Bean product fermentation influence factor analysis method and system

Abstract

The invention provides a bean product fermentation influence factor analysis method and system, and relates to the technical field of food safety detection.The method comprises the steps that multiple sets of key parameters are set within a preset range, physical and chemical index changes under the parameters are monitored in real time, and the physical and chemical index changes comprise the tannin variation, the acid production capacity, the fermentation rate and the active enzyme variation; the method comprises the following steps: establishing a key parameter, constructing a physical and chemical index prediction model, training by taking the key parameter as input, evaluating a fermented bean product, training a sensory evaluation index prediction model by taking the physical and chemical index as input, and finally, determining an optimal combination of the key parameter through a genetic algorithm by taking maximization of the sensory evaluation index as a target, thereby optimizing a fermentation process. By setting and adjusting key fermentation parameters in real time and combining physical and chemical index monitoring and sensory evaluation, the purposes of optimizing the fermentation process and improving the final quality of bean products are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of food safety detection, and in particular to a method and system for analyzing factors affecting fermentation of soy products. Background Art

[0002] As an important food category, soy products have always attracted widespread attention due to their rich nutritional value and diverse flavors. In recent years, research and technology on the fermentation process of soy products and their influencing factors have developed rapidly. Traditional soy product fermentation usually relies on natural fermentation or the use of specific fermentation strains such as lactic acid bacteria and yeast to improve the flavor and texture of soy products. These methods are still widely used in small-scale manual production, and the development of modern technology has made the fermentation process more controllable and efficient. Automated equipment, precision control systems and bioengineering technology have been introduced into the fermentation process of soy products. From simple environmental control to complex biological process control, these technologies make the fermentation process more stable.

[0003] In the prior art, publication number CN103487398A discloses a method for analyzing lysine fermentation broth, which comprises collecting near-infrared spectra of a plurality of lysine fermentation broth standard samples; obtaining the lysine content, OD value of the fermentation strain, reducing sugar content and amino nitrogen content in the plurality of standard samples respectively; obtaining multivariate linear regression equations of the lysine content, OD value of the fermentation strain, reducing sugar content and amino nitrogen content and their corresponding near-infrared absorbance values ​​respectively; and calculating the lysine content, OD value of the fermentation strain, reducing sugar content and amino nitrogen content in the lysine fermentation broth sample to be tested.

[0004] Insufficient existing technology:

[0005] First, the prior art mainly relies on multiple linear regression equations to predict the content of lysine in lysine fermentation broth, the OD value of the fermentation strain, the content of reducing sugars and the content of amino nitrogen. The disadvantage of this method is that the linear regression model has certain limitations, and it assumes that there is a linear relationship between variables. However, in the actual fermentation process, there may be complex nonlinear relationships and interactions between multiple influencing factors. Simple linear regression models are difficult to accurately capture and describe these complex relationships, thereby affecting the prediction accuracy and reliability of the model.

[0006] Secondly, the multivariate linear regression model in the prior art needs to be established based on a large amount of standard sample data. When fermentation conditions such as temperature, pH, stirring speed, etc. or environmental conditions change, the prediction performance of the model may decrease and need to be recalibrated and adjusted. In addition, this method cannot effectively handle the dynamic changes and noise in the fermentation process, and has high requirements for real-time monitoring and optimization, which may lead to problems such as difficulty in adjustment, increased costs, and poor model stability in practical applications.

[0007] Therefore, it is necessary to provide a method and system for analyzing factors affecting soy product fermentation to solve the above problems.

[0008] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0009] The purpose of the present invention is to provide a method and system for analyzing factors affecting soy product fermentation, so as to solve the problems raised in the above-mentioned background technology.

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] A method for analyzing factors affecting soy product fermentation, comprising the following specific steps:

[0012] Step 1: setting multiple key parameters of fermentation based on preset ranges, including temperature, humidity, pH value and fermentation time, wherein the preset ranges are: temperature between 20°C and 40°C, humidity between 50% and 90%, pH value between 4.5 and 7.5, and fermentation time between 12 hours and 72 hours;

[0013] Step 2: Real-time monitoring of the physical and chemical indicators corresponding to each group of key parameters during the fermentation process, including tannin change, acid production capacity, fermentation rate, and active enzyme change;

[0014] Step 3: Establish a physical and chemical index prediction model, take key parameters as input, and use the corresponding physical and chemical indexes as output labels to train the physical and chemical index prediction model;

[0015] Step 4: Evaluate the soy products obtained by fermentation based on each group of key parameters, use the physical and chemical indicators as input, and the corresponding sensory evaluation index as output label to train the sensory evaluation index prediction model;

[0016] Step 5: Taking the maximization of sensory evaluation index as the optimization goal, determine the optimal key parameters based on genetic algorithm.

[0017] Furthermore, the change in tannin content in beans was calculated, and the change in tannin was used as the first indicator affecting fermentation, based on the following method:

[0018] Lactic acid bacteria were selected as fermentation microorganisms. The beans were cleaned, soaked, cooked, and dried to constant weight. The bean samples were ground into powder. The ground bean powder was dissolved in 30 ml of 70% ethanol at a ratio of 1 g bean powder. Ultrasonic extraction was performed for 30 minutes and filtered. The tannin content of the extract was determined using the phenolphthalein reagent method. The formula used was:

[0019]

[0020] Y1=ΔDN=DN 初始 -DN 最终

[0021] Among them, DN represents tannin content, A represents absorbance, measured at a wavelength of 765nm, D represents dilution multiple, i.e., the dilution ratio of the extract, C represents the concentration of the extract, M represents the sample weight, Y1 represents the first indicator affecting fermentation, ΔDNM represents the change in tannin, DNM 初始 DNM 最终 Respectively represent the tannin content before and after ultrasonic extraction.

[0022] Furthermore, the fermentation rate and acid production capacity are evaluated, and the fermentation rate and acid production capacity are used as the second and third indicators respectively, according to the following method:

[0023] For the determination of fermentation rate and acid production capacity, firstly, the bean samples were dissolved in water to form a sample solution with a concentration of 10%, and the sample solution was sterilized by high pressure, inoculated with a lactic acid bacteria solution with the same concentration of 1%, and fermented in a constant temperature box at a temperature of 37%. The pH value of the sample solution before and after fermentation was measured, and the fermentation rate was determined by measuring the change in pH value. The formula based on this is:

[0024]

[0025] Among them, Y2 represents the second index affecting fermentation, namely, fermentation rate v, PH 初始 Indicates the pH value of the sample solution before fermentation. 最终 represents the pH value of the sample solution after fermentation, and Δt represents the fermentation time interval;

[0026] Use titration to determine the total acidity of the sample, record the volume of sodium hydroxide solution required for titration equilibrium, and calculate the acid production capacity of the sample solution based on the following method:

[0027]

[0028] Among them, Y3 represents the third index affecting fermentation, namely, acid production capacity c, V NaOH is the volume of NaPH solution used to neutralize the acidity of the sample solution, C NaPH is the concentration of NaPH solution used for titration, G represents the molar mass of lactic acid, which is 90.08, and V l Indicates the volume of the sample solution.

[0029] Furthermore, the change in active enzyme was calculated and used as the fourth indicator affecting fermentation, according to the following method:

[0030] The chemiluminescence method was used to determine the initial enzyme content in the sample solution and the final enzyme content after fermentation. Specifically, the initial luminescence intensity and the final luminescence intensity of the sample solution were determined by a photomultiplier tube, and a standard curve relationship between the luminescence intensity and the enzyme content was established. The method was based on:

[0031]

[0032] Y4=ΔE=E 最终 -E 初始

[0033] Among them, E 初始 、E 最终 represent the initial and final contents of the enzyme in the sample solution, respectively, and I 初始 ,I 最终 They represent the initial light intensity and final light intensity in the sample solution respectively, a is the slope in the standard curve, which represents the rate of change of luminescence intensity with enzyme content, b represents the intercept of the standard curve, that is, the basic value of luminescence intensity when the enzyme content is zero, ΔE represents the change in enzyme content in the sample solution, and Y4 represents the fourth indicator affecting fermentation.

[0034] Furthermore, a prediction model is constructed with physical and chemical indicators as input and the corresponding sensory evaluation index as output label. The physical and chemical indicator prediction model is trained based on the following formula:

[0035] Q g =β0+β1Y1+β2Y2+β3Y3+β4Y4+∈ q

[0036] Among them, Q g represents the sensory evaluation index, Y1, Y2, Y3, and Y4 are various physical and chemical indicators, β0 is the intercept term in the regression equation, β1, β2, β3, and β4 represent the weight proportional coefficients corresponding to their respective indicators, ∈ is the error term in the regression equation, and β4>β1>β2>β3.

[0037] Furthermore, taking the maximization of sensory evaluation index as the optimization goal, the optimal key parameters were determined based on the genetic algorithm, and the method was based on:

[0038] A plurality of groups of key parameters are randomly generated within a predetermined range to form an initial population, the individuals in the population are input into a physicochemical index prediction model to obtain the physicochemical index corresponding to the individual, the physicochemical index is input into a sensory evaluation model to obtain the sensory evaluation index corresponding to the individual, the individuals in the initial population are sorted according to the sensory evaluation index, the individuals in the front row are selected as the parent generation, crossover and mutation operations are performed to form new individuals, the new individuals and the parent generation are merged to form a new population, and the physicochemical index, the sensory evaluation index, the selection, crossover and mutation operations are performed again until a predetermined number of iterations are completed, and the individual corresponding to the maximum value of the sensory evaluation index is selected as the optimal key parameter.

[0039] The present invention also provides a soy product fermentation influencing factor analysis system, the analysis system is used to execute the above-mentioned soy product fermentation influencing factor analysis method, comprising:

[0040] A fermentation parameter setting module, wherein the fermentation parameter setting module sets multiple groups of key parameters of fermentation based on a preset range, including temperature, humidity, pH value and fermentation time, wherein the preset range is: temperature is between 20°C and 40°C, humidity is between 50% and 90%, pH value is between 4.5 and 7.5, and fermentation time is between 12 hours and 72 hours;

[0041] A real-time monitoring module, which is used to monitor in real time the physical and chemical indicators corresponding to each group of key parameters during the fermentation process, including tannin change, acid production capacity, fermentation rate, and active enzyme change;

[0042] A physicochemical prediction model module, which is used to establish a physicochemical index prediction model, take key parameters as input, take corresponding physicochemical indexes as output labels, and train the physicochemical index prediction model;

[0043] A sensory evaluation index generation module, which is used to evaluate the soy products obtained by fermentation based on each group of key parameters, using physical and chemical indicators as input and corresponding sensory evaluation indexes as output labels to train a sensory evaluation index prediction model;

[0044] The parameter optimization module takes the maximization of the sensory evaluation index as the optimization goal and determines the optimal key parameters based on the genetic algorithm.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention first improves the limitations of the traditional multiple linear regression model by introducing a comprehensive real-time monitoring system and dynamic analysis of multiple physical and chemical indicators. During the fermentation process, in addition to basic parameters such as temperature, humidity, pH value and fermentation time, the change in tannin, acid production capacity, fermentation rate and active enzyme change are also monitored. The comprehensive analysis and real-time adjustment of these physical and chemical indicators make the fermentation conditions more flexible and controllable, and can more accurately reflect the dynamic changes and complex interactions in the fermentation process, thereby improving the accuracy and applicability of the prediction model;

[0047] Secondly, the present invention not only relies on the dynamic adjustment of real-time monitoring data, but also incorporates sensory evaluation and nutritional analysis into the evaluation system. By conducting sensory evaluation on the soy products after fermentation, a comprehensive evaluation is conducted from multiple angles such as flavor, taste and nutritional value to ensure the quality of the final product and user satisfaction. Multiple indicators such as tannin change, fermentation rate, acid production capacity and active enzyme change are calculated, and through standardized processing and statistical analysis, the linear relationship between each indicator and a comprehensive evaluation model are further established. Such a multi-step correction and evaluation method not only eliminates the influence between different dimensions, but also analyzes the comprehensive influence of various physical and chemical indicators on sensory evaluation through multivariate linear regression, thereby comprehensively improving the accuracy of fermentation process control and the reliability of results;

[0048] The present invention optimizes the fermentation process control and improves the prediction accuracy and the quality of the final product by introducing a comprehensive real-time monitoring system and dynamic analysis and adjustment of multiple physical and chemical indicators, and combining sensory evaluation and nutrient component analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the overall method flow of the present invention.

[0050] Figure 2 It is a schematic diagram of the system module flow of the present invention. DETAILED DESCRIPTION

[0051] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0052] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0053] Example:

[0054] See also Figure 1 , the present invention provides a technical solution:

[0055] A method for analyzing factors affecting soy product fermentation, comprising the following specific steps:

[0056] Step 1: setting multiple key parameters of fermentation based on preset ranges, including temperature, humidity, pH value and fermentation time, wherein the preset ranges are: temperature between 20°C and 40°C, humidity between 50% and 90%, pH value between 4.5 and 7.5, and fermentation time between 12 hours and 72 hours;

[0057] Step 2: Real-time monitoring of the physical and chemical indicators corresponding to each group of key parameters during the fermentation process, including tannin change, acid production capacity, fermentation rate, and active enzyme change;

[0058] Step 3: Establish a physical and chemical index prediction model, take key parameters as input, and use the corresponding physical and chemical indexes as output labels to train the physical and chemical index prediction model;

[0059] Step 4: Evaluate the soy products obtained by fermentation based on each group of key parameters, use the physical and chemical indicators as input, and the corresponding sensory evaluation index as output label to train the sensory evaluation index prediction model;

[0060] Step 5: Taking the maximization of sensory evaluation index as the optimization goal, determine the optimal key parameters based on genetic algorithm.

[0061] It should be noted that by selecting lactic acid bacteria as fermentation microorganisms, the phenolphthalein reagent method was used to determine the changes in tannin content in beans, and the change in tannin was used as the first indicator affecting fermentation. This process can evaluate the inhibitory effect of tannin on fermentation, while providing key data support for optimizing fermentation conditions and providing accurate real-time data for the fermentation process, so that the fermentation conditions can be dynamically adjusted according to the change in tannin, thereby effectively controlling the inhibitory effect of tannin on fermentation microorganisms, ensuring the stability of the fermentation process and improving product quality.

[0062] Therefore, it is necessary to calculate the change in tannin content in beans and use the change in tannin as the first indicator affecting fermentation. The method is as follows:

[0063] Lactic acid bacteria were selected as fermentation microorganisms. The beans were cleaned, soaked, cooked, and dried to constant weight. The bean samples were ground into powder. The ground bean powder was dissolved in 30 ml of 70% ethanol at a ratio of 1 g bean powder. Ultrasonic extraction was performed for 30 minutes and filtered. The tannin content of the extract was determined using the phenolphthalein reagent method. The formula used was:

[0064]

[0065] Y1=ΔDN=DN 初始 -DN 最终

[0066] Among them, DN represents tannin content, A represents absorbance, measured at a wavelength of 765nm, D represents dilution multiple, i.e., the dilution ratio of the extract, C represents the concentration of the extract, M represents the sample weight, Y1 represents the first indicator affecting fermentation, ΔDN represents the change in tannin, DN 初始 DN 最终 Respectively represent the tannin content before and after ultrasonic extraction.

[0067] It should be noted that by measuring the fermentation rate and acid production capacity, which are used as the second and third indicators respectively, the effect and efficiency of the fermentation process can be comprehensively evaluated, and the optimization of fermentation conditions can be guided. By monitoring the changes in pH value, the activity and reproduction rate of lactic acid bacteria in the fermentation process can be reflected, and the progress of fermentation can be understood in real time to ensure the efficiency of the fermentation process. By measuring the acidity of the sample, the ability of lactic acid bacteria to produce acid during the fermentation process can be evaluated, reflecting the degree of consumption of the substrate by lactic acid bacteria and the generation of metabolites.

[0068] Therefore, it is necessary to evaluate its fermentation rate and its acid production capacity, and use the fermentation rate and acid production capacity as the second and third indicators respectively, based on the following method:

[0069] For the determination of fermentation rate and acid production capacity, firstly, the bean samples were dissolved in water to form a sample solution with a concentration of 10%, and the sample solution was sterilized by high pressure, inoculated with a lactic acid bacteria solution with the same concentration of 1%, and fermented in a constant temperature box at a temperature of 37%. The pH value of the sample solution before and after fermentation was measured, and the fermentation rate was determined by measuring the change in pH value. The formula based on this is:

[0070]

[0071] Among them, Y2 represents the second index affecting fermentation, namely, fermentation rate v, PH 初始 Indicates the pH value of the sample solution before fermentation. 最终 represents the pH value of the sample solution after fermentation, and Δt represents the fermentation time interval;

[0072] Use titration to determine the total acidity of the sample, record the volume of sodium hydroxide solution required for titration equilibrium, and calculate the acid production capacity of the sample solution based on the following method:

[0073]

[0074] Among them, Y3 represents the third index affecting fermentation, namely, acid production capacity c, V NaOH is the volume of VNaOH solution used to neutralize the acidity of the sample solution, C NaOH V is the concentration of NaOH solution used for titration, G represents the molar mass of lactic acid, which is 90.08, and V l Indicates the volume of the sample solution.

[0075] It should be noted that the chemiluminescence method has high sensitivity and can detect very low concentrations of enzyme activity. This high precision can provide accurate measurement of small changes in enzyme activity and ensure the reliability of data. By establishing a standard curve of luminescence intensity and enzyme content, the results can be standardized to ensure that the measurement data of different batches are comparable. Monitoring changes in enzyme activity can provide a deep understanding of the role and changes of enzymes in the fermentation process, thereby optimizing fermentation parameters and improving fermentation performance and product quality.

[0076] Therefore, it is necessary to calculate the change in active enzyme and use it as the fourth indicator affecting fermentation. The method is based on:

[0077] The chemiluminescence method was used to determine the initial enzyme content in the sample solution and the final enzyme content after fermentation. Specifically, the initial luminescence intensity and the final luminescence intensity of the sample solution were determined by a photomultiplier tube, and a standard curve relationship between the luminescence intensity and the enzyme content was established. The method was based on:

[0078]

[0079] Y4=ΔE=E 最终 -E 初始

[0080] Among them, E 初始 、E 最终 represent the initial and final contents of the enzyme in the sample solution, respectively, and I 初始 ,I 最终 They represent the initial light intensity and final light intensity in the sample solution respectively, a is the slope in the standard curve, which represents the rate of change of luminescence intensity with enzyme content, b represents the intercept of the standard curve, that is, the basic value of luminescence intensity when the enzyme content is zero, ΔE represents the change in enzyme content in the sample solution, and Y4 represents the fourth indicator affecting fermentation.

[0081] It should be noted that the construction of the relevant prediction model structure of physical and chemical indicators takes support vector machine regression SVR as the main structural model, sets the physical and chemical indicators as the output parameters of the model structure, and takes the corresponding key parameters as the output labels. Support vector machine SVR regression is suitable for processing data with complex nonlinear relationships. By using kernel methods such as radial basis function kernels or polynomial kernels, SVR can effectively capture the nonlinear relationship between input features and output labels. The model takes physical and chemical indicators as input parameters and outputs corresponding key parameters. In the training process, the training set data is used to optimize the support vector machine regression model, and the parameters of the model are iteratively adjusted to enable it to accurately fit the training data. The relationship between physical and chemical indicators and key parameters. The Huber loss function is selected as the optimization target in model training. The Huber loss combines the advantages of mean square error and absolute error. When the error is small, it is similar to the mean square error. When the error is large, it turns to linear growth, which makes the model more robust to outliers and reduces the negative impact of outliers on model training. After the training is completed, the model is evaluated using the test set to verify the model's predictive ability for unseen data, and the generalization ability and actual effect of the model are judged by its performance on the test set. When the change of the loss function is less than the preset threshold in several consecutive iterations, it means that the model has basically converged and training can be stopped.

[0082] It should be noted that the structure of the linear regression model is used to construct the prediction sensory evaluation index model. gThe output parameter of the model structure is used, and the corresponding physical and chemical indicators are used as output labels. In order to train this model, the data set is first divided into a training set and a test set. Usually 80% of the data is used for training and 20% of the data is used for testing. During the training process, the training set is used to learn the model parameters, and the mean square error MSE is selected as the loss function to measure the difference between the model prediction value and the actual value. The model training process optimizes the parameters by minimizing the MSE to make the prediction result closer to the true value. After the training is completed, the test set is used to evaluate the performance of the model. If the performance of the model on the test set reaches the expected performance indicator, such as the MSE on the test set is significantly low and stable, and the model meets the preset weight conditions, it can be considered that the model training is completed and has good generalization ability, or a reasonable maximum number of iterations is set. When the change of the loss function is less than the preset threshold within a certain number of iterations, it can be considered that the model has converged, and the training can be stopped or the loss on the test set can be monitored through the early stopping mechanism. If the loss no longer improves after a certain number of iterations, the training is stopped in advance.

[0083] Therefore, it is necessary to build a prediction model with physical and chemical indicators as input and the corresponding sensory evaluation index as output label, and train the physical and chemical indicator prediction model based on the formula:

[0084] Q g =β0+β1Y1+β2Y2+β3Y3+β4Y4+∈ q

[0085] Among them, Q g represents the sensory evaluation index, Y1, Y2, Y3, and Y4 are various physical and chemical indicators, β0 is the intercept term in the regression equation, β1, β2, β3, and β4 represent the weight ratio coefficients corresponding to their respective indicators, ∈ is the error term in the regression equation, and β4>β1>β2>β3; in the above formula, the reason why the weight ratio is set to β4>β1>β2>β3 is that the change in active enzyme contributes most significantly to the sensory evaluation of the final fermented product. Active enzymes catalyze various biochemical reactions during the fermentation process, and their activity changes have a great influence on the product's aroma, taste and other sensory characteristics. Therefore, the weight ratio of the change in active enzyme β4 is the largest; while tannin has a direct impact on the astringency and taste of the fermented product, although the impact is large, it is usually smaller than the change in active enzyme, so the weight ratio of tannin change β1 is relatively high, but lower than the change in active enzyme; acid production capacity affects the acidity of the fermented product, and acidity is one of the important aspects of sensory evaluation, but its influence is second to tannin and active enzyme changes, so the weight ratio of acid production capacity β2 is low, and the fermentation rate affects the overall fermentation time and process, but its direct impact on the final sensory quality is relatively indirect or small. Therefore, the weight ratio of fermentation rate β3 is the lowest among all indicators.

[0086] It should be noted that, when sorting the individuals in the initial population according to the sensory evaluation index, the top 20% of the individuals are selected as the parents. This can avoid slow convergence due to too many parents or too fast convergence due to too few samples, and provide sufficient diversity to explore the parameter space, and retain a certain amount of high-quality genes to facilitate the development of the best solution. In most cases, this ratio can better balance the exploration and development of the algorithm. When performing crossover and mutation operations, be careful not to exceed the predetermined range, otherwise it will cause deviations in the experimental results. Since the number of samples in this prediction model is small, the iteration range is selected between 50 and 100 times, and the changing trend of the fitness value is observed in the initial run. If the fitness value tends to be stable at an early stage, too many iterations may not be required. If the fitness value continues to improve, you can consider increasing the number of iterations.

[0087] Therefore, it is necessary to take the maximization of sensory evaluation index as the optimization goal and determine the optimal key parameters based on genetic algorithm. The method is as follows:

[0088] Randomly generate multiple groups of key parameters within a predetermined range to form an initial population, input individuals in the population into a physical and chemical index prediction model to obtain the physical and chemical index corresponding to the individual, input the physical and chemical index into a sensory evaluation model to obtain the sensory evaluation index corresponding to the individual, sort the individuals in the initial population according to the sensory evaluation index, select the individuals in the front row as the parent generation, perform crossover and mutation operations to form new individuals, merge the new individuals and the parent generation to form a new population, and calculate the physical and chemical index, calculate the sensory evaluation index, select, crossover and mutation operations again until the predetermined number of iterations is completed, and select the individual corresponding to the maximum value of the sensory evaluation index as the optimal key parameter

[0089] See also Figure 2 The present invention also provides a soy product fermentation influencing factor analysis system, the analysis system is used to execute the above-mentioned soy product fermentation influencing factor analysis method, comprising:

[0090] A fermentation parameter setting module, wherein the fermentation parameter setting module sets multiple groups of key parameters of fermentation based on a preset range, including temperature, humidity, pH value and fermentation time, wherein the preset range is: temperature is between 20°C and 40°C, humidity is between 50% and 90%, pH value is between 4.5 and 7.5, and fermentation time is between 12 hours and 72 hours;

[0091] A real-time monitoring module, which is used to monitor in real time the physical and chemical indicators corresponding to each group of key parameters during the fermentation process, including tannin change, acid production capacity, fermentation rate, and active enzyme change;

[0092] A physicochemical prediction model module, which is used to establish a physicochemical index prediction model, take key parameters as input, take corresponding physicochemical indexes as output labels, and train the physicochemical index prediction model;

[0093] A sensory evaluation index generation module, which is used to evaluate the soy products obtained by fermentation based on each group of key parameters, using physical and chemical indicators as input and corresponding sensory evaluation indexes as output labels to train a sensory evaluation index prediction model;

[0094] The parameter optimization module takes the maximization of the sensory evaluation index as the optimization goal and determines the optimal key parameters based on the genetic algorithm.

[0095] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0096] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0097] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0098] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for analyzing factors affecting soy product fermentation, characterized in that: The specific steps include: Step 1: setting multiple key parameters of fermentation based on preset ranges, including temperature, humidity, pH value and fermentation time, wherein the preset ranges are: temperature between 20°C and 40°C, humidity between 50% and 90%, pH value between 4.5 and 7.5, and fermentation time between 12 hours and 72 hours; Step 2: Real-time monitoring of the physical and chemical indicators corresponding to each group of key parameters during the fermentation process, including tannin change, acid production capacity, fermentation rate, and active enzyme change; Step 3: Establish a physical and chemical index prediction model, take key parameters as input, and use the corresponding physical and chemical indexes as output labels to train the physical and chemical index prediction model; Step 4: Evaluate the soy products obtained by fermentation based on each group of key parameters, use the physical and chemical indicators as input, and the corresponding sensory evaluation index as output label to train the sensory evaluation index prediction model; Step 5: Taking the maximization of sensory evaluation index as the optimization goal, determine the optimal key parameters based on genetic algorithm.

2. The method for analyzing factors affecting soybean product fermentation according to claim 1, characterized in that: The method for calculating the change in tannin content in beans and taking the change in tannin as the first indicator affecting fermentation is as follows: Lactic acid bacteria were selected as fermentation microorganisms. The beans were cleaned, soaked, cooked, and dried to constant weight. The bean samples were ground into powder. The ground bean powder was dissolved in 30 ml of 70% ethanol at a ratio of 1 g bean powder. Ultrasonic extraction was performed for 30 minutes and filtered. The tannin content of the extract was determined using the phenolphthalein reagent method. The formula used was: Y1=ΔDN=DN 初始 -DN 最终 Among them, DN represents tannin content, A represents absorbance, measured at a wavelength of 765nm, D represents dilution multiple, i.e., the dilution ratio of the extract, C represents the concentration of the extract, M represents the sample weight, Y1 represents the first indicator affecting fermentation, ΔDN represents the change in tannin, DN 初始 DN 最终 Respectively represent the tannin content before and after ultrasonic extraction.

3. The method for analyzing factors affecting soybean product fermentation according to claim 1, characterized in that: The fermentation rate and acid production capacity are evaluated, and the fermentation rate and acid production capacity are used as the second and third indicators respectively, based on the following method: For the determination of fermentation rate and acid production capacity, firstly, the bean samples were dissolved in water to form a sample solution with a concentration of 10%, and the sample solution was sterilized by high pressure, inoculated with a lactic acid bacteria solution with the same concentration of 1%, and fermented in a constant temperature box at a temperature of 37%. The pH value of the sample solution before and after fermentation was measured, and the fermentation rate was determined by measuring the change in pH value. The formula based on this is: Among them, Y2 represents the second index affecting fermentation, namely, fermentation rate v, PH 初始 Indicates the pH value of the sample solution before fermentation. 最终 represents the pH value of the sample solution after fermentation, and Δt represents the fermentation time interval; Use titration to determine the total acidity of the sample, record the volume of sodium hydroxide solution required for titration equilibrium, and calculate the acid production capacity of the sample solution based on the following method: Among them, Y3 represents the third index affecting fermentation, namely, acid production capacity c, V NaOH is the volume of NaOH solution used to neutralize the acidity of the sample solution, C NaOH is the concentration of NaOH solution used for titration, G represents the molar mass of lactic acid, which is 90.08, and V l Indicates the volume of the sample solution.

4. The method for analyzing factors affecting soybean product fermentation according to claim 1, characterized in that: The change in active enzyme was calculated and used as the fourth indicator affecting fermentation, based on the following method: The chemiluminescence method was used to determine the initial enzyme content in the sample solution and the final enzyme content after fermentation. Specifically, the initial luminescence intensity and the final luminescence intensity of the sample solution were determined by a photomultiplier tube, and a standard curve relationship between the luminescence intensity and the enzyme content was established. The method was based on: Y4=ΔE=E 最终 -THAT 初始 Among them, E 初始 、E 最终 represent the initial and final contents of the enzyme in the sample solution, respectively, and I 初始 ,I 最终 They represent the initial light intensity and final light intensity in the sample solution respectively, a is the slope in the standard curve, which represents the rate of change of luminescence intensity with enzyme content, b represents the intercept of the standard curve, that is, the basic value of luminescence intensity when the enzyme content is zero, ΔE represents the change in enzyme content in the sample solution, and Y4 represents the fourth indicator affecting fermentation.

5. The method for analyzing factors affecting soybean product fermentation according to claim 1, characterized in that: Construct a prediction model, take the physical and chemical indicators as input, and use the corresponding sensory evaluation index as output label. Train the physical and chemical indicator prediction model based on the formula: Q g =β0+β1Y1+β2Y2+β3Y3+β4Y4+∈ q Among them, Q g represents the sensory evaluation index, Y1, Y2, Y3, and Y4 are various physical and chemical indicators, β0 is the intercept term in the regression equation, β1, β2, β3, and β4 represent the weight proportional coefficients corresponding to their respective indicators, ∈ is the error term in the regression equation, and β4>β1>β2>β3.

6. The method for analyzing factors affecting soybean product fermentation according to claim 1, characterized in that: Taking the maximization of sensory evaluation index as the optimization goal, the optimal key parameters are determined based on the genetic algorithm. The method is based on: A plurality of groups of key parameters are randomly generated within a predetermined range to form an initial population, the individuals in the population are input into a physicochemical index prediction model to obtain the physicochemical index corresponding to the individual, the physicochemical index is input into a sensory evaluation model to obtain the sensory evaluation index corresponding to the individual, the individuals in the initial population are sorted according to the sensory evaluation index, the individuals in the front row are selected as the parent generation, crossover and mutation operations are performed to form new individuals, the new individuals and the parent generation are merged to form a new population, and the physicochemical index, the sensory evaluation index, the selection, crossover and mutation operations are performed again until a predetermined number of iterations are completed, and the individual corresponding to the maximum value of the sensory evaluation index is selected as the optimal key parameter.

7. A soybean product fermentation influencing factor analysis system, characterized in that: The analysis system is used to execute the method for analyzing factors affecting fermentation of soy products according to any one of claims 1 to 6, comprising: A fermentation parameter setting module, wherein the fermentation parameter setting module sets multiple groups of key parameters of fermentation based on a preset range, including temperature, humidity, pH value and fermentation time, wherein the preset range is: temperature is between 20°C and 40°C, humidity is between 50% and 90%, pH value is between 4.5 and 7.5, and fermentation time is between 12 hours and 72 hours; A real-time monitoring module, which is used to monitor in real time the physical and chemical indicators corresponding to each group of key parameters during the fermentation process, including tannin change, acid production capacity, fermentation rate, and active enzyme change; A physicochemical prediction model module, which is used to establish a physicochemical index prediction model, take key parameters as input, take corresponding physicochemical indexes as output labels, and train the physicochemical index prediction model; A sensory evaluation index generation module, which is used to evaluate the soy products obtained by fermentation based on each group of key parameters, using physical and chemical indicators as input and corresponding sensory evaluation indexes as output labels to train a sensory evaluation index prediction model; The parameter optimization module takes the maximization of the sensory evaluation index as the optimization goal and determines the optimal key parameters based on the genetic algorithm.

Images