A method for screening wheat suitable for making koji for brewing

Abstract

The present invention discloses a method for screening wheat suitable for making koji for brewing, belonging to the technical field of brewing. The method of the present invention uses a single-grain cereal hardness tester to measure and analyze the hardness of high-purity wheat grains, clarifies the relationship between the standard deviation and the average value of the hardness of single grains, and establishes a determination model for the hardness consistency of single-grain wheat grains, which can be used for the rapid evaluation of the hardness consistency of mixed wheat samples. This method uses a single-grain cereal hardness tester to measure the wheat samples to be tested, obtains the standard deviation and the average value of the sample hardness, and determines the hardness consistency of the wheat samples to be tested according to the determination model. The method provided by the present invention can rapidly and accurately evaluate the hardness consistency of wheat samples.

Description

Technical Field

[0001] The present invention relates to a method for screening wheat suitable for making koji for brewing, belonging to the technical field of brewing. Background Art

[0002] The high or low hardness of wheat grains reflects the texture of wheat endosperm, that is, the tightness of the combination of protein and starch, and determines the energy consumption of flour milling and the size and breakage rate of starch granules. According to the high or low grain hardness, wheat can be classified into hard wheat, mixed wheat and soft wheat. Wheat grain hardness is an important index for evaluating wheat quality. For example, GB 1351-2008 takes grain hardness as the main index for evaluating wheat quality, replaces the horny rate and farinaceous rate with the hardness index as the characterization index of soft and hard wheat. The hardness index of soft wheat is not higher than 45, the hardness index of hard wheat is not lower than 60, and the mixed wheat is between 45 and 60. Grain hardness is the best evaluation index for determining the processing quality and final use of wheat. In addition to being used for processing flour for food, soft wheat is also an excellent raw material for making koji in the brewing industry.

[0003] With the continuous increase in the consumption of Maotai-flavor liquor, the demand for wheat for making koji in brewing increases year by year. Soft wheat suitable for making koji in brewing has a lower grain hardness, is easy to absorb moisture, and can be rolled into plum blossom petals with a "soft core but intact skin", which is beneficial to the growth of microorganisms. Since manual koji stepping is required for making koji, it is required that the hardness difference between grains is small, especially hard grains should not be mixed. In recent years, the scale and standardization level of wheat production have been continuously improved, but the quality differences of wheat products are still large. Although many large liquor enterprises have established their own wheat production bases for brewing to ensure product quality, due to the mixing of varieties, differences in production conditions, etc., product quality differences will occur, especially the excessive difference in grain hardness. Wheat for making koji in brewing not only requires the average grain hardness to meet that of soft wheat, but also requires good consistency in grain hardness, otherwise it will affect the quality of koji making.

[0004] The methods for measuring wheat grain hardness mainly include the grinding volume method, the grinding time method (GT), the particle index method (PSI), the horny rate method, the near-infrared spectroscopy method (NIR), the single-grain cereal property tester method (SKCS), the electron microscope direct observation method (SEM), etc. Among them, the three relatively commonly used methods are PSI, NIR and SKCS. Chinese enterprises generally use the particle index method (PSI) to measure the average hardness of grains during wheat procurement, and have developed the corresponding JYDB 100 type wheat hardness tester, but it cannot detect the hardness of single grains and evaluate the consistency of grain hardness. Therefore, during the procurement of wheat for making koji in brewing, only the average grain hardness can be used to judge whether it meets the requirements. Summary of the Invention

[0005] [Technical Problem]

[0006] The technical problem to be solved by the present invention is that in the process of making koji for brewing Maotai-flavor liquor, manual trampling of koji is required, which requires small hardness differences between wheat grains used for making koji and good consistency in grain hardness. However, in the existing process of purchasing wheat for making koji in liquor brewing, technicians can only obtain the average grain hardness and cannot screen out wheat with good consistency in grain hardness.

[0007] [Technical Solution]

[0008] The present invention provides a method for screening wheat suitable for making koji in liquor brewing, comprising the following steps:

[0009] (1) Detect the grain hardness of a wheat sample

[0010] Use a single-grain cereal hardness tester to measure the hardness of the wheat sample, and calculate the average and standard deviation of the grain hardness of the wheat sample;

[0011] (2) Calculate the average standard deviation (S)

[0012] If the average of the wheat grain hardness is less than 23, substitute the average (X) into the regression equation S = -0.0166X + 13.2737 to calculate S; if the average of the wheat grain hardness is greater than or equal to 23, substitute the average into the regression equation S = 3.7575E-05X 3 - 6.7350E-03X 2 + 0.3424X + 8.1184 to calculate S;

[0013] (3) Select the wheat for making koji in liquor brewing

[0014] Compare the standard deviation obtained in step (1) with S obtained in step (2), and screen out the wheat samples with a standard deviation ≤ S + 2.2 for making koji in liquor brewing.

[0015] In an embodiment of the present invention, in step (1), the wheat sample needs to remove impurities and broken grains, and the water content should be in the range of 11-13%. A water content exceeding 15% has a greater impact on the screening results of low-hardness soft wheat.

[0016] In an embodiment of the present invention, in step (1), each sample is repeatedly detected 2 times, and 300 grains are detected each time. The average of the 2 times is used as the average and standard deviation of the wheat grain hardness. The difference in the average of the wheat grain hardness does not exceed 2, and the standard deviation does not exceed 0.5, otherwise re-measurement is required.

[0017] In an embodiment of the present invention, the average of the wheat grain hardness in step (1) is in the range of -7.20-80.23.

[0018] In an embodiment of the present invention, wheat samples with an average grain hardness ≤ 45 and a standard deviation ≤ S + 2.2 are selected for brewing koji making.

[0019] [Beneficial effects]

[0020] The present invention uses a single-grain cereal hardness tester (Perten SKCS4100) to detect the grain hardness of single wheat grains, and then establishes a model for characterizing the consistency of wheat grain hardness using the average and standard deviation of single-grain hardness. By measuring the grain hardness of the test sample, calculating the average value, and substituting it into the model, wheat suitable for brewing koji making can be screened from the test sample.

[0021] It only takes a few minutes to detect a sample using Perten SKCS4100, and it has good repeatability, enabling rapid and accurate identification of wheat grain hardness and its consistency. Description of the drawings

[0022] Figure 1 Scatter plot of the standard deviation and average of single-grain hardness of 818 wheat samples.

[0023] Figure 2 Relationship between the standard deviation and average of single-grain hardness of 793 high-purity wheat variety samples.

[0024] Figure 3 Relationship between the increase in hardness standard deviation (Y) and the hardness difference (X) between the original sample and the foreign sample for 3 samples with different mixing ratios. Detailed implementation manners

[0025] The single-grain cereal hardness tester (Perten SKCS4100) used in the following examples is an instrument capable of detecting the hardness of single grains. It can detect 300 grains separately at a time, determine whether the sample belongs to "soft", "mixed", or "hard", obtain the average grain hardness and standard deviation, and the data is directly displayed on the computer with good repeatability and a large detectable hardness value range. The instrument automatically separates each sample through a vacuum separation disk, weighs it after it falls into the weighing hopper, and then the sample falls from the hopper into an insulated crescent-shaped crushing device. During the process of the grain being crushed and flattened when it enters the crescent-shaped gap, the corresponding relationship between the crushing force and time and the conductivity between the crushing wheel and the crescent-shaped device are recorded. The microprocessor transmits a series of data of each grain to the main control computer, including sample weight, weighing scale stability, peak crushing force, average conductivity, crushing area, and crushing length, etc. The computer calculates the single-grain hardness index, grain weight, grain size, and grain moisture, etc., and calculates the average and standard deviation of the corresponding indicators. The main control computer checks the series of data of each sample according to certain criteria and rejects the problematic data, that is, invalid data.

[0026] Example 1: Establishing a model for screening wheat suitable for making koji for brewing using 793 high-purity wheat variety samples

[0027] Through this example, clarify the relationship between the standard deviation and the average value of the single-grain kernel hardness detected by Perten SKCS4100 in unadulterated pure-line wheat variety samples, and find a method to evaluate the consistency of wheat kernel hardness using the standard deviation.

[0028] The specific process is as follows:

[0029] 1. Preparation of wheat samples

[0030] There are a total of 818 wheat variety samples, mainly from wheat-certified varieties and stable advanced-generation lines. After field plot planting and identification, there are no miscellaneous plants or a small number of miscellaneous plants, and the impurities are removed before flowering. Artificial harvesting and threshing are carried out to prevent harvesting or threshing from being mixed. Uniformly dried (the moisture range reaches 11-13%), impurities and broken grains are removed. Randomly extract about 100 grams of samples from each sample and store them for later use.

[0031] 2. Detection of wheat kernel hardness

[0032] After the wheat has passed the after-ripening period (40 days after harvesting), the detection of kernel hardness begins.

[0033] During the detection, first turn on the single-grain cereal hardness tester (Perten SKCS4100) and the switch of the supporting computer. Click the test program on the computer desktop, preheat for 30 minutes, and set the number of detected samples to 300 grains. Thoroughly mix the sample to be measured, randomly take more than 300 grains of kernels and put them into the sample hopper at one time, close the sample hopper door, and the sample will automatically enter the grain hopper. Enter the sample number in the test interface, click the "Run Sample" and "Continue" buttons to start the measurement. When the computer shows 300 grains (the set number of grains), the detection is completed. After all samples are measured, clean the remaining kernels and sample residues in the sample hopper, exit the program, and turn off the instrument and the computer.

[0034] Record the average value and standard deviation of the kernel hardness index, the hardness type of the sample (SOFT / MIXTED / HARD), and the average value of the moisture. Each sample is detected twice. The difference in the average kernel hardness between the two times does not exceed 2, and the standard deviation does not exceed 0.5. Otherwise, re-measure. Calculate the average kernel hardness and standard deviation of the two times and save them to an Excel file.

[0035] 3. Review and elimination of data

[0036] Using the average value of the single-grain kernel hardness of the sample as the horizontal coordinate axis and the standard deviation as the vertical coordinate axis, use Excel to make a scatter plot (Figure 1 ), there are some points outside the concentrated distribution area in the figure. These points mostly have a large standard deviation, mainly due to impure or mixed varieties. Remove the samples corresponding to these points. A total of 818 samples were detected, 25 samples (hollow dots in the figure) were excluded, and 793 samples were retained for subsequent data analysis. Among the 793 samples of higher-purity varieties, there are 502 samples of soft wheat with a grain hardness ≤ 45 (brewing koji-making wheat belongs to soft wheat).

[0037] The average value range of the 793 samples of high-purity varieties is -7.02 - 80.23, and the standard deviation range is 10.24 - 15.67, both showing a continuous distribution. There is basically no mixing of different varieties in the 793 samples of high-purity varieties, and the production conditions of the same sample are similar. The difference in single-grain hardness is mainly determined by variety characteristics and is related to the differences in the grains of the variety itself, such as size, shape, plumpness, etc. Therefore, they are samples with consistent grain hardness.

[0038] 4. Determination of the average standard deviation (S) of high-purity samples with different hardnesses

[0039] The difference in single-grain hardness values can reflect the quality of the grain hardness consistency of wheat samples. The values representing variation in statistics mainly include standard deviation and coefficient of variation. The detection results of Perten SKCS4100 found that the average value of the grain hardness of the samples is not in a direct proportional relationship with the coefficient of variation. Therefore, the coefficient of variation cannot be used to evaluate the hardness consistency of the samples, and only the standard deviation can be used. Through the scatter plot, it is found that the standard deviation of the single-grain hardness of the samples is not independent of the average value of the single-grain hardness, and there is a certain relationship between the two, indicating that the standard deviation of the samples varies with the different average values of the grain hardness.

[0040] To avoid the influence of a large number of samples on the analysis of the relationship between the two, the 793 samples of varieties were grouped according to the average value of the grain hardness. A total of 17 groups were divided, with a group interval of 5, and the average grain hardness and average standard deviation of each group of samples were calculated (Table 2). According to the AACC55-31 method, the detected samples mainly include 7 out of all 8 categories, but there is only 1 sample of the very hard type (hardness 80 - 90), and the extremely hard type (hardness greater than 90) is missing. From the grouping results, the hardness of the samples is mainly concentrated between 25 - 70, and there are more of the 4 types of soft (25 - 34), medium-soft (35 - 44), medium-hard (45 - 64), and hard (65 - 80). It can be found from Table 1 that the standard deviation of the grain hardness shows a changing trend of slowly decreasing to slowly increasing and then decreasing significantly with the increase of the average value. No suitable curve regression equation for the standard deviation and the average value was fitted using Excel software.

[0041] Table 1 Grouping results of the grain hardness of wheat samples

[0042]

[0043] Further interval analysis reveals different regression relationships when the kernel hardness is less than 23 and greater than 23. When the kernel hardness is less than 23, the standard deviation shows a linear decreasing trend with the increase of kernel hardness. Its regression equation is S = -0.0166X + 13.2737, reaching a highly significant level ( P = 4.01E-03), and the coefficient of determination R 2 = 0.90, indicating a relatively high fitting degree. When the kernel hardness is greater than 23, the standard deviation shows a trend of increasing first and then decreasing with the increase of kernel hardness, conforming to a cubic polynomial curve. Its regression equation is S = 3.7575E-05X 3 - 6.7350E-03X 2 + 0.3424X + 8.1184, reaching a highly significant level ( P = 3.18E-07), and the coefficient of determination R 2 = 0.98, indicating an extremely high fitting degree. In the two regression equations, X is the average kernel hardness of the variety sample, and S is the predicted standard deviation for a certain X value, that is, the average standard deviation of high-purity samples with different hardnesses. When the kernel hardness is 23, the S values calculated using the two equations are basically equal, both being 12.9. Therefore, taking 23 as the boundary, it is also the integer closest to the average value 23.47 in the 6th group of Table 2. Figure 2 The main results of the regression analysis are shown.

[0044] If the average value of the single-kernel hardness is less than 23, use it as X and substitute it into the linear regression equation S = -0.0166X + 13.2737 to calculate S. If the average value is greater than or equal to 23, use it as X and substitute it into the cubic polynomial regression equation S = 3.7575E-05X 3 - 6.7350E-03X 2 + 0.3424X + 8.1184 to calculate S.

[0045] 5. Judgment of the Consistency of Wheat Kernel Hardness

[0046] The average of the single-kernel hardness of 793 samples was taken as X, substituted into the corresponding regression equation to calculate S, and the difference between the standard deviation of the kernel hardness of each sample and the corresponding S was calculated. The distribution of these differences was used to determine the boundary for the consistency of kernel hardness. Through the descriptive statistics of these differences, the average was 0.006, the standard deviation was 0.939, the kurtosis was -0.302, and the skewness was 0.118, which conformed to the normal distribution and was close to the standardized normal distribution. 95% and 99% are two commonly used confidence probabilities in agricultural scientific research. Considering the many uncertain factors in agricultural production conditions, the boundary was determined according to the confidence probability of 99%. The critical value of the right tail 1% was 2.2. ≤2.2 was the acceptance region of the normal distribution, and >2.2 was the rejection region. Therefore, S + 2.2 was taken as the boundary for evaluating the consistency of kernel hardness using the standard deviation. If the standard deviation of the kernel hardness of the sample ≤ S + 2.2, the kernel hardness consistency was determined as "consistent", otherwise it was "inconsistent".

[0047] In addition, since the wheat for making brewing koji belongs to soft wheat, according to the methods of GB1351-2008 and AACC 55-31, the average kernel hardness of soft wheat ≤ 45. The average kernel hardness of the samples with the hardness type of soft (SOFT) detected by Perten SKCS4100 was not higher than 45, but there were also samples with an average kernel hardness ≤ 45 and the hardness type of mixed (MIXTED), mainly samples mixed with some hard (HARD) ones. However, samples determined as mixed should not be considered as having "inconsistent" kernel hardness because the standard deviation of some mixed samples with an average hardness of about 45 was also relatively low.

[0048] Finally, wheat samples with an average kernel hardness ≤ 45 and a standard deviation ≤ S + 2.2 were selected for making brewing koji.

[0049] Conclusion

[0050] Example 1 details the process of establishing a model for screening wheat suitable for making brewing koji using 793 high-purity wheat variety samples. By analyzing the relationship between the standard deviation and the average of the single-kernel hardness of 793 high-purity wheat variety samples through Excel software, two regression equations with high goodness of fit were obtained, thereby determining the average standard deviation (S) of high-purity samples with different hardnesses. The distribution of the differences between the standard deviation of the kernel hardness of these samples and the corresponding S not only conformed to the normal distribution but was also close to the standardized normal distribution, indicating the scientific nature of the S calculation. Using the critical value of the 1% probability interval of the right tail of the normal distribution as the boundary for evaluating hardness consistency ensured that about 99% of these 793 wheat variety samples were evaluated as having "consistent" hardness.

[0051] According to the method of the present invention, hardness consistency evaluation was carried out on all 818 samples. Among them, 787 samples were evaluated as "consistent" in hardness, accounting for 96.2%, and 31 samples were "inconsistent" (Table 3), accounting for 3.8%. The standard deviations of the hardness of these samples were relatively high, with the lowest being 14.57 and the highest being 20.37, all higher than the corresponding S + 2.2. Among the 793 high-purity samples, 785 samples were evaluated as "consistent" in hardness, accounting for 99.0%, and 8 samples were "inconsistent" (serial numbers 24 - 31 in Table 3), accounting for 1.0%. The lowest was 14.57 and the highest was 15.67, all higher than the corresponding S + 2.2. Among the 818 wheat samples, 502 samples had an average grain hardness ≤ 45 and were evaluated as "consistent" in hardness, being suitable for brewing koji making.

[0052] Table 2 Hardness results of samples with "inconsistent" grain hardness determination

[0053]

[0054] Example 2 Evaluation of hardness consistency of mixed wheat samples using the model (screening conditions) established in Example 1

[0055] 1. Mixing of samples and detection of grain hardness

[0056] Nine high-purity wheat variety samples with different average grain hardness and standard deviation were selected from Example 1. The samples were numbered A, B, C, D, E, F, G, H, and K in ascending order of the average grain hardness. The consistency of the grain hardness of all samples was "consistent". The average grain hardness, standard deviation detected, and the average standard deviation of the corresponding standard samples are listed in Table 3. The nine samples were mixed pairwise with each other according to the proportions of 5%, 15%, and 25% of the number of different samples in the total 300 grains (in the obtained mixed samples, the samples with a proportion ≥ 25% are the original samples, and the samples with a proportion ≤ 25% are the different samples), obtaining 216 mixed sample combinations. Except for the requirement of mixing 300 grains, the preparation and hardness detection of the samples were the same as in Example 1.

[0057] Table 3 Average grain hardness, standard deviation of the selected samples for mixing, and average standard deviation of the corresponding standard samples

[0058]

[0059] 2. Evaluation of grain hardness consistency of mixed samples

[0060] According to the size of the average grain hardness of the mixed samples, substitute them into the regression equations S = -0.0166X + 13.2737 (when the average is less than 23) and S = 3.7575E-05X 3 - 6.7350E-03X 2Calculate S using +0.3424X + 8.1184 (where the average is greater than or equal to 23). Then, determine whether the kernel hardness of each blended sample is consistent based on its standard deviation, and at the same time compare it with the original sample to analyze the changes in the average kernel hardness and standard deviation. The results of kernel hardness detection and consistency evaluation are listed in Table 4.

[0061] Table 4 Detection and Consistency Evaluation Results of Kernel Hardness of Blended Samples

[0062]

[0063] Note: The mixed sample combination 285A + 15B represents a mixture of 285 grains of the original sample A and 15 grains of the different sample B, and the others are similar; 1 and 2 for consistency represent "consistent" and "inconsistent" respectively.

[0064] Analysis of the kernel hardness detection results of three different blending ratios found that the average standard deviations of the 5%, 15%, and 25% blended samples increased by 2.5, 5.6, and 7.6 respectively compared to the original sample, indicating that as the blending ratio increases, the standard deviation shows an obvious increasing trend, but the increasing amplitude decreases continuously. The change ranges of the standard deviations of the samples with three blending ratios are 0.0 - 10.6, 0.1 - 20.6, and 0.2 - 25.9 respectively, and their larger change intervals are also related to the difference in kernel hardness between the original sample and the different sample. From the relationship between the increase in the standard deviation of the kernel hardness of the samples with three different mixing ratios and the hardness difference between the original sample and the different sample ( Figure 3 ) it can be seen that the larger the blending ratio, the greater the increase in the standard deviation. The increase in the standard deviation has a highly significant quadratic polynomial relationship with the hardness difference. As the hardness difference increases, the increase in the standard deviation shows an upward trend, and the upward amplitude increases continuously.

[0065] Analysis of the kernel hardness consistency of the samples with different blending ratios and the change in the hardness difference between the two mixed samples (Table 5) found that there are both "consistent" and "inconsistent" samples in terms of kernel hardness for the three blending ratios. As the blending ratio increases, the number of "consistent" samples and the maximum hardness difference of the mixed samples decrease, and the hardness difference boundaries between the two types of mixed samples become more obvious. The hardness difference of the 25% blended sample is bounded by 21.2.

[0066] Table 5 Changes in Kernel Hardness Consistency of Samples with Different Blending Ratios and Hardness Difference between Two Mixed Samples

[0067]

[0068] Conclusion

[0069] Example 2 characterized the hardness consistency of the mixed samples by using the method established in Example, thus verifying the application effect of the screening method constructed in Example 1. Twenty-one mixed samples were constructed at the mixing ratios of 5%, 15% and 25% by selecting 9 high-purity wheat variety samples with different average values and standard deviations of kernel hardness, and the consistency of their kernel hardness was characterized. Compared with the original samples, the increase in the standard deviation of kernel hardness and the change in consistency of the mixed samples were mainly related to the mixing ratio and the hardness difference between the mixed samples. The greater the mixing ratio and the greater the hardness difference between the mixed samples, the greater the increase in the standard deviation and the worse the hardness consistency, indicating that the present invention can better evaluate the hardness consistency of the mixed samples.

[0070] Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person familiar with this technology can make various modifications and decorations without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be defined by the claims.

Claims

1. A method for screening wheat suitable for making koji for brewing, characterized in that, it includes the following steps: (1) Detect the kernel hardness of the wheat sample Use a single-grain cereal hardness tester to measure the kernel hardness of the wheat sample, and calculate the average and standard deviation of the kernel hardness of the wheat sample; (2) Calculate the average standard deviation S If the average value of the wheat kernel hardness is less than 23, substitute the average value X into the regression equation S = -0.0166X + 13.2737 to calculate S; if the average value of the wheat kernel hardness is greater than or equal to 23, substitute the average value into the regression equation S = 3.7575E-05X 3 - 6.7350E-03X 2 + 0.3424X + 8.1184 to calculate S; (3) Select the wheat for making koji for brewing Compare the standard deviation obtained in step (1) with S obtained in step (2), and screen out the wheat samples with a standard deviation ≤ S + 2.2 for making koji for brewing.

2. The method according to claim 1, characterized in that, in step (1), the wheat sample needs to remove impurities and broken grains, and the moisture content of the wheat sample is 11-13%.

3. The method according to claim 1, characterized in that, in step (1), each sample is repeatedly detected 2 times, and 300 grains are detected each time. The average of the 2 times is used as the average and standard deviation of the kernel hardness of the wheat; the difference in the average of the kernel hardness of the wheat does not exceed 2, and the standard deviation does not exceed 0.5, otherwise re-determine.

4. The method according to claim 1, characterized in that, the average of the kernel hardness of the wheat in step (1) is in the range of -7.20 - 80.

23.

5. The method according to claim 1, characterized in that, select the wheat samples with an average of kernel hardness ≤ 45 and a standard deviation ≤ S + 2.2 in step (3) for making koji for brewing.

6. The method according to claim 1, characterized in that, the wine is Maotai-flavor Baijiu.

Images