercase letters in the same column are used as superscripts to indicate significant differences (P < 0.05). 3. Conclusion In summary, this experiment analyzed the sensory indicators, electronic nose technology, and physicochemical components of five Fujian-style rice wine samples from different years. The results showed significant differences in flavor characteristics between different years, with the flavor improving with age. Weaker W1S, W1W, W2W, and W3S signals, and stronger W3C and W5C signals, indicate a clearer, more transparent, and luminous appearance; a richer and more mellow aroma; a sweeter and more savory taste; a fuller, more lingering, and delicate mouthfeel; and a more harmonious style. These qualities correlated with a higher overall taste evaluation. Differences existed among the five physicochemical indicators for different rice wines, and higher pH values, lower total acidity, lower non-sugar solids, and lower alcohol content generally resulted in a better overall taste. Five physicochemical components and ten flavor characteristic indicators were used as independent variables in a stepwise regression analysis to construct a predictive model for comprehensive sensory evaluation with statistical significance, including five indicators: α-value, L*-value, non-sugar solids, pH, and W5S. The model is defined as: Y = -170.625 + 0.025*a + 0.195*L* + 0.179*non-sugar solids + 52.122*pH - 0.169*W5S (Y represents the predicted comprehensive taste score). While there is a certain error between the model's predicted and actual comprehensive taste scores, the predicted score still provides valuable reference. Therefore, the comprehensive score prediction model based on regression analysis can achieve a comprehensive evaluation of the sensory indicators of Fujian-style rice wine. Physicochemical components and flavor indicators, as objective methods, can effectively compensate for the subjective limitations of sensory analysis and can be applied to the quality evaluation of different types of rice wine. Based on the analysis of sensory indicators, this study demonstrates the feasibility of establishing an objective and scientific quality evaluation method for rice wine using physicochemical components and flavor indicators. This method can provide a data-driven basis for the selection, innovation, and improvement of high-quality resources, and also facilitates commercial management, grading, and classification.

[0057] In summary, this invention provides a data-driven method for comprehensively evaluating the sensory indicators of Fujian-style rice wine by studying different years' vintages, measuring their sensory indicators, electronic nose sensor data, and physicochemical components, and conducting difference analysis, correlation analysis, and stepwise regression analysis. This provides a theoretical basis for exploring key factors of sensory indicators and improving flavor. The difference analysis results show that there are varying degrees of differences in the sensory indicators, electronic nose sensor data, and physicochemical components of Fujian-style rice wine from different years. The correlation analysis results show that the more transparent and glossy the appearance of Fujian-style rice wine, the better. The richer and more mellow the aroma, the more mellow, smooth, and sweet the taste, and the more delicate and lingering the texture, the better the overall taste of Fujian-style rice wine. In addition, the higher the content of aromatic components and the lower the content of sulfides in Fujian-style rice wine, the better its aroma and the better its overall taste. Among the physicochemical components, the higher the pH value and the lower the total acid, non-sugar solids and alcohol content, the better the overall taste of Fujian-style rice wine. A sensory comprehensive evaluation prediction model was established through stepwise regression analysis to achieve prediction. Furthermore, using electronic nose technology and physicochemical component content as objective evaluation indicators can better compensate for the disadvantage of strong subjectivity in sensory analysis, so as to better improve the quality of rice wine.

[0058] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for predicting the quality of rice wine, characterized in that, include: After establishing a sensory evaluation table for rice wine, sensory analysis was conducted on Fujian-style rice wines from different years to obtain a comprehensive taste score. The sensory indicators include appearance, aroma, taste, texture, and style. Correlation and significance analysis were conducted on the physicochemical components and sensory indicators of Fujian-style rice wine. The physicochemical components included pH, total acid, total sugar, non-sugar solids and alcohol content. Correlation and significance analysis were conducted on the color and sensory indicators of Fujian-style rice wine. Color included red-green hue (a), yellow-blue value (b), and brightness. Saturation C; Correlation and significance analysis were conducted on the volatile matter and sensory indicators of Fujian-style rice wine. The volatile matter indicators included aromatic components (W1C); nitrogen oxides (W5S); aromatic components and ammonia (W3C); hydrogen-containing compounds (W6S); short-chain alkanes and aromatic components (W5C); alkane compounds (W1S); sulfur-containing organic compounds (W1W); alcohols, aldehydes and ketones (W2S); sulfides (W2W); and aliphatic hydrocarbons (W3S). Parameters with a significance level <0.05 among physicochemical components, color, and all volatiles were selected as independent variables, and the actual score of comprehensive taste was used as the dependent variable. Stepwise regression analysis was used to establish a predictive model for comprehensive sensory evaluation. Y = -170.625 + 0.025 × a + 0.195 × +0.179×non-sugar solids +52.122×pH -0.169×W5S; Input the information for the Fujian-style Shaoxing wine whose quality needs to be predicted, including red-green hue (a) and brightness. Non-sugar solids and nitrogen oxides (W5S) are predicted by the sensory comprehensive evaluation prediction model, and the predicted result is the comprehensive taste prediction score Y. The Fujian-style rice wine is managed and graded based on the score Y.

2. The method for predicting the quality of rice wine according to claim 1, characterized in that, Volatile matter indicators are measured using an electronic nose, including: Accurately pipette 0.5 mL of Fujian-style rice wine sample into a 20 mL sample bottle at 25±2℃, wait 15 min, then insert an electronic nose to aspirate air from the tip to determine the type of sensitive substance.

3. The method for predicting the quality of rice wine according to claim 2, characterized in that, Electronic nose parameter settings: sample interval time 60s, automatic cleaning time 180s, zeroing time 5s, insertion time 5s, measurement time 180s, inhalation flow rate 180mL / min, injection flow rate 180mL / min; The electronic nose responds after the electrical signal stabilizes and records the response value.

4. The method for predicting the quality of rice wine according to claim 3, characterized in that, After recording the response value, repeat the operation three times in parallel and calculate the average value.

5. The method for predicting the quality of rice wine according to claim 4, characterized in that, The data were organized using Excel 2019 software. The average value of the data obtained from the three parallel experiments was calculated, and the result is expressed as mean ± standard deviation.

6. The method for predicting the quality of rice wine according to claim 3, characterized in that, The electronic nose has 10 sensors that detect 10 volatile substances in the volatile index.

7. The method for predicting the quality of rice wine according to claim 1, characterized in that, SPSS 17.0 software was used to perform stepwise regression analysis, significance analysis, and correlation analysis on the data.

8. The method for predicting the quality of rice wine according to claim 1, characterized in that, To verify the effectiveness of the regression model, statistical analysis was conducted on the actual comprehensive taste scores and the predicted comprehensive taste scores output by the prediction model in the sensory analysis of different Fujian-style rice wines, and the goodness of fit was calculated.

9. The method for predicting the quality of rice wine according to claim 1, characterized in that, The sensory analysis involved 15 evaluators who placed different vintages of Fujian-style rice wine in different cups and numbered them. After tasting each sample, the evaluators rinsed their mouths with water and waited for 20-30 seconds before tasting the next sample.

Images