Fast food fresh-keeping control method based on flash freezing technology

Abstract

The present invention discloses a rapid food preservation control method based on flash freezing technology, which relates to the field of food preservation technology and includes the following steps: obtaining the first data of the food and spraying a nano-coating containing antifreeze proteins with different concentrations on the surface of the food; inputting the first data into a pre-trained neural network model to obtain the second data; synchronously starting a refrigerant injection and pulsed electric field excitation device according to the second data, obtaining food dynamic data, and performing feature extraction and trend prediction on the food dynamic data, wherein the food dynamic data includes real-time core temperature and ice crystal formation dynamics; dynamically adjusting the refrigerant flow rate and electric field frequency according to the prediction results, and synchronously adjusting the release rate of antifreeze proteins in the nano-coating; through the precise synchronous adjustment of the refrigerant flow rate, electric field frequency and antifreeze protein release rate, the present application effectively solves the problems of uneven freezing, too large ice crystals and food quality deterioration in traditional freezing methods, and improves the fineness and preservation effect of food freezing.

Description

Technical Field

[0001] The present invention relates to the technical field of food preservation, and more specifically, to a rapid food preservation control method based on flash freezing technology. Background Art

[0002] With the acceleration of people's living rhythm, the demand for convenient, fresh and nutritious food is increasing day by day, and the fast food market continues to expand. However, traditional freezing preservation technologies have many limitations and are difficult to meet consumers' high requirements for food quality. In this context, flash freezing technology has emerged, bringing new breakthroughs to fast food preservation.

[0003] Traditional freezing preservation mainly relies on air as the heat transfer medium, such as in blast freezers or tunnel freezers. Although it can ensure that the central temperature of the food drops below -18°C within half an hour, crossing the "maximum ice crystal formation zone", the freezing speed is relatively slow. During the freezing process, the internal moisture of the food gradually precipitates, the solution concentration increases, and the freezing point decreases. When the temperature drops to -20°C, there is still about 10% of the moisture unfrozen. The slow freezing speed will lead to the formation of large ice crystals, which will pierce the cell membrane, causing cell death and cytoplasmic leakage, making the texture of the thawed food soft, the color dull, the flavor reduced, and also causing the loss of water-soluble vitamins and minerals, affecting the nutritional value of the food.

[0004] Flash freezing technology, on the other hand, uses a liquid immersion freezing method with a liquid as the heat transfer medium, which has a higher heat transfer coefficient. For example, when using the flash freezing and freshness-locking technology, the food ingredients are immersed in a special food-grade coolant. The coolant is cooled to -35°C - -55°C and circulated at a certain flow rate, which can super-rapidly remove the heat of the frozen food. The freezing ice front uniformly and rapidly advances to the center of the food ingredients, making the central temperature of the food ingredients reach -18°C. The ice front advancing speed can reach 8 - 15 cm / h, which is more than twice that of conventional air heat transfer freezing. At such a fast freezing speed, the ice crystals formed by the internal moisture of food cells are extremely small and will not cause physical damage to the cell membrane / wall. Experiments show that the food ingredients frozen in this way can be stored for more than 12 months at -18°C. After thawing, the inherent juice, nutrients and freshness substances of the food ingredients do not flow out, the color is bright, the elasticity is good, and the freshness is like at the beginning. Taking milk as an example, it takes 15 minutes to cross the "maximum ice crystal formation zone" using the ordinary air freezing method, while it only takes 1.5 minutes using the liquid immersion freezing method, and the freezing speed is increased by a full 10 times. The taste and texture of the thawed milk are not much different from those of fresh milk.

[0005] The quick-freezing technology can not only effectively extend the shelf life of food, but also maximize the preservation of the original flavor, texture and nutritional value of food, meeting consumers' demand for high-quality fast food. At present, the quick-freezing technology has been widely applied in many fields such as meat, aquatic products, fruits and vegetables, and prepared foods, providing strong technical support for the development of the fast food industry. With the continuous progress and improvement of technology, the quick-freezing technology is expected to play a more important role in the field of food preservation.

[0006] For example, a refrigeration control method, device and refrigeration equipment disclosed in the invention patent with the publication number of CN114877613A; the method includes: obtaining a first temperature parameter; the first temperature parameter is the ambient temperature in the freezing space; in the immersion freezing mode, controlling the operation of the refrigeration equipment according to a preset first control logic, and obtaining a second temperature parameter; the second temperature parameter is the temperature of the object placed in the freezing space; judging whether to enter the frozen storage mode according to the second temperature parameter; in the frozen storage mode, controlling the operation of the refrigeration equipment according to a preset second control logic. The solution of this application divides the freezing process into two stages: immersion freezing and frozen storage, and different control logics are adopted in different stages, which can improve the freezing rate, quickly pass through the maximum ice crystal formation zone, and reduce the problem of juice loss caused by a slow freezing rate; and reduce the dry loss problem of frozen food, thereby maintaining the quality of frozen food.

[0007] For example, a fresh-keeping thawing method based on a polynomial interpolation algorithm disclosed in the invention patent with the publication number of CN113892578A includes step S1: obtaining the net weight value of the food to be thawed at room temperature and the gross weight value in the current frozen state and calculating the water content, and placing the food to be thawed on a weighing pan with a percolation function in a sealed metal container; S2: connecting a high-voltage variable-frequency inducer to the sealed metal container to generate a high-voltage variable-frequency electric field acting on the food to be thawed; S3: fitting according to the polynomial interpolation function to obtain the optimal frequency corresponding to different water contents; S4: real-time monitoring the weight of the food to be thawed through the weighing pan to calculate the current remaining water content, and correspondingly adjusting the voltage and frequency of the high-voltage variable-frequency electric field. The high-voltage variable-frequency electric field can shorten the thawing time and thaw evenly. The frequency obtained through the polynomial interpolation function has high accuracy, can greatly improve the effective energy utilization rate, improve the thawing efficiency, and save the total energy and time required for thawing.

[0008] In the above-disclosed technical solutions, there are at least the following technical problems:

[0009] Traditional food preservation cannot effectively control the problems of too large ice crystals, uneven freezing and deterioration of food texture during the freezing process.

[0010] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0011] To overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a rapid food preservation control method based on flash freezing technology. By precisely synchronously adjusting the refrigerant flow rate, electric field frequency, and anti-freeze protein release rate, it effectively solves the problems of uneven freezing, excessive ice crystals, and food quality deterioration in traditional freezing methods, and improves the fineness of food freezing and the preservation effect.

[0012] To achieve the above object, the present invention provides the following technical solutions:

[0013] A rapid food preservation control method based on flash freezing technology, comprising the following steps: obtaining first data of the food and spraying a nano-coating containing anti-freeze proteins with different concentrations on the food surface, the first data including the initial temperature, thickness, and water content; inputting the first data into a pre-trained neural network model to obtain second data, the second data including the refrigerant flow rate, target freezing rate, and pulsed electric field frequency; synchronously starting a refrigerant injection and pulsed electric field excitation device according to the second data, obtaining food dynamic data, and performing feature extraction and trend prediction on the food dynamic data, the food dynamic data including the real-time core temperature and the dynamic formation of ice crystals; dynamically adjusting the refrigerant flow rate and electric field frequency according to the prediction results, and synchronously adjusting the release rate of anti-freeze proteins in the nano-coating.

[0014] In a preferred embodiment, the step of obtaining first data of the food and spraying a nano-coating containing anti-freeze proteins with different concentrations on the food surface, where the first data includes the initial temperature, thickness, and water content, is specifically as follows: placing the food to be processed on a clean platform, removing surface impurities and drying the moisture; collecting the physical characteristic parameters of the food to form a first data set, the first data set including the initial temperature, thickness, and water content of the food to be processed; selecting natural anti-freeze proteins with the function of inhibiting ice crystal growth and dividing them into several groups of anti-freeze protein solutions with different concentrations; based on the first data, obtaining the food change characteristics after freezing under different anti-freeze protein solutions, the food change characteristics including the uniformity of ice crystals, surface heat exchange rate, and texture; inputting the food change characteristics into a preset food change evaluation model to output a food change evaluation result, obtaining the anti-freeze protein solution with the best concentration according to the food change evaluation result, and constructing an experience database; calling the experience database according to the first data to match the best anti-freeze protein concentration, and forming a uniform film on the food surface through a spraying device.

[0015] In a preferred embodiment, the first data is input into a pre-trained neural network model to obtain second data, and the second data includes refrigerant flow rate, target freezing rate, and pulsed electric field frequency, specifically as follows: Obtain the first data of the food, standardize and encode the first data into a feature vector; input the feature vector into the pre-trained neural network model for mapping to obtain second data, and the second data includes refrigerant flow rate, target freezing rate, and pulsed electric field frequency; test the refrigeration equipment to be measured to obtain the constraint conditions of the second data.

[0016] In a preferred embodiment, the testing of the refrigeration equipment to be measured to obtain the constraint conditions of the second data is specifically as follows: Conduct an extreme test on the refrigerant system of the equipment to be measured, gradually increase the refrigerant flow rate until the nozzle is blocked or the pressure exceeds the limit, and record the maximum refrigerant flow rate; Screen through variance calculation to obtain the refrigerant flow rate at the minimum stable injection flow rate as the minimum value, and obtain the range of the refrigerant flow rate; Freeze the food to be measured at different rates, and conduct a texture analysis on the frozen food to obtain the critical point of the ice crystal size and cell damage rate, and determine the maximum target freezing rate through the critical point; Adjust the pulsed electric field frequency under a fixed field strength, detect the juice loss rate and microbial survival rate of the food after thawing, and obtain the safe range of the pulsed electric field frequency.

[0017] In a preferred embodiment, the refrigerant injection and pulsed electric field excitation device are synchronously started according to the second data, food dynamic data is obtained, and feature extraction and trend prediction are performed on the food dynamic data. The food dynamic data includes real-time core temperature and ice crystal formation dynamics, specifically as follows: Place the food to be processed in the refrigeration equipment, and connect the refrigerant injection system and the pulsed electric field excitation device to the control unit; Load the second data obtained from the pre-trained neural network model; The control unit synchronously starts the refrigerant injection and pulsed electric field excitation device according to the second data, performs flash freezing on the food to obtain food dynamic data, and the food dynamic data includes the real-time core temperature of different parts of the food and the formation rate, distribution state, and particle size change of ice crystals; Perform filtering and noise reduction processing on the food dynamic data, and perform feature extraction on the processed food dynamic data. The features include temperature gradient change, ice crystal growth trend, and freezing stability; Predict the extracted features based on the time series method; Evaluate the freezing level according to the prediction results.

[0018] In a preferred embodiment, the freezing level is evaluated according to the prediction result as follows: A plurality of temperature sensors are arranged inside and on the surface of the food to be frozen to form a temperature monitoring network in a three-dimensional space. Each sensor collects temperature data at a preset period to form a dynamic data sequence of temperature changing with time. Based on the temperature data at different sensing points, the local temperature gradient is obtained, and the standard deviation of the local temperature gradient is output to obtain the overall freezing uniformity characteristic. The temperature sequence is analyzed by Fourier transform to obtain the frequency spectrum distribution, and the main frequency component and the high-frequency component intensity are extracted, and the high-frequency energy ratio is statistically calculated as the temperature oscillation frequency. Based on the electrical impedance method, the time-varying curve of the ice crystal particle size is obtained, and the variance of each window is output by the sliding window method, and the ice crystal growth mode is identified according to the variance. The freezing uniformity characteristic, the particle size and the temperature oscillation frequency are input into the support vector machine, and the freezing level is output.

[0019] In a preferred embodiment, the refrigerant flow rate and the electric field frequency are dynamically adjusted according to the prediction result, and the release rate of the antifreeze protein in the nano-coating is synchronously adjusted as follows: Obtain the causal relationship diagram of the coupled data; Obtain the historical data of the coupled data, and based on the Bayesian network model, train to obtain the conditional probability distribution in the causal relationship diagram; Based on the trained Bayesian network model, input the real-time monitored refrigerant flow rate and electric field frequency into the model, and output the predicted values of the current release rate of the antifreeze protein and the freezing result; According to the predicted freezing level, the system automatically adjusts the refrigerant flow rate and the electric field frequency; According to the predicted freezing level, based on the trained Bayesian network model, inversely deduce the optimal release rate of the antifreeze protein, and synchronously adjust the release rate of the antifreeze protein according to the inverse deduction result.

[0020] The technical effects and advantages of a rapid food preservation control method based on the flash freezing technology of the present invention:

[0021] 1. By synchronously adjusting the refrigerant flow rate, the freezing rate, the pulsed electric field frequency and the release rate of the antifreeze protein, the present invention realizes the precise control of the food freezing process. Using the prediction and adjustment of the neural network model can effectively avoid the phenomenon of too large ice crystals or uneven freezing commonly seen in the traditional freezing process, ensure the freezing uniformity of the food and the fineness of the ice crystals, and thus reduce the deterioration of the food texture caused by the destruction of the cell structure by the ice crystals. By dynamically adjusting the refrigerant flow rate and the electric field frequency, it is also possible to flexibly meet the characteristic requirements of different foods, so as to obtain the best freezing effect.

[0022] 2. The present invention effectively inhibits the growth of ice crystals by spraying nano - coatings containing antifreeze proteins with different concentrations, and selects an antifreeze protein solution with an appropriate concentration according to first - hand data such as the initial temperature, thickness, and water content of the food, thereby reducing the damage to food cells caused by the expansion of ice crystal volume during the freezing process and greatly improving the frost resistance of the food. This method enables the food to maintain its original texture and taste after thawing. Especially for sensitive foods such as fruits, vegetables, and meats, the release and control of antifreeze proteins can further improve the quality of the frozen food.

[0023] 3. The present invention precisely evaluates the freezing uniformity of food, the growth trend of ice crystals, and freezing stability by real - time monitoring the core temperature and the dynamic formation of ice crystals in food, extracting features based on sensor data, and combining time - series methods for prediction. Through this technology, dynamic data can be obtained in real - time during the freezing process, and the freezing process can be adjusted according to the prediction results to ensure that the freezing uniformity and ice crystal particle size are controlled within the optimal range. Compared with traditional methods, the present invention can significantly improve the intelligent level of the freezing process, reduce the need for manual intervention, and ensure the consistency and efficiency of food freezing treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic flow chart of a rapid food preservation control method based on flash freezing technology according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] Embodiment 1, Figure 1 A rapid food preservation control method based on flash freezing technology according to the present invention is given, including the following steps:

[0027] S1, Obtain the first - hand data of the food and spray a nano - coating containing antifreeze proteins with different concentrations on the surface of the food. The first - hand data includes the initial temperature, thickness, and water content.

[0028] In this embodiment, obtaining the first - hand data of the food and spraying a nano - coating containing antifreeze proteins with different concentrations on the surface of the food, where the first - hand data includes the initial temperature, thickness, and water content, is specifically as follows:

[0029] Place the food to be processed on a clean platform, remove surface impurities and dry the moisture to ensure the accuracy and stability of subsequent data measurement and coating adhesion;

[0030] Use an integrated multi-functional detection module to sequentially collect the following physical characteristic parameters of the food to form a first data set, where the first data set includes the initial temperature, thickness, and water content of the food to be processed;

[0031] Select a natural antifreeze protein with the function of inhibiting ice crystal growth and divide it into several groups of antifreeze protein solutions with different concentrations;

[0032] Based on the first data, obtain the characteristics of the frozen food under different antifreeze protein solutions, where the food change characteristics include the uniformity of ice crystals, the surface heat exchange rate, and the texture;

[0033] The texture is analyzed by low-field nuclear magnetic resonance method to indirectly reflect the tissue integrity and juiciness by analyzing the migration state of water in the food, so as to obtain the texture result of the food;

[0034] Input the food change characteristics into a preset food change evaluation model to output the food change evaluation result, obtain the best concentration of antifreeze protein solution according to the food change evaluation result, and construct an empirical database;

[0035] According to the first data, call the empirical database to match the best antifreeze protein concentration, and form a uniform film on the food surface through a high-precision nano-spraying device.

[0036] The food change evaluation model is specifically as follows:

[0037] ;

[0038] In the formula: is the food change evaluation result, is the ice crystal uniformity index, is the food surface heat exchange rate, is the food texture index, and are the weight coefficients.

[0039] Exemplarily, from the collected food change evaluation data. Suppose we select the ice crystal uniformity index, the food surface heat exchange rate, and the food texture index as the attributes to be analyzed. At the same time, set the corresponding weighting coefficients. The obtained data is shown in the following table:

[0040]

[0041]

[0042] Table 1. Data table of food antifreeze protein concentration optimization test

[0043] From the experimental data in Table 1, it can be seen that the optimal concentration of antifreeze protein is usually between 0.5% and 0.7%. This concentration range can provide an ideal freezing effect and ensure the quality and structural integrity of food. Through the optimization of these data, the selection of the concentration of antifreeze protein can be effectively guided, thereby improving the efficiency of food freezing technology and product quality.

[0044] It should be noted that the initial temperature, thickness, water content and other physical characteristics of food are accurately collected through the integrated multi-functional detection module to form the first data set. As the input of this data set, combined with the pre-trained neural network model, the optimal freezing parameters of food are automatically matched, so that different types and states of food can be intelligently processed. This data-driven intelligent control avoids the blindness of fixed parameters in the traditional freezing process and improves the accuracy and efficiency of freezing.

[0045] Furthermore, by spraying natural antifreeze protein nano-coatings with different concentrations on the food surface, the growth of ice crystals can be effectively inhibited, and the damage of ice crystals to food tissues can be reduced. Through experiments on antifreeze protein solutions with different concentrations, the change characteristics of food under different concentrations of antifreeze protein are systematically obtained, and the change evaluation results of food are calculated using the food change evaluation model. This process can dynamically adjust the optimal ratio of the concentration of antifreeze protein, thereby significantly improving the texture and tissue integrity of frozen food and avoiding the freezing damage problems that may be caused by single-concentration treatment in the past.

[0046] And by accumulating freezing data under different foods and different concentrations of antifreeze protein, an empirical database is gradually established. Through the intelligent matching of the first data set, the database can provide the optimal freezing strategy for each type of food, and continuously optimize the content of the database according to the actual production situation to ensure that the freezing effect of each batch of food can reach the best state. With the expansion and accumulation of the database, future application scenarios can more widely cover more types of food, further improving the universality and reliability of this technology.

[0047] S2. Input the first data into the pre-trained neural network model to obtain the second data, where the second data includes the refrigerant flow rate, the target freezing rate, and the pulsed electric field frequency.

[0048] In this embodiment, input the first data into the pre-trained neural network model to obtain the second data, where the second data includes the refrigerant flow rate, the target freezing rate, and the pulsed electric field frequency, specifically as follows:

[0049] Obtain the first data of the food, where the first data includes the initial temperature, thickness, and water content;

[0050] Standardize and encode the first data into a feature vector;

[0051] Input the feature vector into the pre-trained neural network model for mapping to obtain the second data, where the second data includes the refrigerant flow rate, the target freezing rate, and the pulsed electric field frequency;

[0052] Test the refrigeration equipment to be measured to obtain the constraint conditions of the second data.

[0053] The specific structure of the neural network model is as follows:

[0054] This model is a Feedforward Neural Network (FNN), which consists of an input layer, several hidden layers, and an output layer. The number of nodes in the input layer is the same as the number of features in the first data, namely the initial temperature, thickness, and water content, usually 3 input nodes;

[0055] According to actual needs, multiple hidden layers can be set, and each hidden layer contains several nodes. The number of nodes in the hidden layer is usually determined by experience or cross-validation methods, aiming to improve the fitting ability of the model so that it can effectively capture the complex relationships between different features;

[0056] The output layer is set to 3 nodes, corresponding to the refrigerant flow rate, the target freezing rate, and the pulsed electric field frequency respectively. These parameters will be used as the prediction results of the neural network;

[0057] Before inputting the first data into the neural network model, it is necessary to standardize the input data. The purpose of standardization is to eliminate the influence of different dimensions and orders of magnitude on the model training results. The commonly used method is to subtract the mean of each input feature and divide it by the standard deviation to make its distribution a standard normal distribution;

[0058] This neural network is trained with historical experimental data. The historical experimental data includes the known first data (temperature, thickness, water content) and the corresponding second data (refrigerant flow rate, freezing rate, and electric field frequency). The training dataset continuously adjusts the network weights through the Backpropagation algorithm to minimize the error between the prediction result and the actual target value;

[0059] After training, the neural network can be used to predict the second data of new input data. By inputting new food data (such as initial temperature, thickness, and water content) into the trained network, the neural network will output the corresponding refrigerant flow rate, target freezing rate, and pulsed electric field frequency;

[0060] The output second data can be directly used in the control system to adjust the refrigerant flow rate and electric field frequency to optimize the freezing process.

[0061] In this embodiment, the refrigeration equipment to be measured is tested to obtain the constraint conditions of the second data, which are specifically as follows:

[0062] Conduct an extreme test on the refrigerant system of the device under test, gradually increase the refrigerant flow rate until the nozzle is blocked or the pressure exceeds the limit, record the maximum refrigerant flow rate, and screen through variance calculation to obtain the refrigerant flow rate at the minimum stable injection flow rate as the minimum value, so as to obtain the range of the refrigerant flow rate;

[0063] Freeze the food under test at different rates, and conduct a texture analysis on the frozen food to obtain the critical point of the ice crystal size and the cell damage rate, and determine the maximum target freezing rate through the critical point;

[0064] Adjust the pulsed electric field frequency under a fixed field strength, detect the juice loss rate and the microbial survival rate of the thawed food, and obtain the safe range interval of the pulsed electric field frequency.

[0065] It should be noted that the parameter prediction mechanism based on the first food data significantly improves the intelligent level of the freezing process. The present invention uses the initial temperature, thickness and water content of the food as the key input parameters, adopts standardization and vector coding means to form a high-dimensional feature input, and performs a non-linear mapping through a pre-trained neural network model to quickly output the refrigerant flow rate, target freezing rate and pulsed electric field frequency suitable for this type of food. This method realizes the comprehensive consideration and automatic association of multiple physical parameters, effectively breaks through the technical bottleneck of the traditional manual experience-based setting of freezing parameters, and improves the scientificity and reliability of parameter matching.

[0066] Secondly, the introduction of the neural network model in parameter prediction improves the system's fitting ability for the relationship between complex variables. The traditional linear control method is difficult to fully describe the non-linear coupling relationship between multiple variables in the food freezing process, while the present invention trains a large amount of freezing experiment data through a deep learning model, has strong generalization and prediction abilities, and can quickly obtain an approximately optimal freezing control strategy according to the input data, so that the freezing system has fast response and high adaptability.

[0067] Thirdly, the parameter constraint mechanism obtained through equipment testing realizes the precise docking between model prediction and equipment operation ability. The present invention does not directly use the model output results for control, but after the second data is generated, an additional extreme performance test for the freezing equipment is added. This test includes the evaluation of the maximum pressure-bearing capacity and the minimum stable jet state of the refrigerant system, the critical point analysis of the freezing rate affecting the food structure, and the relationship modeling between the electric field frequency and the food microbial control and water retention effect. The upper and lower limits of the parameters obtained from this series of tests constitute the physical constraint boundary of the system, ensuring that all the control parameters output by the model are within the acceptable range of equipment safety and food quality control, and effectively improving the practicality and robustness of the system.

[0068] S3. According to the second data synchronization, start the refrigerant injection and pulsed electric field excitation device, obtain the food dynamic data, and perform feature extraction and trend prediction on the food dynamic data. The food dynamic data includes the real-time core temperature and the dynamic formation of ice crystals.

[0069] In this embodiment, according to the second data synchronization, start the refrigerant injection and pulsed electric field excitation device, obtain the food dynamic data, and perform feature extraction and trend prediction on the food dynamic data. The food dynamic data includes the real-time core temperature and the dynamic formation of ice crystals, which are specifically as follows:

[0070] Place the food to be processed in the freezing equipment, and connect the refrigerant injection system, the pulsed electric field excitation device and the control unit;

[0071] Load the second data obtained by the pre-trained neural network model. The second data includes the target refrigerant flow rate, freezing rate and electric field frequency;

[0072] The control unit starts the refrigerant injection and pulsed electric field excitation device according to the second data synchronization, performs flash freezing on the food, and obtains the food dynamic data. The food dynamic data includes the real-time core temperature of different parts of the food and the formation rate, distribution state and particle size change of ice crystals;

[0073] Perform filtering and noise reduction processing on the food dynamic data, and perform feature extraction on the processed food dynamic data. The features include temperature gradient change, ice crystal growth trend, and freezing stability;

[0074] Predict the extracted features based on the time series method;

[0075] Evaluate the freezing level according to the prediction result.

[0076] In this embodiment, evaluate the freezing level according to the prediction result, which is specifically as follows:

[0077] Arrange multiple temperature sensors inside and on the surface of the food to be frozen to form a temperature monitoring network in three-dimensional space. Each sensor collects temperature data once at a preset period to form a dynamic data sequence of temperature changing with time;

[0078] Based on the temperature data of different sensing points, obtain the local temperature gradient, and output the standard deviation of the local temperature gradient to obtain the overall freezing uniformity feature;

[0079] Analyze the temperature sequence through Fourier transform to obtain the frequency spectrum distribution, extract the main frequency component and the intensity of the high-frequency component, and count the proportion of high-frequency energy to obtain the temperature oscillation frequency;

[0080] If the temperature oscillation frequency is lower than the preset threshold, it is determined that the freezing process is stable;

[0081] Based on the electrical impedance method, the time-varying curve of ice crystal particle size is obtained, and the variance of each window is output by the sliding window method. According to the variance, the ice crystal growth mode is identified. If the growth is stable and the particle size is controlled within the target range (such as 10 - 50 μm), it is determined as fine crystals;

[0082] The freezing uniformity characteristics, particle size, and temperature oscillation frequency are input into the support vector machine, and the freezing level is output.

[0083] In the formula: is the freezing level, is the Lagrange multiplier, the weight coefficient obtained from the training process, reflecting the contribution degree of the support vector to the classification decision surface, is the class label corresponding to the support vector, indicating the freezing level category, such as: 0 - poor freezing, 1 - medium freezing, 2 - excellent freezing, is the kernel function, measuring the current sample and the similarity between the support vectors, is the bias, adjusting the position of the classification boundary, is the freezing uniformity characteristic, is the ice crystal particle size, is the temperature oscillation frequency, is the sample feature vector selected in the training set and having a key influence on the classification boundary, is the feature vector of the current sample, is the number of training samples.

[0084] Exemplarily, through multiple tests on the freezing equipment, its performance and freezing effect under different conditions are verified. The specific test items, methods, data records, and test results are shown in the following table:

[0085]

[0086] Table 2. Food Dynamic Data and Freezing Level Evaluation Form

[0087] From the test data in Table 2 above, it can be seen that the adopted freezing equipment and control method can effectively ensure the stability of the refrigerant flow rate, electric field frequency, and food temperature during the freezing process, thereby achieving an ideal freezing effect and ensuring that the quality and nutritional components of the food are not damaged. When analyzing the temperature oscillation frequency, Fourier transform and calculation methods are used to evaluate the high-frequency components of the temperature oscillation frequency. The specific result is that the high-frequency component accounts for 0.02 Hz. This analysis helps to verify the stability and accuracy of temperature control during the freezing process.

[0088] By real-time monitoring of the dynamic data of food and extracting its features and predicting trends, combined with precise adjustment of the refrigerant flow rate and electric field frequency, the quality and freshness preservation effect during the food freezing process are significantly improved. The specific advantages can be summarized as follows:

[0089] By arranging multiple temperature sensors inside and on the surface of the food to form a three-dimensional temperature monitoring network, temperature data is collected in real time to accurately track the temperature change of the food during the freezing process. By calculating the local temperature gradient and performing standard deviation analysis, the uniformity of the freezing process can be judged in real time. Fourier transform analysis of the temperature oscillation frequency further helps to evaluate whether the freezing process is stable. If the oscillation frequency is lower than the preset threshold, the freezing process can be determined to be in a stable state. Control of temperature uniformity and oscillation frequency can avoid local over-freezing or frostbite of the food during freezing, thereby improving the freshness preservation effect of the food;

[0090] By monitoring the formation rate, distribution state and change of particle size of ice crystals, combined with impedance spectroscopy analysis of the time change of ice crystal particle size, the formation process of ice crystals can be tracked in real time. By using the sliding window method to output the variance of each window and identify the growth pattern of ice crystals, if the growth is stable and the particle size is controlled within the target range (such as 10 - 50 μm), the crystals can be determined to be fine, avoiding cell damage caused by too large ice crystals, thereby reducing the texture change and water loss of the food and improving the taste and nutrition retention of the food;

[0091] By introducing a dynamic adjustment method based on neural network and Bayesian network models to synchronously adjust the refrigerant flow rate, electric field frequency and release rate of antifreeze proteins, this method can optimize the freezing process in real time. Especially during the freezing process, when the prediction result shows a lower freezing level, the system can accelerate the freezing by adjusting the refrigerant flow rate or electric field frequency, thereby ensuring that the ice crystal formation is finer, avoiding uneven freezing process, and improving the freshness preservation effect of the frozen product. The system calculates the optimal release rate of antifreeze proteins based on real-time monitoring data and performs synchronous adjustment, enhancing the antifreeze effect and further improving the quality of the frozen product.

[0092] S4. Dynamically adjust the refrigerant flow rate and electric field frequency according to the prediction result, and synchronously adjust the release rate of antifreeze proteins in the nano-coating.

[0093] In this embodiment, dynamically adjusting the refrigerant flow rate and electric field frequency according to the prediction result, and synchronously adjusting the release rate of antifreeze proteins in the nano-coating are as follows:

[0094] Obtain the causal relationship diagram of the coupled data. The nodes in the causal relationship diagram respectively represent the refrigerant flow rate, electric field frequency, anti-freeze protein release rate, and freezing result. The directed edges of each node represent the causal dependence relationship between variables, that is, the change of one variable will affect the values of other variables. The coupled data includes the freezing level, refrigerant flow rate, electric field frequency, and anti-freeze protein release rate;

[0095] Obtain the historical data of the coupled data, and based on the Bayesian network model training, obtain the conditional probability distribution in the causal relationship diagram. For example, calculate the conditional probability of the refrigerant flow rate and electric field frequency on the anti-freeze protein release rate, or the influence of the electric field frequency and anti-freeze protein release rate on the freezing result;

[0096] Based on the trained Bayesian network model, input the real-time monitored refrigerant flow rate and electric field frequency into the model, and calculate the current anti-freeze protein release rate and the predicted value of the freezing result;

[0097] According to the predicted freezing level, the system automatically adjusts the refrigerant flow rate and electric field frequency. For example, if the predicted freezing level is low, the system may accelerate the freezing process by increasing the refrigerant flow rate or adjusting the electric field frequency to ensure that the ice crystal formation is more delicate. If the freezing process is uneven, the system can adjust the parameters according to the results inferred by the Bayesian network to optimize the freezing effect;

[0098] According to the predicted freezing level, based on the trained Bayesian network model, inversely deduce the optimal anti-freeze protein release rate, and adjust according to the inverse deduction result.

[0099] The directed edges of each node represent the causal dependence relationship between variables, that is, the change of one variable will affect the values of other variables. Specifically, it includes:

[0100] Refrigerant flow rate: Affects the freezing rate, ice crystal formation, and freezing uniformity;

[0101] Electric field frequency: Affects the effect of the electric field on ice crystal nucleation, and thus affects the fineness of ice crystals;

[0102] Anti-freeze protein release rate: Affects the distribution of proteins during the freezing process and the ability to inhibit excessive ice crystal growth;

[0103] Freezing result: Evaluate the freezing quality through the actually monitored ice crystal size, freezing uniformity, etc.

[0104] Furthermore, through the precise adjustment of the refrigerant flow rate, electric field frequency, and anti-freeze protein release rate, the quality and efficiency in the food freezing process are significantly improved, and the ice crystal formation and freezing uniformity are effectively optimized, with many technical advantages:

[0105] Traditional freezing methods often rely on fixed operating parameters, resulting in uneven freezing processes, prone to problems such as overly large ice crystals or insufficient freezing. Through the introduction of a Bayesian network model, this invention can predict the dynamic changes during the freezing process in real time and automatically adjust the refrigerant flow rate, electric field frequency, and anti-freeze protein release rate according to the prediction results, thereby precisely adjusting each parameter under different freezing states. Especially during the freezing process, if the predicted freezing level is low, the system can ensure an accelerated freezing process and fine ice crystal formation by increasing the refrigerant flow rate or adjusting the electric field frequency, avoiding the uneven freezing problem in traditional methods and improving the quality of frozen products;

[0106] By precisely adjusting the refrigerant flow rate and electric field frequency, the uniformity during the freezing process can be optimized. The system continuously evaluates the freezing uniformity by real-time monitoring the core temperature of the food, the formation rate of ice crystals, and the progress of freezing. If unevenness is detected during the freezing process, the system will adjust the refrigerant flow rate and electric field frequency based on real-time data feedback, thus achieving a more uniform freezing effect. The improvement of freezing uniformity not only helps to reduce the temperature difference between the surface and the interior of the food but also effectively inhibits quality problems caused by overly large or small ice crystals during the freezing process;

[0107] An accurate freezing process not only helps to enhance the taste and texture of the food but also effectively retains the nutritional components of the food. By controlling the fineness of ice crystals and the uniformity of freezing, this invention can effectively avoid cell damage during the freezing process, reduce water loss, and retain more nutritional components. The dynamic regulation of anti-freeze proteins further helps to protect the food structure and reduce the quality decline caused by overly large ice crystal growth or freeze damage during the freezing process. Ultimately, the taste, appearance, and nutritional components of the frozen products are retained to the maximum extent.

[0108] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0109] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0110] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0111] In addition, each functional module in various embodiments of the present application may be integrated into one processing module, may exist physically alone for each module, or two or more modules may be integrated into one module.

[0112] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claimed rights.

[0113] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A rapid food preservation control method based on the flash freezing technology, characterized in that, It includes the following steps: S01. Obtain the first data of the food, and spray a nano - coating containing antifreeze proteins with different concentrations on the surface of the food. The first data includes the initial temperature, thickness, and water content; S02. Input the first data into a pre - trained neural network model to obtain the second data. The second data includes the refrigerant flow rate, target freezing rate, and pulsed electric field frequency; S03. According to the second data, synchronously start the refrigerant injection and pulsed electric field excitation device, obtain the food dynamic data, and perform feature extraction and trend prediction on the food dynamic data. The food dynamic data includes the real - time core temperature and the dynamic formation of ice crystals; S04. Dynamically adjust the refrigerant flow rate and electric field frequency according to the prediction result, and synchronously adjust the release rate of antifreeze proteins in the nano - coating; The step S03 includes: Place the food to be processed in a freezing device, and connect the refrigerant injection system and the pulsed electric field excitation device to the control unit; Load the second data obtained from the pre - trained neural network model; The control unit synchronously starts the refrigerant injection and pulsed electric field excitation device according to the second data, performs flash freezing on the food, and obtains the food dynamic data. The food dynamic data includes the real - time core temperature of different parts of the food, as well as the formation rate, distribution state, and particle scale change of ice crystals; Perform filtering and noise reduction processing on the food dynamic data, and perform feature extraction on the processed food dynamic data. The features include temperature gradient change, ice crystal growth trend, and freezing stability; Predict the extracted features based on the time - series method; Evaluate the freezing level according to the prediction result; The step S04 includes: Obtain the causal relationship graph of the coupled data. The coupled data includes the freezing level, refrigerant flow rate, electric field frequency, and release rate of antifreeze proteins; Obtain the historical data of the coupled data, and train the conditional probability distribution in the causal relationship graph based on the Bayesian network model; Based on the trained Bayesian network model, input the real - time monitored refrigerant flow rate and electric field frequency into the model, and output the predicted value of the current release rate of antifreeze proteins and the freezing result; Automatically adjust the refrigerant flow rate and electric field frequency according to the predicted freezing level; According to the predicted freezing level, inversely deduce the optimal release rate of antifreeze proteins based on the trained Bayesian network model, and synchronously adjust the release rate of antifreeze proteins according to the inverse deduction result.

2. The rapid food preservation control method based on the flash freezing technology according to claim 1, wherein The obtaining of the first data of the food and spraying a nano - coating containing antifreeze proteins with different concentrations on the surface of the food, where the first data includes the initial temperature, thickness, and water content, is specifically as follows: Place the food to be processed on a clean platform, remove surface impurities and dry the moisture; Collect the physical characteristic parameters of the food to form a first data set. The first data set includes the initial temperature, thickness, and water content of the food to be processed; Select natural antifreeze proteins with the function of inhibiting ice crystal growth, and divide them into several groups of antifreeze protein solutions with different concentrations; Based on the first data, obtain the food change characteristics after freezing under different antifreeze protein solutions. The food change characteristics include the uniformity of ice crystals, surface heat exchange rate, and texture; Input the food change characteristics into a preset food change evaluation model to output the food change evaluation result, obtain the antifreeze protein solution with the optimal concentration according to the food change evaluation result, and construct an experience database; Call the experience database according to the first data to match the optimal antifreeze protein concentration, and form a uniform film on the food surface through a spraying device.

3. The rapid food preservation control method based on the flash freezing technology according to claim 2, wherein Input the first data into a pre-trained neural network model to obtain the second data, where the second data includes the refrigerant flow rate, the target freezing rate, and the pulsed electric field frequency, specifically as follows: Obtain the first data of the food, standardize and encode the first data into a feature vector; Input the feature vector into a pre-trained neural network model for mapping to obtain the second data, where the second data includes the refrigerant flow rate, the target freezing rate, and the pulsed electric field frequency; Test the to-be-tested freezing device to obtain the constraint conditions of the second data.

4. The rapid food preservation control method based on the flash freezing technology according to claim 3, characterized in that, The testing of the to-be-tested freezing device to obtain the constraint conditions of the second data is specifically as follows: Conduct an extreme test on the refrigerant system of the to-be-tested device, gradually increase the refrigerant flow rate until the nozzle is blocked or the pressure exceeds the limit, and record the maximum refrigerant flow rate; Screen through variance calculation to obtain the refrigerant flow rate at the minimum stable injection flow rate as the minimum value, and obtain the range of the refrigerant flow rate; Freeze the to-be-tested food at different rates, and conduct a texture analysis on the frozen food to obtain the critical point of the ice crystal size and the cell breakage rate, and determine the maximum target freezing rate through the critical point; Adjust the pulsed electric field frequency under a fixed field strength, detect the juice loss rate and the microbial survival rate of the thawed food, and obtain the safety range interval of the pulsed electric field frequency.

5. The rapid food preservation control method based on the flash freezing technology according to claim 4, characterized in that, The evaluation of the freezing level according to the prediction result is specifically as follows: Arrange multiple temperature sensors inside and on the surface of the food to be frozen to form a temperature monitoring network in three-dimensional space. Each sensor collects temperature data once at a preset period to form a dynamic data sequence of temperature changing with time; Based on the temperature data of different sensing points, obtain the local temperature gradient, and output the standard deviation of the local temperature gradient to obtain the overall freezing uniformity characteristics; Analyze the temperature sequence through Fourier transform to obtain the frequency spectrum distribution, extract the main frequency component and the high-frequency component intensity, and statistically calculate the high-frequency energy ratio as the temperature oscillation frequency; Obtain the time-varying curve of the ice crystal particle size based on the electrical impedance method, and output the variance of each window through the sliding window method, and identify the ice crystal growth mode according to the variance; Input the freezing uniformity characteristics, particle size, and temperature oscillation frequency into a support vector machine to output the freezing level.

Images