Prediction method for total ester content of zero-generation fermentation refreshing beer

Abstract

The invention relates to a method for predicting the total ester content of zero-generation fermented fresh beer, and aims to solve the technical problems of low total ester content and poor flavor quality of zero-generation fermented beer. According to the method, key factors influencing the total ester content are determined by collecting production data of each factory of a group company, and the key factors comprise the pH value of wort, the total acid of the wort, the amino nitrogen content of the wort, the tank type, the contact time of yeast and oxygen, the oxygenation amount per ton of wine, the number of yeast in a full tank, the two-stage main fermentation temperature, the increased sugar degree and the wine storage pressure. Based on the key factors, a partial least square method is adopted to establish a total ester content regulation and control model. The fitting degree (R-sq) of the model reaches 85% or above, and the deviation ratio of a predicted value and a measured value is 10% or below. Through actual production verification, the model can accurately predict the total ester content and guide process adjustment, and the total ester content and flavor quality of the zero-generation fermented beer are remarkably improved. The invention provides an effective means for predicting, regulating and controlling the total ester content for beer production, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to the technical field of beer brewing, and particularly to a method for predicting the total ester content of zero-generation fermented refreshing beer, and more particularly to a prediction model for improving the flavor quality of beer by regulating fermentation process parameters. Background Art

[0002] The flavor quality of beer is an important factor determining its overall quality, and ester compounds are one of the main components of beer flavor. Appropriate esters can endow beer with rich fruity aroma and harmonious taste, but too high or too low ester content will affect the flavor balance of beer. Especially for zero-generation fermented beer, due to the weak vitality and adaptability of its yeast when first used, the total ester content is often low, affecting the flavor quality of beer.

[0003] In the prior art, although there have been some studies on increasing the ester content of beer, these studies mainly focus on the optimization of yeast strains and the adjustment of fermentation processes, and there is a lack of specific research on zero-generation fermented beer. In addition, the process adjustment measures in the prior art are relatively random, lacking systematicness and purposefulness, and it is difficult to effectively predict and control the total ester content in beer.

[0004] Therefore, developing a method that can accurately predict the total ester content of zero-generation fermented refreshing beer is of great significance for improving the flavor quality of beer. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for predicting the total ester content of zero-generation fermented refreshing beer, and by establishing a total ester content regulation model, to achieve accurate prediction and regulation of the total ester content of beer, thereby improving the flavor quality of beer.

[0006] The present application provides a method for predicting the total ester content of zero-generation fermented refreshing beer, including the following steps:

[0007] Step 1: Determine the key factors affecting the total ester content, and the key factors include wort pH value, wort total acid, wort amino nitrogen content, tank type, yeast oxygen contact time, oxygen charge per ton of wine, full-tank yeast count, secondary main fermentation temperature, boosting sugar degree, and storage pressure;

[0008] Step 2: Based on the key factors, perform regression analysis using the partial least squares method to establish a total ester content regulation model, and the equation of the model is:

[0009] Total ester = 80.907 - 0.015 × tank type - 0.004 × oxygen charge per ton of wine + 0.3 × secondary main fermentation temperature - 8.516 × boosting sugar degree - 318.2 × storage pressure;

[0010] Step 3: Verify the model with actual production data and adjust the fermentation process parameters according to the model prediction results to increase the total ester content of the zero-generation fermented beer.

[0011] Furthermore, the goodness of fit (R-sq) of the model reaches over 85%.

[0012] Furthermore, the deviation rate between the predicted value and the measured value of the model is below 10%.

[0013] Furthermore, the method is applied to the production process of zero-generation fermented light beer. By adjusting process parameters such as tank type, oxygen charging amount, and secondary main fermentation temperature, the total ester content of the beer is increased.

[0014] Furthermore, the tank type is the capacity of the fermentation tank, ranging from 250 tons to 480 tons.

[0015] Furthermore, the oxygen charging amount per ton of wine ranges from 150 g / L to 786 g / L.

[0016] Furthermore, the secondary main fermentation temperature ranges from 10°C to 12°C.

[0017] Furthermore, the boosting sugar degree ranges from 3.0°P to 3.8°P.

[0018] Furthermore, the storage pressure ranges from 0.07 MPa to 0.10 MPa.

[0019] On the other hand, the present application provides a production process optimization method for increasing the total ester content of zero-generation fermented light beer, including the following steps:

[0020] S1: Determine the key process parameters affecting the total ester content according to the above prediction method;

[0021] S2: Adjust the key process parameters, including:

[0022] Adjust the tank type of the fermentation tank to 250 tons to 480 tons;

[0023] Adjust the oxygen charging amount per ton of wine to 150 g / T to 160 g / T;

[0024] Adjust the secondary main fermentation temperature to 10°C to 12°C;

[0025] Adjust the boosting sugar degree to 3.0°P to 3.5°P;

[0026] Adjust the storage pressure to 0.06 MPa to 0.08 MPa.

[0027] Advantages of the invention:

[0028] 1. Precise prediction: By establishing a prediction model for the total ester content, it is possible to accurately predict the total ester content of the zero-generation fermented light beer, with the deviation rate between the predicted value and the measured value being less than 10%, and the goodness of fit (R-sq) reaching over 85%.

[0029] 2. Process optimization: By adjusting key process parameters such as tank type, oxygen charge per ton of beer, secondary main fermentation temperature, boosting sugar content, and storage pressure, the total ester content of the zero-generation fermented light beer has been significantly increased, reaching 25.5 mg / L to 43.3 mg / L, and the alcohol-ester ratio has been reduced to 2.5 to 6.4.

[0030] 3. Systematic control: Through the total ester content control system, real-time monitoring and precise control of key process parameters during beer production have been achieved, improving the stability and consistency of beer flavor quality. Description of the Drawings

[0031] Figure 1 It is the Spearman correlation analysis result diagram in the embodiment of the present invention;

[0032] Figure 2 It is the Pearson correlation analysis result in the embodiment of the present invention

[0033] Figure 3 It is the total ester control equation diagram of the light zero-generation fermented beer of the present invention

[0034] Figure 4 It is the fitting degree diagram of the predicted value and the measured value of the present invention

[0035] Figure 5 It is the linear analysis diagram for model verification in the embodiment of the present invention Detailed Embodiments

[0036] The following further elaborates on the present invention in conjunction with the drawings and embodiments.

[0037] 1. Data collection and preprocessing

[0038] 1.1 Data collection

[0039] First, collect the production data of light beer from each factory of the group company. The data sources include the annual production data of more than 10 factories such as those in Guangdong, Henan, Hebei, Baotou, and Beijing in 2023, with the data volume being approximately 220 groups. The key process (PI) data collected include: wort pH value, wort total acid, wort amino nitrogen content, tank type (capacity of the fermentation tank), yeast-oxygen contact time, oxygen charge per ton of beer (g / L), full-tank yeast count, secondary main fermentation temperature (°C), boosting sugar content (°P), and storage pressure (MPa).

[0040] 1.2 Data preprocessing

[0041] Clean and organize the collected data to ensure data integrity and accuracy. Remove outliers and missing values to ensure that each set of data contains all of the above key process (PI) indicators. After data collation, perform standardization processing so that data with different dimensions can be subjected to unified regression analysis.

[0042] 2. Determine the key factors affecting the total ester content

[0043] 2.1 Correlation analysis

[0044] Use Spearman and Pearson correlation analysis methods to analyze the collected production data and determine the correlation between each factor and the total ester content. The specific steps are as follows:

[0045] Import data using statistical software (such as Minitab, SPSS, etc.).

[0046] Perform Spearman and Pearson correlation analyses separately to calculate the correlation coefficients between each factor and the total ester content.

[0047] Based on the analysis results, screen out the factors that are significantly correlated with the total ester content.

[0048] 2.2 Key factor determination

[0049] As Figure 1-2 shown, through correlation analysis, the following factors are determined to have a significant impact on the total ester content:

[0050] Wort pH value, total acid of wort, amino nitrogen content of wort, tank type, yeast oxygen contact time, oxygen charge per ton of wine, number of yeast at full tank, secondary main fermentation temperature, boosting sugar degree, storage wine pressure.

[0051] From Figure 1-2 the analysis results in, it can be seen that by using Spearman and Pearson correlation analysis methods to conduct correlation analysis on each production factor and the total ester content, the key factors with significant correlation in fermentation are basically the same, mainly including: Wort pH value, total acid of wort, amino nitrogen content of wort, tank type, yeast oxygen contact time, oxygen charge per ton of wine, number of yeast at full tank, secondary main fermentation temperature, boosting sugar degree and storage wine pressure, etc. 8 - 9 factors have a significant correlation with the total ester content in 16 days, and the analysis results show "**".

[0052] 3. Establish a total ester content regulation model

[0053] 3.1 Model construction

[0054] As shown in Table 1, based on the above key factors, use partial least squares (PLS) method for regression analysis to establish a total ester content regulation model. The specific steps are as follows:

[0055] Using Minitab Statistical Software or other statistical software, take the above key factor data as independent variables and the total ester content data as the response variable.

[0056] Conduct partial least squares regression analysis to fit the regulation equation.

[0057] Through the stepwise elimination method, delete the factors with low correlation coefficients to optimize the model.

[0058] Table 1 Comparison of Different Prediction Formulas

[0059] Prediction formula PRESS R-sq <![CDATA[y = 17.221 - 1.524x1 - 0.569x2 - 0.003x3 - 0.009x4 + 0.001x5 - 0.006x6 + 3.158x7 + 0.713x8 - 0.001x9 - 311.311x 10 > 280.74 0.797 <![CDATA[y = 20.408 + 8.468x1 + 0.279x2 - 0.03x3 + 0.008x4 - 0.006x6 - 0.162x8 - 0.026x 9- 484.088x 10 > 254.452 0.817 <![CDATA[y = 37.972 + 6.375x1 - 0.029x3 + 0.009x4 - 0.006x6 - 0.169x8 - 0.107x 9- 558.665x 10 > 259.959 0.813 <![CDATA[y = 80.907 - 0.015x4 - 0.004x6 + 0.3x8 - 8.516x 9- 318.2x 10 > 207.714 0.85.2

[0060] In Table 1: y is the total ester content of beer, x1 is the wort pH value, x2 is the total acid of wort, x3 is the wort α - amino nitrogen, x4 is the tank type, x5 is the yeast - oxygen contact time, x6 is the oxygen charge per ton of wine, x7 is the primary fermentation temperature in the first stage, x8 is the primary fermentation temperature in the second stage, x9 is the boosting sugar degree, and x 10 is the storage pressure.

[0061] PRESS: It is the predicted sum of squares, which evaluates the prediction ability of the model. Generally speaking, the smaller the PRESS value, the stronger the prediction ability of the model.

[0062] R - sq: Also called R 2 , which is the coefficient of determination, explaining the variance score of the regression model. The value range is [0, 1]. The larger the value, the better the fitting effect.

[0063] 3.2 Establish the Model Equation

[0064] The finally obtained total ester content regulation model equation is as Figure 3 shown:

[0065] Total ester = 80.907 - 0.015 × Tank type - 0.004 × Oxygen charge per ton of wine + 0.3 × Primary fermentation temperature in the second stage - 8.516 × Boosting sugar degree - 318.2 × Storage pressure

[0066] The goodness of fit (R - sq) of this model reaches more than 85%, indicating that the model has a high prediction accuracy.

[0067] Using the factory production process data (before May 2024) of the key influencing factors determined in the early stage, import them into the above - established partial least squares prediction model, compare the prediction results with the measured data, and conduct a linear comparison analysis on the two groups of data. The results are as Figure 4-5 shown. The coincidence degree between the prediction results and the measured data is good. At the same time, the linear analysis R 2 value between the predicted value and the actual detection result can reach about 0.9, indicating that the coincidence degree between the prediction results and the actual detection data is relatively high and the equation fitting is good.

[0068] 4. Model Validation

[0069] 4.1 Validation Data Preparation

[0070] The model is validated using production data from different batches. Five batches of actual production data are selected, and the key influencing factor index data for each batch are obtained, including:

[0071] Tank type, oxygen charging amount per ton of wine, main fermentation temperature in the second stage, boosting sugar degree, and wine storage pressure.

[0072] 4.2 Comparison between Model Prediction and Actual Measurement

[0073] The above data are imported into the model to calculate the predicted value of the total ester content in 16 days and compare it with the actual test results. The specific steps are as follows:

[0074] Substitute the key factor data of each batch into the model equation to calculate the predicted value.

[0075] Conduct actual tests on the fermentation broth of the same batch to obtain the measured value of the total ester content.

[0076] Compare the predicted value with the measured value and calculate the deviation rate.

[0077] 4.3 Validation Results

[0078] The validation results show that the deviation rate between the predicted value and the measured value is less than 10%, indicating that the model has high prediction accuracy. The specific deviation rates are shown in Table 2 below:

[0079] Table 2 Prediction Effect of Total Ester Prediction Model for Refreshing Zero - generation Fermented Beer

[0080]

[0081]

[0082] 5. Production Application

[0083] 5.1 Tank Type Adjustment Test

[0084] A tank type adjustment test is carried out in the Beijing factory. A 250 - ton fermentation tank is used for brewing production, and a 480 - ton fermentation tank is used as the control group. The specific inlet tank process parameters are shown in Table 3. The specific process parameters are as follows:

[0085] Control group: 480 - ton fermentation tank, current process of four - sugar inlet to the second - stage CO tank.

[0086] Test group: 250 - ton fermentation tank, test process of four - sugar inlet to the second - stage CO tank.

[0087] Other fermentation processes follow the current process, and the key PIs remain unchanged.

[0088] The specific parameters for charging into the tank are shown in Table 3 below.

[0089] Table 3 Process parameters for charging into the tank

[0090]

[0091] 5.2 Test results

[0092] After 16 days of fermentation, the contents of the main flavor substances in the fermentation broth were detected, and the results are shown in Table 4:

[0093] Table 4 Flavor detection results of the test tank after 16 days (mg / L)

[0094]

[0095]

[0096] According to the data in Table 4, the total ester content in the test tanks (126, 128) increased by about 30% compared with the control tank and entered the standard range (25.5 - 43.3 mg / L). According to the model prediction, the theoretical total ester content was 27.4 mg / L, and the prediction deviation was about 6%.

[0097] 5.3 Comprehensive multi-factor regulation test

[0098] A comprehensive multi-factor regulation test was carried out in the third fermentation workshop of the Beijing factory, and the specific adjustment plan is shown in Table 5.

[0099] Table 5 Comprehensive multi-factor adjustment plan

[0100]

[0101] 5.4 Test results

[0102] After 16 days of test fermentation, the conventional physical and chemical and flavor indexes of the fermentation broth were detected, and the results are shown in Table 6 below:

[0103] Table 6 Detection results of key quality indexes

[0104]

[0105]

[0106] The total ester content of the test samples increased significantly by 58.5% (8.6 mg / L) compared with the control and was close to the target range. According to the model prediction, the theoretical total ester content was 24.15 mg / L, and the prediction deviation was 3.5%.

[0107] By the method of the present invention, a prediction model for the total ester content of the zero-generation fermented refreshing beer has been successfully established, and the accuracy of the model has been verified through actual production. This method can effectively guide the process adjustment in the beer production process and improve the flavor quality of the beer.

Claims

1. A method for predicting the total ester content of zero-generation fermented refreshing beer, characterized in that: The following steps are involved: Step 1: Determine the key factors affecting the total ester content, wherein the key factors include wort pH value, wort total acidity, wort amino nitrogen content, tank type, yeast oxygen contact time, oxygenation amount per ton of wine, yeast number in a full tank, second-stage main fermentation temperature, boost sugar content and wine storage pressure; Step 2: Based on the key factors, partial least squares method is used for regression analysis to establish a total ester content control model, and the equation of the model is: Total ester = 80.907-0.015×tank type-0.004×amount of oxygen per ton of wine+0.3×second stage main fermentation temperature-8.516×boost sugar content-318.2×wine storage pressure; Step 3: Verify the model through actual production data, and adjust the fermentation process parameters according to the model prediction results to increase the total ester content of zero-generation fermented beer.

2. The prediction method according to claim 1, characterized in that: The goodness of fit (R-sq) of the model reached more than 85%.

3. The prediction method according to claim 1, characterized in that: The deviation rate between the predicted value of the model and the measured value is less than 10%.

4. The prediction method according to claim 1, characterized in that: The tank type is the capacity of the fermentation tank, ranging from 250 tons to 480 tons.

5. The prediction method according to claim 1, characterized in that: The range of oxygenation amount per ton of wine is 150g / L to 786g / L.

6. The prediction method according to claim 1, characterized in that: The temperature of the second stage main fermentation ranges from 10°C to 12°C.

7. The prediction method according to claim 1, characterized in that: The boosted sugar content ranges from 3.0°P to 3.8°P.

8. The prediction method according to claim 1, characterized in that: The wine storage pressure ranges from 0.07MPa to 0.10MPa.

Images