Method and system for analyzing influence factors of beany flavor of soybeans and products

Abstract

The invention provides a soybean and product beany flavor influence factor analysis method and system, and relates to the technical field of soybean product analysis, and the method comprises the following steps: selecting a plurality of groups of soybean products of the same variety and the same quality as samples, scoring the beany flavor of the samples, and forming a sensory score data set; collecting the concentration of volatile organic compounds in the sample, simultaneously obtaining planting environment data and processing condition data of the sample, and calculating a compound concentration index, an environment comprehensive index and a processing comprehensive index; analyzing the correlation between the compound concentration index, the environmental comprehensive index and the processing comprehensive index and the final sensory score by using a Pearson's correlation coefficient method; and establishing a regression model by adopting a multiple linear regression analysis method, changing each index one by one, observing the change degree of the final sensory score, and calculating a sensitivity index. According to the method, producers can make more accurate decisions in actual production, and the sensory quality and market competitiveness of products are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of soybean product analysis, and specifically provides a method and system for analyzing the influencing factors of the beany flavor of soybeans and their products. Background Art

[0002] Soybeans and their products are widely consumed foods globally and are favored by an increasing number of consumers due to their rich nutrition and health benefits. However, the beany flavor, as an important sensory characteristic of soybeans and their products, often affects consumers' acceptance and willingness to purchase. The generation of the beany flavor is closely related to various factors, including soybean variety, planting environment, processing methods, and the release of volatile organic compounds, etc. However, existing research often adopts a single-factor analysis method and lacks a systematic evaluation of the comprehensive influence of multiple factors. This one-sided method makes it difficult for us to accurately grasp the formation mechanism of the beany flavor and also unclear about the interaction between various factors.

[0003] In addition, the current evaluation of the beany flavor mostly relies on sensory evaluation. Although it can provide intuitive feelings, it lacks scientific quantitative analysis methods and often has subjectivity and limitations. Many studies only focus on some factors, such as the influence of a certain volatile organic compound, while ignoring the role of other environmental and processing conditions. This method not only cannot comprehensively reflect the complexity of the beany flavor but may also lead to incorrect judgments on product quality. In addition, the existing technology is also relatively simple in data processing, lacking systematic modeling and multivariate analysis of influencing factors, resulting in difficulty in providing effective optimization suggestions for production, restricting the quality improvement and market competitiveness of soybean products. Therefore, developing a scientific analysis method that can comprehensively evaluate the influencing factors of the beany flavor has become an important issue that the industry urgently needs to solve.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for analyzing the influencing factors of the beany flavor of soybeans and their products to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A method for analyzing the influencing factors of the beany flavor of soybeans and their products, the specific steps include:

[0008] Step 1: Select several groups of soy products with the same variety and the same quality as samples, organize 10 professionally trained sensory evaluators to score the beany flavor of the samples, and take the average value as the final sensory score of the samples to form a sensory score data set;

[0009] Step 2: Collect the concentration of volatile organic compounds in the samples. At the same time, obtain the planting environment data and processing condition data of the samples. After dimensionless processing of the final sensory score, the concentration of volatile organic compounds, the planting environment data, and the processing condition data, calculate the compound concentration index, the environmental comprehensive index, and the processing comprehensive index;

[0010] Step 3: Use the Pearson correlation coefficient method to analyze the correlation between the compound concentration index, the environmental comprehensive index, the processing comprehensive index and the final sensory score, calculate the correlation coefficients respectively, and combine the analysis results to judge the influence degree of different indexes on the beany flavor;

[0011] Step 4: Adopt the multiple linear regression analysis method. Take the final sensory score as the dependent variable, and the compound concentration index, the environmental comprehensive index, and the processing comprehensive index as the independent variables to establish a regression model. Vary each index one by one, observe the change degree of the final sensory score, and calculate the sensitivity index. Sort the influencing factors according to the sensitivity index of each factor to provide a basis for optimization.

[0012] Furthermore, the specific logic for scoring the beany flavor of the samples is as follows: The scoring standard is 0 - 10 points. 0 points means no beany flavor, and 10 points means a very strong beany flavor. Each evaluator scores each sample three times independently, and takes the average value as the final sensory score. Record the sensory score data of each sample to form a sensory score data set S:

[0013] S = {s1, s2, …, s n}

[0014] where s n represents the sensory score of the nth sample, and n is the index of the sample.

[0015] Furthermore, collect the concentration of volatile organic compounds in the samples, specifically: Use a gas chromatography - mass spectrometry (GC - MS) instrument to detect the concentration of volatile organic compounds in the samples. The volatile organic compounds include hexanal, acetone, and hexanol. Each sample is detected three times repeatedly, and the average value is taken as the final concentration value of each volatile organic compound;

[0016] The planting environment data includes cumulative rainfall, cumulative fertilization amount, and soil humidity. The processing condition data includes processing time, processing temperature at the end of processing, and soil pH value at the end of processing. And for the same sample, the raw materials used are all in the same planting environment.

[0017] Furthermore, the formula for calculating the compound concentration index is as follows:

[0018]

[0019] In the formula, CCI is the compound concentration index, and C a is the concentration of the a-th volatile organic compound, where a is the index of the volatile organic compound. When a = 1, it corresponds to hexanal; when a = 2, it corresponds to acetone; when a = 3, it corresponds to hexanol;

[0020] The formula for calculating the environmental comprehensive index is as follows:

[0021]

[0022] In the formula, ECI is the environmental comprehensive index, and E b is the value of the b-th environmental factor, where b is the index of the environmental factor. When b = 1, it corresponds to the cumulative rainfall; when b = 2, it corresponds to the cumulative fertilization amount; when b = 3, it corresponds to the soil humidity. α is used to adjust the influence degree of the index to ensure that the result is dimensionless, and 0 < α < 1;

[0023] The formula for calculating the processing comprehensive index is as follows:

[0024]

[0025] In the formula, PCI is the processing comprehensive index, and p k is the value of the k-th processing condition, where k is the index of the processing condition. When k = 1, it corresponds to the processing time; when k = 2, it corresponds to the processing temperature at the end of processing; when k = 3, it corresponds to the processing temperature at the end of processing. γ is used to adjust the influence degree of the index to ensure that the result is dimensionless, and 0 < γ < 1.

[0026] Furthermore, the Pearson correlation coefficient method is used to analyze the correlation between the compound concentration index, the environmental comprehensive index, the processing comprehensive index and the final sensory score. The formula is as follows:

[0027]

[0028] Among them, r is the Pearson correlation coefficient, and X i represents the compound concentration index, the environmental comprehensive index or the processing comprehensive index of the i-th sample, which is used to analyze the correlation with the corresponding final sensory score respectively. Y i represents the final sensory score of the i-th sample, is the corresponding mean value, is the mean value of the final sensory score;

[0029] The relationship between each index and the final sensory score is analyzed through the calculated Pearson correlation coefficient value r:

[0030] When r > 0, it indicates a positive correlation between this index and the final sensory score, that is, it means that when the index increases, the final sensory score also increases;

[0031] When r < 0, it indicates a negative correlation between this index and the final sensory score, meaning that when the index increases, the final sensory score decreases instead;

[0032] When r = 0, there is no correlation, that is, it means there is no linear relationship between this index and the final sensory score;

[0033] Among them, when |r| is closer to 1, it means the stronger the correlation.

[0034] Furthermore, using the multiple linear regression analysis method, with the final sensory score as the dependent variable and the compound concentration index, environmental comprehensive index, and processing comprehensive index as the independent variables, a regression model is established, and the expression is as follows:

[0035] M = β0 + β1CCI + β2ECI + β3PCI + ∈

[0036] In the formula, M is the final sensory score, CCI is the compound concentration index, ECI is the environmental comprehensive index, PCI is the processing comprehensive index, β1, β2, and β3 are the regression coefficients of their respective independent variables, β0 is the intercept, and ∈ is the error term;

[0037] Use Python for model fitting, obtain the regression coefficients and statistical significance, and calculate the coefficient of determination:

[0038]

[0039] In the formula, R 2 is the coefficient of determination, used to evaluate the goodness of fit of the regression model to the final sensory score, M i is the final sensory score of the i-th sample, represents the predicted final sensory score of the i-th sample by the model, represents the mean of the final sensory scores of all samples.

[0040] Furthermore, calculate the sensitivity index. Specifically: Set the dependent variable as the final sensory score, and set the independent variables as the compound concentration index, environmental comprehensive index, and processing comprehensive index. Vary each independent variable one by one, record the initial value and change value of each independent variable, as well as the initial value and change value of the corresponding final sensory score, and calculate the sensitivity index of each independent variable according to the following formula:

[0041]

[0042] In the formula, SI z is the sensitivity index of the z-th independent variable, M oldis the initial value of the final sensory score, ΔM represents the change value of the final sensory score, N old represents the initial value of the independent variable, ΔN represents the change value of the independent variable, z is the index of the independent variable, z ∈ {1, 2, 3}. When z = 1, the corresponding independent variable is the compound concentration index; when z = 2, the corresponding independent variable is the environmental comprehensive index; when z = 3, the corresponding independent variable is the processing comprehensive index;

[0043] When |SI| > 1, it indicates that the change of this independent variable leads to a relatively large change in the sensory score, indicating that this independent variable has a strong influence on the sensory score;

[0044] When |SI| < 1, it indicates that the change of this independent variable leads to a relatively small change in the sensory score, indicating that this index has a weak influence on the sensory score;

[0045] When |SI| = 1, it indicates that the influence of this index is moderate;

[0046] Arrange the sensitivity index values from high to low. According to the results of the sensitivity analysis, the researchers conduct in-depth research and optimization on the factors with high sensitivity index to improve the sensory quality of soy products.

[0047] The present invention also further provides an analysis system for influencing factors of beany flavor in soybeans and their products. The analysis system for influencing factors of beany flavor in soybeans and their products is used to execute the above-mentioned analysis method for influencing factors of beany flavor in soybeans and their products, including:

[0048] A sample selection and scoring module, which is used to select several groups of soy products of the same variety and the same quality as samples, organize 10 professionally trained sensory evaluators to score the beany flavor of the samples, and take the average value as the final sensory score of the samples to form a sensory score data set;

[0049] A data collection and processing module, which is used to collect the concentration of volatile organic compounds in the samples, and at the same time obtain the planting environment data and processing condition data of the samples. After dimensionless processing of the final sensory score, the concentration of volatile organic compounds, the planting environment data and the processing condition data, calculate the compound concentration index, the environmental comprehensive index and the processing comprehensive index;

[0050] A correlation analysis module, which is used to analyze the correlation between the compound concentration index, the environmental comprehensive index and the processing comprehensive index and the final sensory score by using the Pearson correlation coefficient method, calculate the correlation coefficients respectively, and combine the analysis results to judge the influence degree of different indexes on the beany flavor;

[0051] Regression model and sensitivity module are used to establish a regression model by means of multiple linear regression analysis method, with the final sensory score as the dependent variable and the compound concentration index, environmental comprehensive index and processing comprehensive index as the independent variables. Each index is changed one by one to observe the degree of change of the final sensory score, and the sensitivity index is calculated. The influencing factors are ranked according to the sensitivity index of each factor, providing a basis for optimization.

[0052] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0053] Through dimensionless processing and comprehensive index calculation in the present invention, data from different sources can be unified, facilitating multi-dimensional analysis. By using the Pearson correlation coefficient and multiple linear regression analysis, the relationship between various factors and the soybean odor can be quantified, and a prediction model can be established to guide producers to make scientific decisions during the processing. In addition, the calculation of the sensitivity index shows the relative influence degree of each influencing factor on the change of the sensory score, helping producers to prioritize factors with high sensitivity, thereby effectively optimizing the planting and processing conditions. This series of methods not only provides a scientific and systematic analysis of the influencing factors of soybean odor, but also provides theoretical support and practical basis for improving the overall quality and market competitiveness of soybeans and their products. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0055] Figure 2 It is a schematic diagram of the system module of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to specific embodiments.

[0057] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those of ordinary skill in the field to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before the term cover the elements or objects listed after the term and their equivalents, without excluding other elements or objects. The terms such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right" are only used to represent relative positional relationships, and when the absolute position of the object to be described changes, the relative positional relationship may also change accordingly.

[0058] Example:

[0059] Please refer to Figure 1 , the present invention provides a technical solution:

[0060] A method for analyzing the influencing factors of soybean odor in soybeans and their products, the specific steps include:

[0061] Step 1: Select several groups of soy products of the same variety and the same quality as samples, organize 10 professionally trained sensory evaluators, score the soybean odor of the samples, and take the average value as the final sensory score of the samples, forming a sensory score data set;

[0062] In this embodiment, when scoring the soybean odor of the samples, the specific logic is as follows: The scoring standard is 0-10 points, 0 points means no soybean odor, 10 points means very strong soybean odor. Each evaluator independently scores each sample three times, and takes the average value as the final sensory score, records the sensory score data of each sample, and forms a sensory score data set S:

[0063] S = {s1, s2, …, s n}

[0064] In the formula, s n represents the sensory score of the nth sample, and n is the index of the sample.

[0065] The advantage of Step 1 is that by organizing professionally trained sensory evaluators to score soy products of the same variety and the same quality, the objectivity and consistency of the sensory scores can be ensured. Such an independent scoring mechanism reduces individual subjective biases, and the obtained average value is more reliable, providing a solid foundation for subsequent analysis. Compared with the prior art, this step improves the accuracy and comparability of the data through a standardized scoring system and the participation of professionals, making the research results more persuasive.

[0066] In this solution, adopting this step can effectively construct a reliable sensory score data set, laying a foundation for subsequent multivariate analysis and model establishment. High-quality sensory score data not only enhances the correlation analysis with volatile organic compound and environmental and processing condition data, but also makes the final regression model more accurate, thereby improving the analysis ability of the influencing factors of soybean odor and providing a scientific basis and reliable support for the improvement of the quality of soy products. This promoting effect helps to drive innovation and development in the quality control and product optimization of the soy product industry.

[0067] Step 2: Collect the concentration of volatile organic compounds in the samples, and simultaneously obtain the planting environment data and processing condition data of the samples. After dimensionless processing of the final sensory score, the concentration of volatile organic compounds, the planting environment data, and the processing condition data, calculate the compound concentration index, the environmental comprehensive index, and the processing comprehensive index;

[0068] In this embodiment, collecting the concentration of volatile organic compounds in the samples specifically means: using a gas chromatography-mass spectrometry (GC-MS) instrument to detect the concentration of volatile organic compounds in the samples. The volatile organic compounds include hexanal, acetone, and hexanol. Each sample is detected three times repetitively, and the average value is taken as the final concentration value of each volatile organic compound;

[0069] The planting environment data includes cumulative rainfall, cumulative fertilization amount, and soil humidity. The processing condition data includes processing time, processing temperature at the end of processing, and soil pH value at the end of processing. For the same sample, the raw materials used are all in the same planting environment.

[0070] The formula for calculating the compound concentration index is as follows:

[0071]

[0072] In the formula, CCI is the compound concentration index, C a is the concentration of the a-th volatile organic compound, and a is the index of the volatile organic compound. Among them, when a = 1, it corresponds to hexanal; when a = 2, it corresponds to acetone; when a = 3, it corresponds to hexanol;

[0073] The value of CCI can reflect the overall concentration level of volatile organic compounds in the sample. A higher CCI value means that the concentration of volatile organic compounds in the sample is generally higher, which may be related to a strong bean smell. Using the natural logarithm to transform the concentration of each compound can reduce the skewed distribution of the concentration data. Especially when the concentration of some compounds is very low, the logarithmic transformation can also avoid infinity during calculation. By squaring the logarithmic concentration of each compound, then summing, and finally taking the square root, a comprehensive and multi-dimensional concentration index is formed. This processing method emphasizes the influence of high-concentration compounds on the overall concentration index because the squaring process will amplify the relative contribution of larger values.

[0074] The formula for calculating the environmental comprehensive index is as follows:

[0075]

[0076] In the formula, ECI is the environmental comprehensive index, E bis the value of the b-th environmental factor, where b is the index of the environmental factor. When b = 1, it corresponds to the cumulative rainfall; when b = 2, it corresponds to the cumulative fertilizer application; when b = 3, it corresponds to the soil humidity. α is used to adjust the influence degree of the exponent to ensure that the result is dimensionless, and 0 < α < 1;

[0077] The formula uses to process each environmental factor. The purpose of this transformation is to ensure that environmental factors with different units and ranges are converted into a dimensionless value, which helps to eliminate the influence of dimensions in comprehensive evaluation. Continuous functions such as logarithms, squares, and square roots are used, and the entire formula is smooth and differentiable within a reasonable range. Such characteristics make the calculation and analysis of ECI relatively easy in practical applications, and the result changes smoothly.

[0078] The formula for calculating the processing comprehensive index is as follows:

[0079]

[0080] In the formula, PCI is the processing comprehensive index, and p k is the value of the k-th processing condition, where k is the index of the processing condition. When k = 1, it corresponds to the processing time; when k = 2, it corresponds to the processing temperature at the end of processing; when k = 3, it corresponds to the processing temperature at the end of processing. γ is used to adjust the influence degree of the exponent to ensure that the result is dimensionless, and 0 < γ < 1.

[0081] The use of the square root in the formula ensures that PCI is smooth and differentiable within a reasonable range, which means that in practical applications, the calculation and analysis of PCI will be more stable, and the change of the result will also be smoother. Using for averaging ensures fair consideration of each processing condition and can reduce the influence of individual extreme values on the comprehensive index.

[0082] The advantage of step 2 is that by systematically collecting the concentrations of volatile organic compounds, planting environment data, and processing condition data in the samples, the influencing factors of beany flavor are comprehensively analyzed from multiple dimensions. This comprehensive data collection method can reveal the potential sources affecting the beany flavor, provide a more accurate basis, and make the subsequent analysis more scientific and systematic. Compared with the existing technology, this step eliminates the interference of different units and magnitudes through dimensionless processing and index calculation, enhancing the comparability of data and the accuracy of analysis.

[0083] In this solution, adopting this step can provide a rich data basis for subsequent correlation analysis and regression modeling, ensuring the reliability and effectiveness of the analysis results. The data processing and index calculation of the system make the relationships between various influencing factors clearer, thus providing practical guidance for optimizing the production and processing of soy products. This promoting effect not only enhances the understanding of the causes of beany flavor but also provides strong support for subsequent quality control and product improvement.

[0084] Step 3: Use the Pearson correlation coefficient method to analyze the correlations between the compound concentration index, the comprehensive environmental index, and the comprehensive processing index and the final sensory score, calculate the correlation coefficients respectively, and based on the analysis results, judge the influence degree of different indexes on the beany flavor;

[0085] In this embodiment, the Pearson correlation coefficient method is used to analyze the correlations between the compound concentration index, the comprehensive environmental index, and the comprehensive processing index and the final sensory score. The formula is as follows:

[0086]

[0087] where r is the Pearson correlation coefficient, X i represents the compound concentration index, the comprehensive environmental index, or the comprehensive processing index of the i-th sample, which is used to conduct correlation analysis with the corresponding final sensory score respectively, and Y i represents the final sensory score of the i-th sample, is the corresponding mean value, is the mean value of the final sensory score;

[0088] Based on the calculated Pearson correlation coefficient value r, analyze the relationships between each index and the final sensory score:

[0089] When r > 0, it indicates a positive correlation between this index and the final sensory score, that is, it means that when the index increases, the final sensory score also increases;

[0090] When r < 0, it indicates a negative correlation between this index and the final sensory score, meaning that when the index increases, the final sensory score decreases instead;

[0091] When r = 0, there is no correlation, that is, it means there is no linear relationship between this index and the final sensory score;

[0092] Among them, when |r| is closer to 1, it means the stronger the correlation.

[0093] The advantage of Step 3 lies in quantitatively analyzing the correlations between the compound concentration index, the environmental comprehensive index, the processing comprehensive index, and the final sensory score using the Pearson correlation coefficient method. This method can systematically reveal the influence degrees of various factors on the beany flavor, enabling researchers to identify the factors that significantly affect the beany flavor. Compared with the qualitative evaluation of the prior art, this analysis method provides more rigorous and quantitative research results, laying a solid theoretical foundation for subsequent optimization.

[0094] In this solution, adopting this step can effectively integrate data from different sources, ensuring the scientificity and reliability of the analysis results. By calculating the correlation coefficient, researchers can prioritize the factors that are highly correlated with the sensory score, and concentrate resources on solving the most critical problems during the production and processing of soy products. This promoting effect will improve the overall quality of soy products, help meet consumers' demands for sensory quality, and thus enhance market competitiveness.

[0095] In Step 3, the purpose of calculating the Pearson correlation coefficient is to initially understand the relationships between each independent variable, such as the compound concentration index, the environmental comprehensive index, and the processing comprehensive index, and the dependent variable, the final sensory score. If the correlation coefficient indicates a strong correlation between certain variables and the dependent variable, this will guide the subsequent multiple linear regression analysis.

[0096] Step 4: Adopt the multiple linear regression analysis method. Taking the final sensory score as the dependent variable and the compound concentration index, the environmental comprehensive index, and the processing comprehensive index as independent variables, establish a regression model. Vary each index one by one, observe the change degree of the final sensory score, and calculate the sensitivity index. Rank the influencing factors according to the sensitivity index of each factor to provide a basis for optimization;

[0097] In this embodiment, adopt the multiple linear regression analysis method. Taking the final sensory score as the dependent variable and the compound concentration index, the environmental comprehensive index, and the processing comprehensive index as independent variables, establish a regression model, and the expression is as follows:

[0098] M = β0 + β1CCI + β2ECI + β3PCI + ∈

[0099] In the formula, M is the final sensory score, CCI is the compound concentration index, ECI is the environmental comprehensive index, PCI is the processing comprehensive index, β1, β2, and β3 are the regression coefficients of their respective independent variables, β0 is the intercept, and ∈ is the error term;

[0100] Use Python for model fitting, obtain the regression coefficients and statistical significance, and calculate the coefficient of determination:

[0101]

[0102] where R 2 is the coefficient of determination, which is used to evaluate the goodness of fit of the regression model to the final sensory score. The larger the coefficient of determination, the higher the goodness of fit. M i is the final sensory score of the i-th sample, represents the predicted final sensory score of the i-th sample by the model, represents the mean value of the final sensory scores of all samples.

[0103] Calculate the sensitivity index. Specifically: Set the dependent variable as the final sensory score, and set the independent variables as the compound concentration index, the comprehensive environmental index, and the comprehensive processing index. Vary each independent variable one by one, record the initial value and the change value of each independent variable, as well as the initial value and the change value of the corresponding final sensory score. Calculate the sensitivity index of each independent variable according to the following formula:

[0104]

[0105] where SI z is the sensitivity index of the z-th independent variable, M old is the initial value of the final sensory score, ΔM represents the change value of the final sensory score, N old represents the initial value of the independent variable, ΔN represents the change value of the independent variable, z is the index of the independent variable, z ∈ {1, 2, 3}. When z = 1, the corresponding independent variable is the compound concentration index; when z = 2, the corresponding independent variable is the comprehensive environmental index; when z = 3, the corresponding independent variable is the comprehensive processing index;

[0106] When |SI| > 1, it indicates that the change of this independent variable leads to a relatively large change in the sensory score, indicating that this independent variable has a strong influence on the sensory score;

[0107] When |SI| < 1, it indicates that the change of this independent variable leads to a relatively small change in the sensory score, indicating that this index has a weak influence on the sensory score;

[0108] When |SI| = 1, it indicates that the influence of this index is moderate;

[0109] Arrange the sensitivity index values from high to low. According to the results of the sensitivity analysis, the researchers conduct in-depth research and optimization on the factors with high sensitivity indexes to improve the sensory quality of soy products.

[0110] The advantage of Step 4 is that through multiple linear regression analysis, a quantitative relationship can be systematically established between the final sensory score and multiple influencing factors. This method can not only evaluate the comprehensive influence of the compound concentration index, environmental comprehensive index, and processing comprehensive index on the soybean odor, but also provide the specific contribution of each factor to the sensory score. Compared with the single-factor analysis of the prior art, it can consider the interaction between various factors more comprehensively, improving the depth and accuracy of the analysis.

[0111] In this solution, adopting this step can provide a powerful prediction model for the overall analysis, enabling researchers to more accurately identify the key factors affecting the soybean odor. At the same time, by calculating the regression coefficients and sensitivity indicators, researchers can prioritize the factors that have the greatest impact on the sensory score, thus providing a practical basis for the production and processing optimization of soy products. This not only improves the sensory quality of soy products but also enhances their market competitiveness, making the products more in line with the taste requirements of consumers.

[0112] Please refer to Figure 2 , a system for analyzing the influencing factors of soybean odor in soybeans and their products, including:

[0113] A sample selection and scoring module, which is used to select several groups of soy products of the same variety and the same quality as samples, organize 10 professionally trained sensory evaluators to score the soybean odor of the samples, and take the average value as the final sensory score of the samples to form a sensory score data set;

[0114] A data collection and processing module, which is used to collect the concentration of volatile organic compounds in the samples, and at the same time obtain the planting environment data and processing condition data of the samples. After dimensionless processing of the final sensory score, volatile organic compound concentration, planting environment data, and processing condition data, calculate the compound concentration index, environmental comprehensive index, and processing comprehensive index;

[0115] A correlation analysis module, which is used to analyze the correlation between the compound concentration index, environmental comprehensive index, and processing comprehensive index and the final sensory score by using the Pearson correlation coefficient method, calculate the correlation coefficients respectively, and combine the analysis results to judge the influence degree of different indexes on the soybean odor;

[0116] A regression model and sensitivity module, which is used to adopt the multiple linear regression analysis method, with the final sensory score as the dependent variable and the compound concentration index, environmental comprehensive index, and processing comprehensive index as the independent variables, establish a regression model, change each index one by one, observe the change degree of the final sensory score, and calculate the sensitivity indicators, and rank the influencing factors according to the sensitivity indicators of each factor to provide a basis for optimization.

[0117] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0118] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art will realize that the units and algorithm steps of each example described in connection with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0119] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, and may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0120] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application.

Claims

1. A method for analyzing the influencing factors of the beany flavor in soybeans and their products, characterized in that, The specific steps include: Step 1: Select several groups of soy products of the same variety and the same quality as samples. Organize 10 professionally trained sensory evaluators to score the beany flavor of the samples, and take the average value as the final sensory score of the samples, forming a sensory score data set. Step 2: Collect the concentrations of volatile organic compounds in the samples. At the same time, obtain the planting environment data and processing condition data of the samples. After dimensionless processing of the final sensory score, the concentrations of volatile organic compounds, the planting environment data, and the processing condition data, calculate the compound concentration index, the environmental comprehensive index, and the processing comprehensive index. Step 3: Use the Pearson correlation coefficient method to analyze the correlations between the compound concentration index, the environmental comprehensive index, and the processing comprehensive index and the final sensory score, calculate the correlation coefficients respectively, and combine the analysis results to judge the influence degree of different indexes on the beany flavor. Step 4: Adopt the multiple linear regression analysis method. Take the final sensory score as the dependent variable, and the compound concentration index, the environmental comprehensive index, and the processing comprehensive index as the independent variables to establish a regression model. Vary each index one by one, observe the change degree of the final sensory score, and calculate the sensitivity index. Sort the influencing factors according to the sensitivity index of each factor to provide a basis for optimization.

2. The method for analyzing the influencing factors of the beany flavor of soybeans and their products according to claim 1, characterized in that: When scoring the beany flavor of the samples, the specific logic is as follows: The scoring standard is 0 - 10 points, 0 points means no beany flavor, and 10 points means a very strong beany flavor. Each evaluator independently scores each sample three times, and takes the average value as the final sensory score, records the sensory score data of each sample, and forms a sensory score data set S. S = {s1, s2, …, s n} where s n represents the sensory score of the nth sample, and n is the index of the sample.

3. The method for analyzing the influencing factors of the beany flavor of soybeans and their products according to claim 1, characterized in that: Collect the concentrations of volatile organic compounds in the samples, specifically: Use a gas chromatography - mass spectrometry (GC - MS) instrument to detect the concentrations of volatile organic compounds in the samples. The volatile organic compounds include hexanal, acetone, and hexanol. Each sample is detected three times repeatedly, and the average value is taken as the final concentration value of each volatile organic compound. The planting environment data includes the cumulative rainfall, the cumulative fertilization amount, and the soil humidity. The processing condition data includes the processing time, the processing temperature at the end of processing, and the soil pH value at the end of processing. And for the same sample, the raw materials used are all in the same planting environment.

4. The method for analyzing the influencing factors of the beany flavor of soybeans and their products according to claim 1, characterized in that: When calculating the compound concentration index, the formula is as follows: where CCI is the compound concentration index, and C a is the concentration of the a-th volatile organic compound, and a is the index of the volatile organic compound. When a = 1, it corresponds to hexanal; when a = 2, it corresponds to acetone; when a = 3, it corresponds to hexanol. When calculating the environmental comprehensive index, the formula is as follows: In the formula, ECI is the comprehensive environmental index, and E b is the value of the b-th environmental factor, where b is the index of the environmental factor. When b = 1, it corresponds to the cumulative rainfall; when b = 2, it corresponds to the cumulative fertilization amount; when b = 3, it corresponds to the soil humidity. α is used to adjust the influence degree of the index to ensure that the result is dimensionless, and 0 < α < 1; When calculating the processing comprehensive index, the formula is as follows: where PCI is the processing comprehensive index, p k is the value of the k-th processing condition, k is the index of the processing condition, where when k = 1, it corresponds to the processing time, when k = 2, it corresponds to the processing temperature at the end of processing, when k = 3, it is the processing temperature at the end of processing, γ is used to adjust the influence degree of the index to ensure that the result is dimensionless, and 0 < γ < 1.

5. The method for analyzing the influencing factors of the beany flavor of soybeans and their products according to claim 4, wherein: When using the Pearson correlation coefficient method to analyze the correlations between the compound concentration index, the environmental comprehensive index, and the processing comprehensive index and the final sensory score, the formula is as follows: Among them, r is the Pearson correlation coefficient, X i represents the compound concentration index, environmental comprehensive index or processing comprehensive index of the i-th sample, and is used to perform correlation analysis with the corresponding final sensory score respectively. Y i represents the final sensory score of the i-th sample, is the corresponding mean value, is the mean value of the final sensory score; Based on the calculated Pearson correlation coefficient value r, analyze the relationships between each index and the final sensory score: When r > 0, it indicates a positive correlation between this index and the final sensory score, that is, it means that when the index increases, the final sensory score also increases; When r < 0, it indicates a negative correlation between this index and the final sensory score, meaning that when the index increases, the final sensory score decreases instead; When r = 0, there is no correlation, that is, it means there is no linear relationship between this index and the final sensory score; Among them, when |r| is closer to 1, it means the stronger the correlation.

6. A method for analyzing influencing factors of soybean and its product beany flavor according to claim 1, characterized in that: Using the multiple linear regression analysis method, with the final sensory score as the dependent variable and the compound concentration index, environmental comprehensive index, and processing comprehensive index as independent variables, a regression model is established, and the expression is as follows: M = β0 + β1CCI + β2ECI + β3PCI + ∈ In the formula, M is the final sensory score, CCI is the compound concentration index, ECI is the environmental comprehensive index, PCI is the processing comprehensive index, β1, β2, and β3 are the regression coefficients of their respective independent variables, β0 is the intercept, and ∈ is the error term; Use Python for model fitting, obtain the regression coefficients, and calculate the coefficient of determination: where R 2 is the coefficient of determination, which is used to evaluate the goodness of fit of the regression model to the final sensory score, M i is the final sensory score of the i-th sample, represents the predicted final sensory score of the i-th sample by the model, represents the mean of the final sensory scores of all samples.

7. A method for analyzing influencing factors of soybean and its product beany flavor according to claim 1, characterized in that: Calculate the sensitivity index, specifically: set the dependent variable as the final sensory score, set the independent variables as the compound concentration index, environmental comprehensive index, and processing comprehensive index, change each independent variable one by one, record the initial value and change value of each independent variable, and the initial value and change value of the corresponding final sensory score, and calculate the sensitivity index of each independent variable according to the following formula: where SI z is the sensitivity index of the z-th independent variable, M old is the initial value of the final sensory score, ΔM represents the change value of the final sensory score, N old represents the initial value of the independent variable, ΔN represents the change value of the independent variable, z is the index of the independent variable, z ∈ {1, 2, 3}, when z = 1, the corresponding independent variable is the compound concentration index, when z = 2, the corresponding independent variable is the environmental comprehensive index, and when z = 3, the corresponding independent variable is the processing comprehensive index; When |SI| > 1, it indicates that the change of this independent variable causes a relatively large change in the sensory score, indicating that this independent variable has a strong influence on the sensory score; When |SI| < 1, it indicates that the change of this independent variable causes a relatively small change in the sensory score, indicating that this index has a weak influence on the sensory score; When |SI| = 1, it means the influence of this index is moderate; Arrange the sensitivity index values from high to low. According to the results of the sensitivity analysis, researchers conduct in-depth research and optimization on the factors with high sensitivity indexes to improve the sensory quality of soy products.

8. An analysis system for influencing factors of soybean and soybean product beany flavor, characterized in that: The described system for analyzing influencing factors of soybean and its product beany flavor is used to execute the method for analyzing influencing factors of soybean and its product beany flavor according to any one of claims 1 - 7, including: A sample selection and scoring module, used to select several groups of soy products of the same variety and the same quality as samples, organize 10 professionally trained sensory evaluators to score the beany flavor of the samples, and take the average value as the final sensory score of the samples to form a sensory score data set; A data collection and processing module, used to collect the concentration of volatile organic compounds in the samples, and at the same time obtain the planting environment data and processing condition data of the samples. After dimensionless processing of the final sensory score, volatile organic compound concentration, planting environment data, and processing condition data, calculate the compound concentration index, environmental comprehensive index, and processing comprehensive index; The correlation analysis module is used to analyze the correlations between the compound concentration index, the environmental comprehensive index, the processing comprehensive index and the final sensory score by using the Pearson correlation coefficient method, calculate the correlation coefficients respectively, and combine the analysis results to judge the influence degree of different indexes on the beany flavor; The regression model and sensitivity module is used to adopt the multiple linear regression analysis method, with the final sensory score as the dependent variable and the compound concentration index, the environmental comprehensive index and the processing comprehensive index as the independent variables, establish a regression model, change each index one by one, observe the change degree of the final sensory score, calculate the sensitivity index, and rank the influencing factors according to the sensitivity index of each factor to provide a basis for optimization.

Images