A method for regulating the fermentation rate to maintain the natural activity of fruits and vegetables and a fermentation device

Abstract

The present application provides a method and a fermentation device for regulating the fermentation speed of maintaining the natural activity of fruits and vegetables, which relates to the technical field of fruit and vegetable fermentation. It collects multiple types of bacterial communities corresponding to multiple liquid inlet ports of a fruit and vegetable fermentation tank; conducts an oxygen adaptability analysis based on the multiple types of bacterial communities, and outputs multiple groups of low-oxygen regulation parameters; controls multiple low-oxygen regulation sub-modules to activate the bacterial communities at the multiple liquid inlet ports according to the multiple groups of low-oxygen regulation parameters; constructs a mixed fermentation speed regulation model, inputs the types of bacterial communities in the mixer into the mixed fermentation speed regulation model for analysis of activity retention, and outputs mixed-low-oxygen regulation parameters; adjusts the oxygen in the mixer according to the mixed-low-oxygen regulation parameters. The present application solves the technical problems in the prior art that due to the lack of fine regulation of the oxygen demand during the fermentation process of multiple bacterial communities, the fermentation efficiency is low and the nutritional components are lost, and achieves the technical effect of improving the fermentation efficiency and retaining the natural components of fruits and vegetables.

Description

Technical Field

[0001] This application relates to the technical field of fruit and vegetable fermentation, and specifically relates to a method and equipment for regulating the fermentation speed while maintaining the natural activity of fruits and vegetables. Background Art

[0002] Fruit and vegetable fermentation is a processing technology that converts substances such as sugars and acids in fruits and vegetables into functional fermentation products through the action of microorganisms. It can not only extend the shelf life of fruits and vegetables, but also enhance their nutritional value and flavor characteristics. Existing methods for regulating fruit and vegetable fermentation mostly focus on the cultivation of single strains or simple microbial communities, and initially optimize the fermentation process by adjusting temperature, pH value, or oxygen supply. These methods have certain effects in traditional fermentation environments. However, with the application of multi-microbial community fermentation technology, traditional regulation methods can no longer meet complex regulation requirements. First, existing methods lack in-depth research and optimization of the co-fermentation of multiple strains, resulting in the insufficient exertion of the diversity and synergistic effects of microbial communities during the fermentation process, thereby affecting fermentation efficiency and product quality. Second, existing fermentation regulation technologies do not control key environmental factors such as oxygen precisely enough. Oxygen is an important factor affecting microbial metabolism and the generation of fermentation products, but traditional methods cannot be precisely regulated according to the specific needs of different microbial communities. This not only limits the improvement of fermentation efficiency, but also easily causes the loss of nutrients in fruits and vegetables under unfavorable fermentation conditions. Summary of the Invention

[0003] This application provides a method and equipment for regulating the fermentation speed while maintaining the natural activity of fruits and vegetables, solving the technical problems of low fermentation efficiency and nutrient loss in the prior art due to the lack of precise regulation of the oxygen demand during the multi-microbial fermentation process, and achieving the technical effect of improving fermentation efficiency and retaining the natural components of fruits and vegetables.

[0004] In view of the above problems, on the one hand, the present application provides a method for regulating the fermentation rate to maintain the natural activity of fruits and vegetables. The method includes: obtaining a fruit and vegetable fermentation tank, which includes a plurality of liquid inlet ports and a mixer. Each liquid inlet port corresponds to a low-oxygen regulation sub-module, and the mixer is provided with a low-oxygen regulation main module; collecting a plurality of bacterial community types corresponding to fermentation for the plurality of liquid inlet ports; performing oxygen adaptability analysis according to the plurality of bacterial community types, and outputting multiple groups of low-oxygen regulation parameters, where each group of low-oxygen regulation parameters includes an oxygen concentration and a feeding rate; controlling a plurality of low-oxygen regulation sub-modules to activate the bacterial communities at the plurality of liquid inlet ports according to the multiple groups of low-oxygen regulation parameters, and the activated bacterial communities flow into the mixer through the plurality of liquid inlet ports; constructing a mixed fermentation rate regulation model, inputting the bacterial community types in the mixer into the mixed fermentation rate regulation model for activity retention analysis, and outputting mixed-low-oxygen regulation parameters; controlling the low-oxygen regulation main module to adjust the oxygen in the mixer according to the mixed-low-oxygen regulation parameters.

[0005] On the other hand, the present application also provides a fermentation device for maintaining the natural activity of fruits and vegetables. The device includes: a fermentation tank connection unit for obtaining a fruit and vegetable fermentation tank, which includes a plurality of liquid inlet ports and a mixer. Each liquid inlet port corresponds to a low-oxygen regulation sub-module, and the mixer is provided with a low-oxygen regulation main module; a bacterial community type collection unit for collecting a plurality of bacterial community types corresponding to fermentation for the plurality of liquid inlet ports; an oxygen adaptability analysis unit for performing oxygen adaptability analysis according to the plurality of bacterial community types and outputting multiple groups of low-oxygen regulation parameters, where each group of low-oxygen regulation parameters includes an oxygen concentration and a feeding rate; a low-oxygen regulation unit for controlling a plurality of low-oxygen regulation sub-modules to activate the bacterial communities at the plurality of liquid inlet ports according to the multiple groups of low-oxygen regulation parameters, and the activated bacterial communities flow into the mixer through the plurality of liquid inlet ports; an activity retention analysis unit for constructing a mixed fermentation rate regulation model, inputting the bacterial community types in the mixer into the mixed fermentation rate regulation model for activity retention analysis, and outputting mixed-low-oxygen regulation parameters; an oxygen regulation unit for controlling the low-oxygen regulation main module to adjust the oxygen in the mixer according to the mixed-low-oxygen regulation parameters.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0007] By collecting multiple liquid inlets for fermenting corresponding multiple types of microbial communities, the types of microbial communities involved in the fermentation process are clarified, providing input conditions for subsequent oxygen demand analysis and setting of regulation parameters. Oxygen adaptability analysis is carried out according to the multiple types of microbial communities, and multiple sets of low-oxygen regulation parameters are output to ensure the optimal growth conditions of the microbial communities during the activation stage. Control multiple low-oxygen regulation sub-modules to activate the microbial communities at the multiple liquid inlets according to the multiple sets of low-oxygen regulation parameters, ensuring that the microbial communities enter the mixed fermentation stage in a highly active state and improving the initial fermentation efficiency. Construct a mixed fermentation speed regulation model, input the types of microbial communities in the mixer into the mixed fermentation speed regulation model for activity retention analysis, achieve the balance between fermentation speed optimization and activity retention, and provide a decision-making basis for the mixed fermentation stage. Control the low-oxygen regulation main module to globally regulate the oxygen environment in the mixer according to the mixed-low-oxygen regulation parameters output by the mixed fermentation speed regulation model, ensuring that the fermentation process proceeds evenly and efficiently and retaining the natural activity of fruits and vegetables.

[0008] In summary, this application collects multiple types of microbial communities and conducts oxygen adaptability analysis, providing a suitable growth environment for different microbial communities, giving full play to the synergistic effect of multiple microbial communities, and promoting the generation and transformation of various beneficial components in fruits and vegetables. The precise regulation of the constructed mixed fermentation speed regulation model and the low-oxygen regulation main module further optimizes the fermentation conditions, improves the fermentation efficiency and the activity of the microbial communities, effectively reduces the loss of nutritional components of fruits and vegetables, and maintains the natural activity of fruits and vegetables.

[0009] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Brief Description of the Drawings

[0010] Figure 1 It is a flowchart of a fermentation speed regulation method for maintaining the natural activity of fruits and vegetables provided by an embodiment of this application.

[0011] Figure 2 It is a flowchart of obtaining multiple sets of low-oxygen regulation parameters in a fermentation speed regulation method for maintaining the natural activity of fruits and vegetables provided by an embodiment of this application.

[0012] Figure 3 It is a flowchart of constructing a mixed fermentation speed regulation model in a fermentation speed regulation method for maintaining the natural activity of fruits and vegetables provided by an embodiment of this application.

[0013] Figure 4 It is a structural diagram of a fermentation device for maintaining the natural activity of fruits and vegetables provided by an embodiment of this application.

[0014] Explanation of the reference numerals: fermenter connection unit 10 , bacterial flora type collection unit 20 , oxygen adaptability analysis unit 30 , hypoxia control unit 40 , activity retention analysis unit 50 , oxygen regulation unit 60 . DETAILED DESCRIPTION

[0015] The embodiments of the present application provide a fermentation speed control method and fermentation equipment for maintaining the natural activity of fruits and vegetables, thereby solving the technical problems of low fermentation efficiency and loss of nutrients in the prior art due to the lack of fine control of oxygen demand in the multi-bacteria fermentation process, thereby achieving the technical effect of improving fermentation efficiency and retaining the natural ingredients of fruits and vegetables.

[0016] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides a method for controlling the fermentation speed for maintaining the natural activity of fruits and vegetables, the method comprising:

[0017] Step S1: obtaining a fruit and vegetable fermentation tank, wherein the fruit and vegetable fermentation tank comprises a plurality of liquid inlets and a mixer, wherein each liquid inlet corresponds to a hypoxia regulation submodule, and the mixer is provided with a hypoxia regulation main module.

[0018] Specifically, before fermentation, first determine a fruit and vegetable fermentation tank for fermentation. This fermentation tank has multiple liquid inlets and a mixer. Each liquid inlet is equipped with a low oxygen regulation submodule for independently controlling the oxygen environment of the bacterial community entering the inlet. A low oxygen regulation main module is provided inside the mixer to comprehensively adjust the oxygen environment in the mixer. During the fermentation process, different bacterial communities can enter the fermentation tank through multiple liquid inlets. The low oxygen regulation submodule equipped at each inlet can independently regulate the oxygen requirements of different bacterial communities to ensure that each bacterial community can grow and ferment in a suitable oxygen environment. As the main place for fermentation, the mixer is equipped with a low oxygen regulation main module responsible for comprehensive regulation of the oxygen concentration of the entire fermentation environment to ensure the stability and efficiency of the fermentation process.

[0019] The fruit and vegetable fermentation tank provides a fermentation system that can carry out multi-bacteria fermentation and has preliminary oxygen regulation function, providing the hardware foundation for the subsequent precise regulation of different bacteria communities.

[0020] Step S2: collecting the multiple liquid inlets for fermenting the corresponding multiple bacterial flora types.

[0021] Specifically, the flora type refers to different types of microbial populations used for fermentation, such as lactic acid bacteria, yeasts, acetic acid bacteria, etc. Before fermentation, determine the corresponding flora type for each liquid inlet according to the fermentation process design document. For example, according to the fermentation process design, lactic acid bacteria need to be introduced into the first liquid inlet, yeasts into the second liquid inlet, and acetic acid bacteria into the third liquid inlet. These flora can be obtained from the strain bank or the bacterial liquid cultured in the laboratory through aseptic operation and transported through pipelines to the corresponding liquid inlets of the fermentation tank.

[0022] By collecting multiple flora types corresponding to the fermentation of multiple liquid inlets, the different flora participating in the fermentation are clarified, providing a basis for subsequent oxygen regulation according to the characteristics of different flora.

[0023] Step S3: Conduct oxygen adaptability analysis based on the multiple flora types and output multiple sets of low-oxygen regulation parameters, where each set of low-oxygen regulation parameters includes oxygen concentration and dosing rate.

[0024] Specifically, oxygen adaptability analysis refers to analyzing the growth and metabolic characteristics of the flora under different oxygen conditions to determine its oxygen demand and adaptability. The low-oxygen regulation parameters include parameters such as oxygen concentration and dosing rate, which are used to guide the regulation of the oxygen environment where the flora is located.

[0025] Utilize existing microbiological research data and experimental analysis to analyze the adaptability of different flora to the oxygen environment (such as oxygen concentration, oxygen supply rate, etc.) during fermentation. For example, through laboratory culture and measurement, it can be obtained that the aerobic threshold of lactic acid bacteria is 0.5% oxygen concentration, the low-oxygen growth rate is 0.2 generations per hour, and the oxygen metabolism sensitivity is medium. According to the collected flora type data, determine the oxygen concentration and dosing rate suitable for each flora through machine learning algorithms or by establishing mathematical models (such as linear regression models, etc.). For example, for lactic acid bacteria, it can be determined that the optimal oxygen concentration range is 0.3% to 0.7%, and the dosing rate is 0.1 liters of oxygen per minute.

[0026] Through oxygen adaptability analysis, the oxygen demand of different flora can be accurately understood, providing a scientific basis for subsequent oxygen regulation, so as to be able to create a suitable growth environment for each flora and improve fermentation efficiency and product quality.

[0027] Step S4: Control multiple low-oxygen regulation sub-modules to activate the flora at the multiple liquid inlets according to the multiple sets of low-oxygen regulation parameters, and the activated flora flows into the mixer through the multiple liquid inlets.

[0028] Specifically, activation refers to restoring the bacterial community from a dormant state or a low-activity state to an active growth and metabolic state by adjusting conditions such as the oxygen environment. Control multiple hypoxic regulation sub-modules, and according to multiple groups of hypoxic regulation parameters output in step S3, adjust the oxygen concentration and feeding rate of each liquid inlet respectively to make them conform to the hypoxic regulation parameters of the corresponding bacterial community, so as to activate the bacterial community at the liquid inlet. For example, for the liquid inlet of lactic acid bacteria, the oxygen concentration can be adjusted to 0.5%, and the feeding rate can be adjusted to 0.1 liters of oxygen per minute, so that it can quickly regain its vitality in a suitable oxygen environment and enter an active growth and fermentation state. The activated bacterial community flows into the mixer through multiple liquid inlets, ready for mixed fermentation.

[0029] By activating the bacterial community at multiple liquid inlets, the bacterial community can be in a suitable initial fermentation state before entering the mixer, improving the activity of the bacterial community, preparing for the subsequent fermentation in the mixer, and ensuring the smooth progress of the fermentation process.

[0030] Step S5: Construct a mixed fermentation rate regulation model, input the types of bacterial communities in the mixer into the mixed fermentation rate regulation model for active retention analysis, and output mixed-hypoxic regulation parameters.

[0031] Specifically, the mixed fermentation rate regulation model is a mathematical model established according to the types and characteristics of the bacterial communities in the mixer, and is used to predict and optimize the oxygen regulation parameters in the mixed fermentation process. Use machine learning algorithms to construct a mixed fermentation rate regulation model, input the data of the types of bacterial communities in the mixer into this model, and the model analyzes the active retention of the bacterial communities according to the preset algorithms, and outputs mixed-hypoxic regulation parameters. These mixed-hypoxic regulation parameters are used to guide the regulation of the oxygen environment in the mixer.

[0032] Using the mixed fermentation rate regulation model can accurately analyze the regulation parameters required for active retention according to the types of bacterial communities in the mixer, which helps to retain the natural active ingredients of fruits and vegetables during the fermentation process.

[0033] Step S6: Control the main hypoxic regulation module to adjust the oxygen in the mixer according to the mixed-hypoxic regulation parameters.

[0034] Specifically, according to the mixed-hypoxic regulation parameters output in step S5, control the main hypoxic regulation module to adjust the oxygen concentration and feeding rate in the mixer to make them conform to the mixed-hypoxic regulation parameters. For example, if the mixed-hypoxic regulation parameters require an oxygen concentration of 3%, then the main hypoxic regulation module will adjust the oxygen supply device (such as an oxygen nozzle, etc.) in the mixer to adjust the oxygen concentration in the mixer to 3% to optimize the fermentation conditions, improve the activity of the bacterial community and the fermentation efficiency, and at the same time better retain the natural active ingredients of fruits and vegetables.

[0035] Furthermore, as Figure 2 shown, step S3 includes:

[0036] Step S31: Collect a microbial flora sample library, and identify the oxygen adaptability data of each microbial flora sample in the microbial flora sample library, where the oxygen adaptability data includes the aerobic threshold, the low-oxygen growth rate, and the oxygen metabolism sensitivity.

[0037] Step S32: Group the oxygen adaptability data through a clustering analysis algorithm, and output the classification result of the microbial flora types.

[0038] Step S33: Generate a first fermentation regulation model according to the classification result of the microbial flora types.

[0039] Step S34: Perform oxygen adaptability analysis on the multiple microbial flora types respectively according to the first fermentation regulation model, calculate the optimal oxygen concentration range and the feeding rate of each microbial flora, and output multiple groups of low-oxygen regulation parameters.

[0040] Specifically, the microbial flora sample library is a collection that stores various data related to microbial flora. The oxygen adaptability data is the data that describes the growth and metabolic characteristics of microbial flora under different oxygen conditions, including indicators such as the aerobic threshold, the low-oxygen growth rate, and the oxygen metabolism sensitivity. Among them, the aerobic threshold is the lowest or highest boundary value of the oxygen demand of the microbial flora during the growth and metabolism process. The low-oxygen growth rate is the growth rate of the microbial flora under low-oxygen conditions. The oxygen metabolism sensitivity is the sensitivity degree of the microbial flora to the change of oxygen concentration, that is, the influence degree of the change of oxygen concentration on the metabolic activities of the microbial flora.

[0041] Obtain multiple microbial flora samples from laboratory cultures or strain banks, such as lactic acid bacteria, yeast, acetic acid bacteria, etc., and store them in the microbial flora sample library. Then, determine the oxygen adaptability data of each microbial flora sample through experiments. For example, the microbial flora samples can be placed in culture environments with different oxygen concentrations respectively, observe and record their growth conditions and metabolic activities, use an oxygen sensor to accurately measure the growth conditions of the microbial flora under different oxygen concentrations to determine the aerobic threshold; calculate the low-oxygen growth rate by regularly measuring the growth of the microbial flora quantity; use a metabolite analysis instrument (such as a gas chromatography-mass spectrometry instrument) to analyze the changes in metabolites of the microbial flora under different oxygen concentrations, so as to judge the oxygen metabolism sensitivity. The detailed oxygen adaptability data of these microbial flora samples provides the basic data support for the subsequent classification of microbial flora and the calculation of oxygen regulation parameters.

[0042] Input the oxygen adaptability data of the collected microbial community samples into a clustering analysis algorithm (such as the K-means clustering algorithm, hierarchical clustering algorithm, etc.). The algorithm will automatically classify the microbial community into different classes according to the similarity of data features. Exemplarily, the K-means clustering algorithm can be used. First, determine the number of groups to be divided (such as 3 groups), and then the algorithm will classify the microbial community samples into different groups according to characteristics such as the aerobic threshold, low-oxygen growth rate, and oxygen metabolism sensitivity. For example, lactic acid bacteria are anaerobic, yeast are low-oxygen tolerant, and acetic acid bacteria are aerobic. Through the clustering analysis of oxygen adaptability data, classifying microbial community samples with similar oxygen adaptability characteristics into one category can simplify the subsequent analysis and regulation processes of the microbial community, and improve the accuracy and effectiveness of oxygen regulation.

[0043] The first fermentation regulation model is a model constructed based on the classification results of microbial community types and is used for preliminary regulation planning of the fermentation process. According to the classification results of microbial community types and combined with the oxygen adaptability data of each microbial community, the first fermentation regulation model is established. This model can adopt mathematical modeling or machine learning methods, taking the microbial community type and oxygen adaptability data as inputs and outputting corresponding oxygen regulation strategies. For example, a linear programming model can be adopted. According to the oxygen demand characteristics of different microbial community categories (such as the growth advantage of a certain type of microbial community at a specific oxygen concentration, etc.), set the objective function and constraints, perform regression fitting with the oxygen concentration interval and dosing rate as independent variables and the low-oxygen growth rate as the dependent variable to establish the first fermentation regulation model. The first fermentation regulation model can also be constructed using a neural network: use a deep learning framework such as TensorFlow, PyTorch, etc. to construct a neural network model, collect the oxygen adaptability data of multiple microbial community types as the training data set, use the training data to iteratively train the neural network model, and adopt optimization algorithms such as the gradient descent method to continuously adjust the weights and biases of the model to make the model output as close as possible to the actual values of the training data to improve the performance of the model. The trained first fermentation regulation model can accurately predict the growth and metabolic performance of different microbial communities under different oxygen conditions, provide personalized oxygen regulation strategies for different microbial communities, and guide the oxygen regulation in the fermentation process to improve the fermentation efficiency and product quality.

[0044] According to the first fermentation regulation model, perform oxygen adaptability analysis on each microbial community type separately. Input the oxygen adaptability data and classification results of each microbial community type into the first fermentation regulation model. The model will calculate the optimal oxygen concentration interval and dosing rate for each microbial community according to this input information, making the fermentation process more in line with the growth and metabolic needs of the microbial community, further improving the fermentation efficiency and product quality, and at the same time helping to better retain the natural active ingredients of fruits and vegetables.

[0045] Furthermore, as Figure 3As shown, in step S5, constructing a mixed fermentation rate regulation model includes:

[0046] Step S51: Obtain the mixed flora sample data, where the mixed flora sample data includes data under the mixed fermentation environment, including oxygen concentration change, flora metabolic rate, and active ingredient retention rate.

[0047] Step S52: Use the first fermentation regulation model as a pre-trained model to output transfer learning parameters.

[0048] Step S53: Determine the objective function, and based on the objective function, perform transfer training with the transfer learning parameters and the mixed flora sample data to construct a mixed fermentation rate regulation model.

[0049] Specifically, the mixed flora sample data refers to the flora sample data collected under the mixed fermentation environment, including indicators such as oxygen concentration change, flora metabolic rate, and active ingredient retention rate. Among them, the oxygen concentration change is the change of oxygen concentration over time during the fermentation process, reflecting the dynamic change of oxygen supply. The flora metabolic rate is the rate at which the flora consumes nutrients and produces metabolites during the fermentation process, and is an important indicator to measure the activity of the flora. The active ingredient retention rate is the retention ratio of natural active ingredients (such as vitamins, antioxidants, etc.) in fruits and vegetables during the fermentation process, reflecting the impact of fermentation on the quality of fruits and vegetables. During the mixed fermentation process, data such as oxygen concentration change, flora metabolic rate, and active ingredient retention rate are monitored and recorded in real time through sensors and detection devices. For example, an oxygen sensor can be used to monitor the oxygen concentration, the flora metabolic rate can be determined by a biosensor or chemical analysis method, and high-performance liquid chromatography (HPLC) and other technologies can be used to detect the content change of active ingredients, so as to obtain the mixed flora sample data.

[0050] Take the first fermentation regulation model as a pre-trained model, extract the parameters and feature representations learned during its training process as transfer learning parameters. For example, if the first fermentation regulation model is a neural network, parameters such as the weights and biases of its hidden layers can be extracted, as well as the learned feature representations regarding the oxygen adaptability of the microbial community. During the actual operation process, the Python language can be used in combination with deep learning frameworks (such as TensorFlow or PyTorch) to extract the parameters in the pre-trained model by calling the corresponding functions. For example, in TensorFlow, relevant model saving and loading functions can be used to load the first fermentation regulation model, and then specific interface functions can be used to obtain the parameters useful for constructing the mixed fermentation rate regulation model. These parameters, after being sorted out, become the transfer learning parameters. By using the first fermentation regulation model as a pre-trained model, the training process of the mixed fermentation rate regulation model can be accelerated, the initial performance of the model can be improved, and it can better capture the complex laws in the mixed fermentation process even with limited sample data.

[0051] The objective function is a function used to evaluate the model performance and optimization direction during the model training process, reflecting the objective that the model needs to optimize. Determining the objective function of the mixed fermentation rate regulation model usually includes indicators such as the active ingredient retention rate and the degradation rate to ensure that the model can optimize the oxygen regulation parameters in the fermentation process, improve the active ingredient retention rate, and reduce the degradation rate. For example, the objective function can be expressed as the maximization of the active ingredient retention rate and the minimization of the degradation rate. Based on the objective function, transfer training is carried out using the transfer learning parameters and the mixed microbial community sample data. Take the transfer learning parameters as the initial parameters of the mixed fermentation rate regulation model, input the mixed microbial community sample data for training, and adjust the model parameters of the mixed fermentation rate regulation model through an optimization algorithm to make the objective function value output by the model optimal. For example, optimization algorithms such as the gradient descent method can be adopted. According to the gradient information of the objective function, the model parameters are gradually adjusted to better fit the sample data and achieve the purpose of optimizing the fermentation process.

[0052] By determining the objective function and carrying out transfer training, the constructed mixed fermentation rate regulation model can accurately predict and optimize the oxygen regulation parameters in the mixed fermentation process, improve the fermentation efficiency and product quality, better retain the natural active ingredients of fruits and vegetables, and achieve the intelligent and precise control of the fermentation process.

[0053] Furthermore, the objective function described in step S53 includes the active ingredient retention rate and the active ingredient degradation rate, where the expression of the objective function includes:

[0054] ; where is the retention rate of the k-th type of active ingredient in fruits and vegetables at time t, represents the degradation rate of the active ingredient at time t, represents the weight of the k-th type of active ingredient, and m is the number of active ingredients in the fruits and vegetables, is the weight coefficient for the degradation of the active ingredient, used to balance the relationship between the retention rate and degradation of the active ingredient.

[0055] Specifically, the above objective function comprehensively considers the retention rate of the active ingredient and the degradation rate of the active ingredient. By introducing the weight coefficient, it highlights the retention contribution of important ingredients and at the same time balances the relationship between the retention rate and degradation rate of the active ingredient. In this function expression, the retention rate of the active ingredient refers to the retention ratio of the k-th type of active ingredient in the fruits and vegetables at time t, reflecting the degree to which this ingredient is not degraded or lost during the fermentation process. The degradation rate of the active ingredient refers to the rate at which the active ingredient is decomposed or lost at time t, usually expressed as a percentage or in mass units per unit time. The first term in the function represents the weighted sum of the retention rates of all active ingredients, multiplying the retention rate of each ingredient by its weight for weighting. Ingredients with higher weights have a greater impact on the optimization result. The second term represents the negative impact of the degradation rate of the active ingredient on the objective, and is controlled by the weight coefficient . By adjusting the value, the relative importance between the retention rate and degradation rate of the active ingredient in the objective function can be changed. A larger indicates a greater emphasis on reducing the degradation rate; a smaller indicates a preference for improving the retention rate.

[0056] During the optimization process, by solving the objective function , the optimal oxygen regulation parameters are found to achieve the best retention effect of the active ingredient during the fermentation process while effectively controlling the degradation rate, thereby improving the quality and nutritional value of the fermented product.

[0057] Furthermore, the expression for calculating the degradation rate of the active ingredient includes:

[0058] ; where n is the total number of microbial populations in the mixer, is the metabolic sensitivity coefficient of the i-th microbial population in the mixer, is the metabolic rate of the i-th microbial population at time t, is the oxygen concentration in the mixer, is the optimal oxygen concentration of the i-th microbial population.

[0059] Specifically, the expression for the degradation rate of the active ingredient can comprehensively consider the metabolic rates, metabolic sensitivity coefficients, and oxygen concentration deviations of different microbial communities in the mixer, and calculate the total degradation rate of the active ingredient through calculation, providing an important reference basis for optimizing oxygen regulation. When calculating the degradation rate of the active ingredient, for each microbial community i in the mixer, calculate the oxygen concentration deviation at time t, that is . Among them, is the oxygen concentration in the mixer, is the optimal oxygen concentration of microbial community i, that is, the oxygen concentration corresponding to the highest metabolic efficiency of the microbial community. Then, square the oxygen concentration deviation to obtain , to eliminate the positive and negative effects of the deviation and amplify the contribution of the deviation. Multiply the squared deviation by the metabolic rate and the metabolic sensitivity coefficient of microbial community i to obtain the degradation contribution of microbial community i at time t. Sum the degradation contributions of all microbial communities to obtain the total active ingredient degradation rate of the mixer at time t. By minimizing this degradation rate, the loss of the active ingredient can be effectively controlled, and the quality and nutritional value of the fermentation product can be improved.

[0060] Furthermore, inputting the types of microbial communities in the mixer into the mixed fermentation rate regulation model in step S5 for active ingredient retention analysis includes:

[0061] Step S54: Identify the types of microbial communities in the mixer and extract the key characteristic data of each microbial community.

[0062] Step S55: Input the key characteristic data for feature vectorization processing and output a set of feature vectors.

[0063] Step S56: Call the mixed fermentation rate regulation model and predict the active ingredient retention rate and active ingredient degradation rate of the mixer according to the set of feature vectors.

[0064] Step S57: Output the mixing-low oxygen regulation parameters according to the relationship between the active ingredient retention rate and the active ingredient degradation rate.

[0065] Specifically, key feature data are data that can reflect the important characteristics of the microbial community during fermentation, such as metabolic rate, oxygen sensitivity, etc. Sensors and detection devices, such as microbial detectors, metabolic analyzers, etc., are used to monitor and identify the microbial community in the mixer in real time to determine the types of microbial communities present. Then, the key feature data of each microbial community, such as metabolic rate, oxygen sensitivity, etc., are extracted. For example, by analyzing the change in the concentration of metabolites of the microbial community, its metabolic rate can be calculated; by observing the growth of the microbial community under different oxygen concentrations, its oxygen sensitivity can be evaluated. Identifying the types of microbial communities in the mixer and extracting key feature data provide accurate input information for subsequent active ingredient retention analysis, enabling the model to make accurate predictions and regulations based on the actual state of the microbial community.

[0066] The extracted key feature data are subjected to feature vectorization processing. For example, data such as metabolic rate and oxygen sensitivity are converted into numerical forms and normalized so that they are within the same numerical range, facilitating the calculation and comparison of the model. The processed feature data are organized into a set of feature vectors, and each feature vector contains the key feature information of a microbial community type. For example, during the fermentation of tomato juice, there may be lactic acid bacteria, yeast, and acetic acid bacteria in the mixer. After the key feature data such as their metabolic rate and oxygen sensitivity are subjected to feature vectorization processing, a corresponding set of feature vectors is obtained, such as {[0.3, 0.5], [0.2, 0.8], [0.4, 0.3]}.

[0067] The obtained set of feature vectors is used as input data to load the already constructed mixed fermentation rate regulation model. The model calculates the input set of feature vectors according to its internal algorithm structure (such as the weights and activation functions of each layer in the neural network structure), and based on the patterns and rules learned during the construction and training process before, predicts the retention rate of active ingredients and the degradation rate of active ingredients in the mixer. By invoking the mixed fermentation rate regulation model, it is possible to quickly and accurately predict the retention rate of active ingredients and the degradation rate of active ingredients during the mixed fermentation process using the existing data and model structure, providing an important basis for the monitoring and regulation of the fermentation process.

[0068] According to the values of the predicted retention rate of active ingredients and the degradation rate of active ingredients, mathematical calculations and logical judgments are used to determine the mixing-low oxygen regulation parameters. For example, if the retention rate of active ingredients is low and the degradation rate is high, it may be necessary to reduce the oxygen supply to reduce the metabolic rate of the microbial community, thereby increasing the retention rate of active ingredients. A simple algorithm can be set, such as when the retention rate of active ingredients is lower than a certain threshold (such as 30%) and the degradation rate is higher than a certain value (such as 0.1 unit / hour), the oxygen concentration is reduced by a certain proportion (such as 10%), so as to determine the corresponding mixing-low oxygen regulation parameters, including oxygen concentration and feeding rate, etc.

[0069] The mixed-hypoxic regulation parameters determined through the above steps can provide specific regulation directions and numerical bases for the mixed fermentation process, which helps to maximize the retention of active ingredients during the fermentation process and improve the quality and nutritional value of the fermentation products.

[0070] Furthermore, after step S6, it further includes:

[0071] Step S71: Obtain the feedback data set of the mixer after regulation, where the feedback data set includes oxygen concentration feedback data and activity detection feedback data.

[0072] Step S72: Perform feedback optimization on the mixed fermentation rate regulation model according to the feedback data set, and output the optimized mixed fermentation rate regulation model.

[0073] Specifically, the feedback data set is the data obtained from the mixer after adjusting the oxygen environment, which is used to evaluate the adjustment effect and model performance. It includes oxygen concentration feedback data and activity detection feedback data. Among them, the oxygen concentration feedback data is the actual oxygen concentration data in the mixer after adjustment, which reflects the actual effect of oxygen adjustment. The activity detection feedback data is the activity data of the bacteria population and the retention of active ingredients in the mixer after adjustment, which is used to evaluate the quality and efficiency of the fermentation process.

[0074] After adjusting the oxygen environment in the mixer, use sensors and detection devices to monitor and record the oxygen concentration feedback data and activity detection feedback data in real time. For example, the percentage concentration of oxygen in the mixer is measured in real time through an oxygen sensor. The activity of the bacteria population and the retention of active ingredients are detected by biosensors or chemical analysis methods.

[0075] Input the obtained feedback data set into the mixed fermentation rate regulation model, and evaluate the prediction error and performance metrics of the model, such as the mean squared error (MSE) or the coefficient of determination (R²), etc. According to the evaluation results, use the model optimization algorithm in machine learning to optimize and adjust the model, including adjusting model parameters, improving model structure or updating model algorithms, etc. Exemplarily, for the mixed fermentation rate regulation model constructed based on a neural network, the backpropagation algorithm can be used for optimization. First, divide the feedback data set into a training set and a test set. Using the training set data as input, calculate the error between the predicted output of the model and the actual feedback data (oxygen concentration feedback data and activity detection feedback data). Then, adjust the weights and parameters of the model according to the error backpropagation. For example, if there is a large deviation between the predicted oxygen concentration of the model and the actual oxygen concentration feedback data, through the backpropagation algorithm, the weights of the neural network layer related to oxygen concentration prediction in the model will be adjusted to reduce this deviation. During the parameter adjustment process, an optimizer such as the Adam optimizer can be used to adaptively adjust the learning rate and accelerate the convergence speed of the model. After completing the parameter adjustment, use the test set data to verify the optimized model to ensure that the performance of the model is improved rather than overfitted. Finally, output the optimized mixed fermentation rate regulation model.

[0076] Through feedback optimization, the mixed fermentation rate regulation model can better adapt to the actual mixed fermentation process. Improve the accuracy in predicting oxygen concentration, active ingredient retention rate, active ingredient degradation rate, etc., so as to more effectively guide the regulation operations in the mixed fermentation process and improve fermentation efficiency and product quality.

[0077] In summary, the fermentation rate regulation method for maintaining the natural activity of fruits and vegetables provided by the embodiments of the present application has the following technical effects:

[0078] By obtaining a fruit and vegetable fermenter with multiple liquid inlets and a mixer, an independent oxygen regulation environment is provided for different bacterial communities, ensuring that each bacterial community can grow and ferment under suitable oxygen conditions. Then, collect various types of bacterial communities and conduct oxygen adaptability analysis, and output multiple sets of low-oxygen regulation parameters, enabling the fermentation process to be individually regulated according to the specific needs of the bacterial communities and optimizing the fermentation conditions. Then, construct a mixed fermentation rate regulation model, and generate a first fermentation regulation model through oxygen adaptability data and clustering analysis algorithm, further improving the accuracy and scientificity of oxygen regulation. In addition, by real-time monitoring the bacterial community types and key feature data in the mixer and performing feature vectorization processing, the model can accurately predict the retention rate and degradation rate of active ingredients, thereby outputting the optimal mixed-low oxygen regulation parameters and realizing the dynamic optimization of the fermentation process. Finally, by obtaining the adjusted feedback data set and performing feedback optimization on the model, the continuous improvement and stable operation of the fermentation process are ensured.

[0079] Overall, in the embodiments of the present application, through the phased fine regulation of low oxygen according to the oxygen demand characteristics of multiple bacterial communities, the dynamic optimization of the fermentation conditions throughout the whole process from bacterial community activation to mixed fermentation is realized. This not only ensures the independent activity of each bacterial community but also realizes the efficient and unified regulation of the oxygen environment in the mixed fermentation stage, thus significantly improving the fermentation speed, enhancing the quality of the fermentation products, and successfully retaining the natural activity of fruits and vegetables.

[0080] Embodiment 2, as Figure 4 shown, based on the same inventive concept as Embodiment 1, the embodiments of the present application provide a fermentation device for maintaining the natural activity of fruits and vegetables, and the device includes:

[0081] A fermentation tank connection unit 10, configured to obtain a fruit and vegetable fermentation tank, where the fruit and vegetable fermentation tank includes a plurality of liquid inlet ports and a mixer, and each liquid inlet port corresponds to a low oxygen regulation sub-module, and the mixer is provided with a low oxygen regulation main module.

[0082] A bacterial community type collection unit 20, configured to collect a plurality of bacterial community types corresponding to fermentation for the plurality of liquid inlet ports.

[0083] An oxygen adaptability analysis unit 30, configured to perform oxygen adaptability analysis according to the plurality of bacterial community types and output multiple groups of low oxygen regulation parameters, where each group of low oxygen regulation parameters includes an oxygen concentration and a feeding rate.

[0084] A low oxygen regulation unit 40, configured to control a plurality of low oxygen regulation sub-modules to activate the bacterial communities at the plurality of liquid inlet ports according to the multiple groups of low oxygen regulation parameters, and the activated bacterial communities flow into the mixer through the plurality of liquid inlet ports.

[0085] An activity retention analysis unit 50, configured to construct a mixed fermentation speed regulation model, input the bacterial community types in the mixer into the mixed fermentation speed regulation model for activity retention analysis, and output mixed-low oxygen regulation parameters.

[0086] An oxygen regulation unit 60, configured to control the low oxygen regulation main module to perform oxygen regulation on the mixer according to the mixed-low oxygen regulation parameters.

[0087] Furthermore, the oxygen adaptability analysis unit 30 in the embodiments of the present application is configured to perform the following steps:

[0088] Collect a microbial community sample library, identify the oxygen adaptability data of each microbial community sample in the microbial community sample library, where the oxygen adaptability data includes aerobic threshold, low-oxygen growth rate, and oxygen metabolism sensitivity; group the oxygen adaptability data through a clustering analysis algorithm, and output the classification result of the microbial community type; generate a first fermentation regulation model according to the classification result of the microbial community type; perform oxygen adaptability analysis on the multiple microbial community types according to the first fermentation regulation model, calculate the optimal oxygen concentration range and dosing rate for each microbial community, and output multiple groups of low-oxygen regulation parameters.

[0089] Further, the activity retention analysis unit 50 in the embodiment of the present application is used to perform the following steps:

[0090] Obtain mixed microbial community sample data, where the mixed microbial community sample data includes data in a mixed fermentation environment, including oxygen concentration change, microbial community metabolic rate, and active ingredient retention rate; use the first fermentation regulation model as a pre-trained model to output transfer learning parameters; determine an objective function, and perform transfer training based on the objective function with the transfer learning parameters and the mixed microbial community sample data to construct a mixed fermentation rate regulation model.

[0091] Further, the objective function includes the active ingredient retention rate and the active ingredient degradation rate, where the expression of the objective function includes:

[0092] ; where represents the retention rate of the k-th type of active ingredient in fruits and vegetables at time t, represents the degradation rate of the active ingredient at time t, represents the weight of the k-th type of active ingredient, and m is the number of active ingredients in fruits and vegetables, is the weight coefficient of active ingredient degradation, which is used to balance the relationship between the active ingredient retention rate and degradation.

[0093] Further, the expression for calculating the active ingredient degradation rate includes:

[0094] ; where n is the total number of microbial communities in the mixer, is the metabolic sensitivity coefficient of the i-th microbial community in the mixer, is the metabolic rate of the i-th microbial community at time t, is the oxygen concentration of the mixer, is the optimal oxygen concentration of the i-th microbial community.

[0095] Further, the activity retention analysis unit 50 in the embodiment of the present application is also used to perform the following steps:

[0096] Identify the types of bacterial flora in the mixer, and extract the key feature data of each type of bacterial flora; input the key feature data for feature vectorization processing, and output a set of feature vectors; call the mixed fermentation rate regulation model, and predict the retention rate of active ingredients and the degradation rate of active ingredients in the mixer according to the set of feature vectors; according to the relationship between the retention rate of active ingredients and the degradation rate of active ingredients, output the mixing-hypoxia regulation parameters.

[0097] Further, the device according to the embodiment of the present application further includes a feedback optimization unit, and the feedback optimization unit is used to perform the following steps:

[0098] Obtain the feedback data set of the mixer after adjustment, and the feedback data set includes oxygen concentration feedback data and activity detection feedback data; perform feedback optimization on the mixed fermentation rate regulation model according to the feedback data set, and output the optimized mixed fermentation rate regulation model.

[0099] Through the foregoing detailed description of a method for regulating the fermentation rate of maintaining the natural activity of fruits and vegetables in this specification, those skilled in the art can clearly know a fermentation device for maintaining the natural activity of fruits and vegetables in this embodiment. For the device disclosed in Embodiment 2, since it corresponds to the method disclosed in Embodiment 1, it has corresponding functional units and beneficial effects. For the relevant parts, refer to the description in the method part.

[0100] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for controlling the fermentation rate to maintain the natural activity of fruits and vegetables, characterized in that: The method comprises: Obtain a fruit and vegetable fermentation tank, the fruit and vegetable fermentation tank comprising a plurality of liquid inlets and a mixer, wherein each liquid inlet corresponds to a hypoxia regulation submodule, and the mixer is provided with a hypoxia regulation main module; Collect the multiple liquid inlets for fermenting the corresponding multiple bacterial colony types; Performing oxygen adaptability analysis according to the multiple bacterial colony types, and outputting multiple groups of hypoxia control parameters, wherein each group of hypoxia control parameters includes oxygen concentration and delivery rate; Controlling a plurality of hypoxia control submodules to activate the bacterial colonies at the plurality of liquid inlets according to the plurality of groups of hypoxia control parameters, so that the activated bacterial colonies flow into the mixer through the plurality of liquid inlets; Constructing a mixed fermentation speed control model, inputting the bacterial flora type in the mixer into the mixed fermentation speed control model for activity retention analysis, and outputting a mixed-hypoxia control parameter; Controlling the hypoxia control main module to regulate oxygen in the mixer according to the mixing-hypoxia control parameters; Performing oxygen adaptability analysis according to the multiple bacterial colony types and outputting multiple groups of hypoxia control parameters, the method includes: Collecting a bacterial sample library, identifying oxygen adaptability data of each bacterial sample in the bacterial sample library, wherein the oxygen adaptability data includes aerobic threshold, hypoxic growth rate, and oxygen metabolic sensitivity; Grouping the oxygen adaptability data by a cluster analysis algorithm, and outputting a bacterial flora type classification result; generating a first fermentation regulation model according to the bacterial flora type classification result; performing oxygen adaptability analysis on the multiple bacterial colony types according to the first fermentation regulation model, calculating the optimal oxygen concentration range and delivery rate for each bacterial colony, and outputting multiple groups of hypoxia regulation parameters; Constructing a mixed fermentation speed control model, the method includes: Acquiring mixed bacterial flora sample data, wherein the mixed bacterial flora sample data includes data under a mixed fermentation environment, including oxygen concentration changes, bacterial flora metabolic rate, and active ingredient retention rate; Using the first fermentation regulation model as a pre-training model, outputting transfer learning parameters; Determine an objective function, perform migration training based on the objective function using the migration learning parameters and the mixed flora sample data, and construct a mixed fermentation speed control model; The objective function includes the active ingredient retention rate and the active ingredient degradation rate, wherein the expression of the objective function includes: Among them, Active k (t) represents the retention rate of the kth type of active ingredient in fruits and vegetables at time t, Decay(t) represents the degradation rate of the active ingredient at time t, ω k represents the weight of the kth type of active ingredient, m is the number of active ingredients in fruits and vegetables, and λ is the weight coefficient of active ingredient degradation, which is used to balance the relationship between active ingredient retention rate and degradation.

2. A method for controlling the fermentation rate for maintaining the natural activity of fruits and vegetables as claimed in claim 1, characterized in that: The expression for calculating the degradation rate of the active ingredient includes: Wherein, n is the total number of bacteria in the mixer, α i is the metabolic sensitivity coefficient of bacterial colony i in the mixer, i (t) is the metabolic rate of bacterial colony i at time t, C(t) is the oxygen concentration in the mixer, C opt,i is the optimal oxygen concentration for bacterial colony i.

3. A method for controlling the fermentation speed for maintaining the natural activity of fruits and vegetables as claimed in claim 1, characterized in that: Inputting the bacterial flora type in the mixer into the mixed fermentation speed control model to perform activity retention analysis, the method comprising: Identify the type of bacterial colonies in the mixer and extract key characteristic data of each bacterial colony; Input the key feature data to perform feature vectorization processing, and output a feature vector set; Calling the mixed fermentation speed control model to predict the active ingredient retention rate and active ingredient degradation rate of the mixer according to the feature vector set; According to the relationship between the active ingredient retention rate and the active ingredient degradation rate, the mixed-hypoxia control parameter is output.

4. A method for controlling the fermentation rate for maintaining the natural activity of fruits and vegetables as claimed in claim 1, characterized in that: After controlling the hypoxia control main module to adjust the oxygen in the mixer according to the mixing-hypoxia control parameter, the method further includes: Acquire a feedback data set of the mixer after adjustment, wherein the feedback data set includes oxygen concentration feedback data and activity detection feedback data; Feedback optimization is performed on the mixed fermentation speed control model according to the feedback data set, and an optimized mixed fermentation speed control model is output.

5. A fermentation device for maintaining the natural activity of fruits and vegetables, characterized in that: The device is used to implement a fermentation speed control method for maintaining the natural activity of fruits and vegetables as described in any one of claims 1 to 4, comprising: A fermentation tank connection unit, used to obtain a fruit and vegetable fermentation tank, wherein the fruit and vegetable fermentation tank comprises a plurality of liquid inlets and a mixer, wherein each liquid inlet corresponds to a hypoxia regulation submodule, and the mixer is provided with a hypoxia regulation main module; A bacterial flora type collection unit, used for collecting the plurality of bacterial flora types corresponding to the plurality of liquid inlets for fermentation; An oxygen adaptability analysis unit, configured to perform oxygen adaptability analysis according to the plurality of bacterial colony types, and output a plurality of groups of hypoxia control parameters, wherein each group of hypoxia control parameters includes oxygen concentration and delivery rate; A hypoxia control unit, used for controlling the plurality of hypoxia control submodules to activate the bacterial colonies at the plurality of liquid inlets according to the plurality of groups of hypoxia control parameters, so that the activated bacterial colonies flow into the mixer through the plurality of liquid inlets; An activity retention analysis unit is used to construct a mixed fermentation speed control model, input the bacterial flora type in the mixer into the mixed fermentation speed control model to perform activity retention analysis, and output a mixed-hypoxia control parameter; An oxygen regulating unit, used for controlling the hypoxia regulating main module to regulate oxygen in the mixer according to the mixing-hypoxia regulating parameters; The Oxygen Suitability Analysis Unit is used to perform the following steps: Collect a bacterial sample library, identify oxygen adaptability data of each bacterial sample in the bacterial sample library, wherein the oxygen adaptability data includes an aerobic threshold, a hypoxic growth rate, and oxygen metabolic sensitivity; group the oxygen adaptability data by a cluster analysis algorithm, and output a bacterial type classification result; generate a first fermentation regulation model according to the bacterial type classification result; perform oxygen adaptability analysis on the multiple bacterial types according to the first fermentation regulation model, calculate the optimal oxygen concentration range and delivery rate of each bacterial population, and output multiple groups of hypoxia regulation parameters; The Activity Retention Assay Unit is used to perform the following steps: Acquire mixed flora sample data, the mixed flora sample data including data in a mixed fermentation environment, including oxygen concentration changes, flora metabolic rate, and active ingredient retention rate; use the first fermentation regulation model as a pre-training model and output transfer learning parameters; determine an objective function, perform migration training based on the objective function using the transfer learning parameters and the mixed flora sample data, and construct a mixed fermentation speed regulation model; Furthermore, the objective function includes active ingredient retention rate and active ingredient degradation rate, wherein, The expression of the objective function includes: Among them, Active k (t) represents the retention rate of the kth type of active ingredient in fruits and vegetables at time t, Decay(t) represents the degradation rate of the active ingredient at time t, ω k represents the weight of the kth type of active ingredient, m is the number of active ingredients in fruits and vegetables, and λ is the weight coefficient of active ingredient degradation, which is used to balance the relationship between active ingredient retention rate and degradation.

Images