Control methods, apparatus, equipment and storage media for the preparation of hypotonic nutritional foods

Abstract

This invention provides a control method, apparatus, equipment, and storage medium for the preparation of low-osmotic nutritional foods, relating to the field of nutritional product preparation control technology. The method includes: obtaining the measured osmotic pressure of a mixture of materials to be prepared as low-osmotic nutritional foods; measuring the osmotic pressure of a sample of the mixture using the freezing point method; inputting the measured osmotic pressure into a trained linear regression analysis model to obtain a predicted value for intestinal motility promotion rate; this model is trained based on the osmotic pressure of reconstituted milk powder and the intestinal motility promotion rate of zebrafish; determining an optimized preparation process scheme based on the predicted intestinal motility promotion rate and intestinal motility promotion rate warning conditions, and adjusting process parameters according to the optimized preparation process scheme; wherein, the process parameters in the optimized preparation process scheme include one or more of the following: the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature. This invention can quickly complete functional evaluation and optimize preparation parameters.

Description

Technical Field

[0001] This invention relates to the field of nutritional product preparation control technology, and in particular to a control method, apparatus, equipment and storage medium for the preparation of low-osmotic nutritional foods. Background Technology

[0002] During infancy and early childhood, the intestines are sensitive and fragile. Consuming foods with high osmotic pressure can easily cause strong intestinal stimulation and accelerate intestinal peristalsis, resulting in discomfort such as diarrhea or vomiting. The appropriate rate of intestinal stimulation exhibited by infants and young children after eating can, to some extent, reflect the good adaptability of the food to their intestines.

[0003] In the field of infant formula processing, the relationship between osmotic pressure regulation and intestinal adaptability involves complex physiological processes. Existing technologies have four major shortcomings in this field: First, they use fixed electrolyte ratios without dynamic adjustment based on infant age and intestinal maturity, lacking individualized adaptation; second, the ratio of extensively hydrolyzed proteins is unreasonable, as although they have low allergenicity, the short peptide chains increase the osmotic pressure of the solution, creating a contradiction of "low allergenicity - low osmotic pressure"; third, the thermal denaturation of whey proteins during spray drying makes the actual osmotic pressure 15% to 20% higher than the theoretical value; and fourth, a correlation has not been established between osmotic pressure and intestinal motility evaluation, making it impossible to quickly assess intestinal adaptability.

[0004] The existing technological system urgently needs to shift from simple component adjustment to a systematic solution encompassing "low-osmotic formulation design - low-osmotic preparation process - intestinal adaptability evaluation." This requires breakthroughs in traditional nutritional frameworks, food processing, biology, and other multidisciplinary technologies to ultimately achieve precise osmotic pressure control in infant formula. Therefore, existing technologies have significant shortcomings in achieving rapid functional evaluation, effective osmotic reduction, and synergistic optimization of components. Summary of the Invention

[0005] This invention provides a control method, apparatus, equipment, and storage medium for the preparation of low-osmotic nutritional foods to address the issues of rapid functional evaluation, effective osmotic reduction, and synergistic optimization of components.

[0006] In a first aspect, embodiments of the present invention provide a method for controlling the preparation of low-osmotic nutritional foods, comprising:

[0007] The osmotic pressure of the mixture of materials to be prepared into a low-osmotic nutritional food is obtained; wherein the osmotic pressure is determined by measuring the osmotic pressure of the material sample of the mixture using the freezing point method;

[0008] The measured osmotic pressure is input into a trained linear regression analysis model to obtain a predicted value of the intestinal motility promotion rate; wherein, the linear regression analysis model is trained based on the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish;

[0009] Based on the predicted value of the intestinal peristalsis promotion rate and the early warning condition of the intestinal peristalsis promotion rate, an optimization scheme for the preparation process is determined, and the process parameters are adjusted according to the optimization scheme.

[0010] The process parameters in the optimized preparation process include one or more of the following: the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature; the types of proteins include hydrolyzed whey protein, hydrolyzed casein, and non-hydrolyzed whey protein.

[0011] In one possible implementation, the control method further includes:

[0012] Prepare n groups of nutritional food samples with different formulations. Each group of samples is measured m times to obtain data on osmotic pressure and zebrafish intestinal peristalsis promotion rate, and obtain n×m groups of raw data.

[0013] Preprocess the n×m sets of original data, and divide the preprocessed data into training set and test set;

[0014] Based on the training set, the initial coefficients are obtained by fitting the initial linear regression analysis model using the least squares method.

[0015] Five-fold cross-validation was used to optimize the parameters of the initial linear regression analysis model by minimizing the root mean square error, resulting in the optimized linear regression analysis model.

[0016] The optimized linear regression analysis model was validated based on the training set and test set, and the mean deviation, maximum deviation and coefficient of determination were selected as evaluation indicators.

[0017] When all evaluation indicators meet the discrimination criteria, a well-trained linear regression analysis model is obtained.

[0018] In one possible implementation, the preprocessing of the n×m sets of original data includes:

[0019] Outliers are removed using Z-score standardization.

[0020] In one possible implementation, the outlier is a value that deviates from a set range of the mean; wherein the set range is 3 times the standard deviation.

[0021] In one possible implementation, the preparation of n groups of nutritional food samples with different formulations, with each group of samples being measured m times repeatedly to obtain osmotic pressure and zebrafish intestinal motility promotion rate data, includes:

[0022] According to the mixing ratio indicated on the nutritional food label, mix milk powder and solvent to prepare n sets of nutritional food samples with different formulations.

[0023] The osmotic pressure of n groups of nutritional food samples with different formulations was measured using the freezing point method, and the osmotic pressure data were recorded.

[0024] Female and male zebrafish were mated to screen for zebrafish with swim bladder stage. Nile red working solution of a set concentration was prepared and the screened zebrafish were treated with Nile red working solution to construct an intestinal fluorescent labeling model.

[0025] Each group of samples was applied to the intestinal fluorescent labeling model and measured m times. After the application was completed, the zebrafish corresponding to each group of intestinal fluorescent labeling models were placed under a fluorescence microscope to take pictures and obtain intestinal fluorescence images. The intestinal peristalsis promotion rate of zebrafish was determined based on the intestinal fluorescence images.

[0026] In one possible implementation, obtaining n×m sets of original data includes:

[0027] The osmotic pressure data and zebrafish intestinal motility promotion rate data corresponding to a single measurement were used as a set of data;

[0028] Based on the data of each group of n different formula nutritional food samples, construct n×m sets of original data.

[0029] In one possible implementation, the linear regression analysis model is:

[0030] Y=β0+β1X

[0031] Where Y is the zebrafish intestinal peristalsis promotion rate, X is the measured osmotic pressure, β0 is a constant term, and β1 is the regression coefficient of osmotic pressure.

[0032] Secondly, embodiments of the present invention provide a control device for preparing low-osmotic nutritional foods, comprising:

[0033] The acquisition module is used to acquire the measured osmotic pressure of the mixture of materials to be prepared into a low-osmotic nutritional food; wherein the measured osmotic pressure is obtained by measuring the osmotic pressure of the material sample of the mixture using the freezing point method;

[0034] The prediction module is used to input the measured osmotic pressure into a trained linear regression analysis model to obtain a predicted value of the intestinal motility promotion rate; wherein, the linear regression analysis model is trained based on the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish.

[0035] The control module is used to determine the preparation process optimization scheme based on the predicted value of the intestinal peristalsis promotion rate and the early warning condition of the intestinal peristalsis promotion rate, so as to adjust the process parameters according to the preparation process optimization scheme;

[0036] The process parameters in the optimized preparation process include one or more of the following: the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature; the types of proteins include hydrolyzed whey protein, hydrolyzed casein, and non-hydrolyzed whey protein.

[0037] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect or any possible implementation thereof.

[0038] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect or any possible implementation thereof.

[0039] In this embodiment of the invention, the osmotic pressure of the mixture of ingredients for the low-osmotic nutritional food to be prepared is accurately measured by the freezing point method. This osmotic pressure is then input into a linear regression analysis model trained on the osmotic pressure of milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish, rapidly obtaining a predicted value for the intestinal motility promotion rate without relying on traditional time-consuming animal experiments or clinical observations. Subsequently, based on this predicted value and early warning conditions, an optimized preparation process scheme is determined. Key process parameters such as the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature are adjusted accordingly. This allows for synergistic control of the osmotic pressure of the low-osmotic nutritional food to be prepared from both the formulation and preparation process perspectives, effectively reducing the osmotic pressure of its reconstituted solution and minimizing the risk of intestinal irritation, diarrhea, or vomiting in infants caused by high osmotic pressure. Simultaneously, it enables dynamic control of the low-osmotic nutritional food preparation process, ensuring that the final product has good intestinal adaptability. Attached Figure Description

[0040] Figure 1 This is an application scenario diagram of the control method for preparing low-osmotic nutritional food provided in an embodiment of the present invention;

[0041] Figure 2 This is a flowchart illustrating the implementation of a control method for preparing low-osmotic nutritional food according to an embodiment of the present invention;

[0042] Figure 3 This is a flowchart of the preparation process of a low-osmotic formula nutritional food according to an embodiment of the present invention;

[0043] Figure 4 This is a fluorescence imaging image of the zebrafish intestine provided in an embodiment of the present invention;

[0044] Figure 5 This is a schematic diagram of the control device for preparing low-osmotic nutritional food according to an embodiment of the present invention;

[0045] Figure 6 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0046] This application aims to develop an innovative low-osmotic nutritional food formula with lower osmotic pressure, better component synergy, and efficient evaluation methods, along with supporting technologies. The low-osmotic nutritional food is a dairy product, such as milk powder. This application provides a linear regression analysis model to evaluate the intestinal motility promotion rate. This model can be used not only to control the process parameters of low-osmotic nutritional food preparation but also for quality control and consumer selection. For example, in the factory testing process of low-osmotic nutritional foods, osmotic pressure can be measured to predict intestinal adaptability, replacing the traditional time-consuming zebrafish experiment, shortening the testing process from 72 hours or more using the zebrafish experiment to approximately 10 minutes. Suitable low-osmotic nutritional products can be recommended based on the results of the linear regression analysis model for different age groups and those with a history of diarrhea.

[0047] The control method for preparing low-osmotic nutritional food provided in the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0048] Figure 1 This is an application scenario diagram of the control method for preparing low-osmotic nutritional food provided in an embodiment of the present invention. For example... Figure 1 As shown, it includes a freezing point osmotic pressure meter, a control terminal, and nutritional food preparation equipment.

[0049] In the specific implementation process, samples of the mixture to be prepared into low-osmotic nutritional food are taken, and the samples are placed in a freezing point osmometer to obtain the osmotic pressure. The control terminal communicates with the freezing point osmometer via wired or wireless means to obtain the corresponding osmotic pressure of the mixture, and predicts the intestinal motility promotion rate based on the measured osmotic pressure and a linear regression analysis model. The nutritional food corresponding to the mixture is evaluated based on the intestinal motility promotion rate. If it does not meet the preparation requirements, the process parameters of the nutritional food preparation equipment are adjusted under the guidance of the predicted intestinal motility promotion rate.

[0050] Figure 1 In the illustrated embodiment, the control terminal is independent of the nutritional food preparation equipment, facilitating remote control by staff. In other possible implementations, the freezing point osmotic pressure meter communicates directly with the controller within the nutritional food preparation equipment system, eliminating the need for an additional control terminal.

[0051] Figure 2 This is a flowchart illustrating the implementation of a control method for preparing low-osmotic nutritional foods according to an embodiment of the present invention. Figure 2 As shown, it includes the following steps:

[0052] S201, Obtain the measured osmotic pressure of the mixture of materials to be prepared as a low-osmotic nutritional food; wherein, the measured osmotic pressure is obtained by measuring the osmotic pressure of the material sample of the mixture using the freezing point method.

[0053] The execution entity in the various embodiments of this application can be a server, processor, microprocessor, or other device with data processing capabilities. In actual implementation, the specific implementation method of the execution entity can be selected according to actual needs. This embodiment does not impose any particular limitation on this; any device with data processing capabilities is acceptable. For ease of understanding, in subsequent embodiments, [the following will be used as an example]. Figure 1 The control terminal shown is the execution entity for explanation.

[0054] In the preparation of hypotonic nutritional foods, after the raw materials are mixed according to a preset formula to form a mixture, a sample is extracted from the mixing system. The osmotic pressure of this sample is measured using the freezing point method. Utilizing the quantitative relationship between solution osmotic pressure and freezing point depression, the osmotic pressure is calculated by measuring the degree of freezing point depression of the sample, thus obtaining the measured osmotic pressure. This method is simple to operate and provides stable results, accurately reflecting the osmotic pressure characteristics of the mixture.

[0055] Among them, the application ratio of different proteins is optimized for the mixture of ingredients to be prepared low-osmotic nutritional foods, and the molecular weight distribution range of hydrolyzed proteins is controlled. While ensuring that functional proteins that are easy to digest and improve immunity are present, high osmotic pressure caused by over-reliance on small molecule proteins is avoided.

[0056] Furthermore, optimizing the ratio of exogenous calcium to milk-derived calcium reduces the level of excess free calcium in milk-derived calcium, and further reduces the level of free calcium through the chelation effect of potassium citrate and sodium citrate, thereby lowering the osmotic pressure. In addition, the optimized calcium-to-phosphorus ratio is more conducive to the formation of more stable complexes (such as calcium phosphate clusters) of calcium, rather than its existence as free ions with high osmotic pressure.

[0057] S202, the measured osmotic pressure is input into the trained linear regression analysis model to obtain the predicted value of the intestinal peristalsis promotion rate; wherein, the linear regression analysis model is trained based on the osmotic pressure of milk powder reconstituted solution and the intestinal peristalsis promotion rate of zebrafish.

[0058] The linear regression analysis model was trained in advance using experimental data. The training process used the osmotic pressure of the milk powder reconstituted solution as the input variable and the zebrafish intestinal peristalsis promotion rate as the output variable. The correlation between the two was established through linear fitting.

[0059] Since zebrafish have certain similarities in intestinal physiology to mammals, and their intestinal peristalsis can be observed and quantified through standardized experiments, the predicted intestinal peristalsis promotion rate obtained based on this model can effectively reflect the trend of the influence of nutritional foods on intestinal peristalsis.

[0060] S203. Based on the predicted value of intestinal motility promotion rate and the early warning condition of intestinal motility promotion rate, determine the preparation process optimization scheme, and adjust the process parameters according to the preparation process optimization scheme; wherein, the process parameters in the preparation process optimization scheme include one or more of the following: the content range of different types of protein, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure and spray drying temperature; the types of protein include hydrolyzed whey protein, hydrolyzed casein and non-hydrolyzed whey protein.

[0061] In practice, the warning criteria for the intestinal motility promotion rate are based on the intestinal tolerance standards of the target population (such as infants, the elderly, and postoperative patients). For example, when the predicted value is higher than a certain threshold, it indicates a higher risk of intestinal adaptation.

[0062] In one possible implementation, an optimized preparation process is determined based on the predicted value of the intestinal motility promotion rate and the early warning conditions for the intestinal motility promotion rate, including:

[0063] Compare the predicted value of intestinal motility promotion rate with the early warning conditions;

[0064] If the predicted value meets the warning conditions (such as being within a suitable range), then the current process parameters shall be maintained.

[0065] If the predicted value does not meet the early warning conditions (such as exceeding the risk threshold), the preparation process optimization scheme is determined based on the correlation feedback from the model, and the process parameters are adjusted accordingly.

[0066] Optional, adjustable process parameters include, but are not limited to, the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature.

[0067] Among them, the content range of different types of proteins: by optimizing the application ratio of hydrolyzed whey protein, hydrolyzed casein and non-hydrolyzed whey protein, the molecular weight distribution range of hydrolyzed protein is controlled, thereby affecting the osmotic pressure of the material;

[0068] The ratio of exogenous calcium to milk-derived calcium: By adjusting the ratio of the two calcium sources, the contribution of minerals to osmotic pressure is balanced;

[0069] Homogenization pressure: By changing the homogenization pressure, the dispersion state and particle size distribution of the material are affected, thus indirectly regulating the osmotic pressure;

[0070] Spray drying temperature: By adjusting the drying temperature, the moisture content and particle morphology of the material can be controlled, which helps to optimize the osmotic pressure.

[0071] According to the process optimization plan, the above adjustments can bring the osmotic pressure of the mixture closer to the target range, ultimately ensuring that the low-osmotic nutritional food has a suitable effect on promoting intestinal peristalsis and reducing the risk of intestinal discomfort.

[0072] In different embodiments, the process parameters in the process optimization scheme include one or more. In one specific embodiment, by controlling the homogenization pressure and optimizing the spray drying temperature in a synergistic manner, the particle size D50 of the milk powder reconstituted solution is controlled to be ≥0.18μm, thereby reducing the osmotic pressure; at the same time, the loss of heat-sensitive substances is reduced.

[0073] In the specific implementation process, based on the control method for preparing low-osmotic nutritional foods provided in this application, the synergistic effect of optimizing the formula combination and improving the process can significantly reduce the reconstituted osmotic pressure of milk powder to about 280 mOsm / kg (the traditional formula is 315 mOsm / kg), effectively reducing the risk of intestinal discomfort.

[0074] In this embodiment, the osmotic pressure of the mixture of ingredients for the low-osmotic nutritional food to be prepared is accurately measured using the freezing point method. This osmotic pressure is then input into a linear regression analysis model trained on the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish, rapidly obtaining a predicted value for the intestinal motility promotion rate without relying on traditional time-consuming animal experiments or clinical observations. Subsequently, based on this predicted value and early warning conditions, an optimized preparation process scheme is determined. Key process parameters such as the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature are adjusted accordingly. This allows for synergistic control of the osmotic pressure of the low-osmotic nutritional food from both the formulation and preparation process perspectives, effectively reducing the osmotic pressure of its reconstituted solution and minimizing the risk of intestinal irritation, diarrhea, or vomiting in infants caused by high osmotic pressure. Simultaneously, it enables dynamic control of the low-osmotic nutritional food preparation process, ensuring that the final product has good intestinal adaptability.

[0075] Based on the aforementioned embodiments, in addition to determining the optimized preparation process scheme according to the predicted value of intestinal motility promotion rate, the composition of the mixture of materials to be prepared low-osmotic nutritional food is also an important factor in ensuring the preparation efficiency of low-osmotic nutritional food and the good intestinal adaptability of the final product.

[0076] In one specific embodiment, the low-osmotic nutritional food is a powdered food, relating to a low-osmotic-pressure infant formula milk powder, the composition of which (per 100g of milk powder) includes essential components and selective components, as detailed below:

[0077] (1) Proteins

[0078] Hydrolyzed whey protein: 0.5-3g, processed using a specific enzymatic hydrolysis process, with the proportion of components with a molecular weight ≤5kDa >80%, optimizing intestinal absorption efficiency;

[0079] Hydrolyzed casein: 0.1-1g, processed with a specific enzymatic hydrolysis process to ensure that the proportion of components with a molecular weight ≤3kDa is >90%, thereby improving protein digestibility and utilization.

[0080] Non-hydrolyzed whey protein: 9.5-16g, containing 0.02-0.04g of immunoglobulin (IgG), which can form a slow-release absorption structure and enhance immune function.

[0081] (2) Minerals

[0082] Calcium (Ca): 280-500mg, preferably a mixture of exogenous calcium and milk-derived calcium in a ratio of 1:(1-1.5);

[0083] Phosphorus (P): 190-300mg. Through innovative optimization of the calcium source ratio, the calcium-to-phosphorus ratio is controlled at 1:(0.55-0.80) to balance the level of free calcium ions in the milk powder particles.

[0084] (3) Other materials

[0085] It is composed of raw milk, lactose, compound minerals, compound vitamins, blended vegetable oil, choline chloride, potassium citrate, sodium citrate, arachidonic acid oil powder, docosahexaenoic acid oil powder, galactooligosaccharides (GOS), 2'-alglucosyl lactose (2'-FL), nucleotides, etc., to ensure nutritional balance and synergistic function.

[0086] Figure 3 This is a flowchart illustrating the preparation process of a low-osmotic formula nutritional food according to an embodiment of the present invention, as shown below. Figure 3 As shown, it includes at least the following main steps:

[0087] (1) Mixing: Raw milk, lactose, compound minerals, compound vitamins, edible vegetable oil, choline chloride, potassium citrate, sodium citrate, arachidonic acid oil powder, docosahexaenoic acid oil powder, nucleotides, hydrolyzed whey protein, hydrolyzed casein, non-hydrolyzed whey protein, prebiotics and other materials other than immunoglobulin (IgG) are mixed in a set ratio to obtain premixed materials;

[0088] (2) Homogenization: Homogenize the premixed materials and control the homogenization pressure to 145-185 bar to ensure that the materials are mixed evenly and maintain a suitable particle size level (D50≥0.18μm).

[0089] (3) Sterilization and concentration: The homogenized material is sterilized and the water is removed by the concentration process to obtain the concentrate;

[0090] (4) Spray drying: The concentrate is sent into the spray drying equipment, the inlet air temperature is controlled at 170-175℃ and the exhaust air temperature is controlled at 70-90℃, and semi-finished milk powder is obtained after drying.

[0091] (5) Physical mixing (dry mixing): Immunoglobulin (IgG) is physically mixed with the semi-finished milk powder obtained from the spray drying step. By controlling the mixing conditions, the biological activity of immunoglobulin is preserved, and finally low osmotic pressure infant formula is obtained.

[0092] The above scheme mainly introduces the calculation of the predicted value of the intestinal peristalsis promotion rate using a trained linear regression analysis model, the determination of the preparation process optimization scheme based on the predicted value of the intestinal peristalsis promotion rate, and the adjustment of process parameters based on the preparation process optimization scheme. The following focuses on the training process of the linear regression analysis model.

[0093] In one possible implementation, the method further includes:

[0094] Prepare n groups of nutritional food samples with different formulations. Each group of samples is measured m times to obtain data on osmotic pressure and zebrafish intestinal peristalsis promotion rate, and obtain n×m groups of raw data.

[0095] Preprocess the n×m sets of original data, and divide the preprocessed data into training set and test set;

[0096] Based on the training set, the initial linear regression analysis model is fitted using the least squares method to obtain the initial coefficients;

[0097] Five-fold cross-validation was used to optimize the parameters of the initial linear regression analysis model by minimizing the root mean square error, resulting in the optimized linear regression analysis model.

[0098] The optimized linear regression analysis model was validated based on the training set and the test set, and the mean deviation, maximum deviation and coefficient of determination were selected as evaluation indicators.

[0099] When all evaluation indicators meet the discrimination criteria, a well-trained linear regression analysis model is obtained.

[0100] This study prepared n groups of nutritional food samples with different formulations, each group differing in raw material composition or process parameters to cover a broad range of formulation variables. For each group of nutritional food samples, the osmotic pressure was repeatedly measured m times using the freezing point method, and the intestinal motility promotion rate was simultaneously measured m times using a zebrafish model, thus obtaining n×m sets of raw data. Each set of data includes the osmotic pressure value and intestinal motility promotion rate value of the corresponding sample. Multiple repeated measurements reduce the impact of single-experiment errors on data reliability, providing rich and stable basic data for model training.

[0101] The above n×m sets of original data are preprocessed to remove outliers or noise, making the data distribution more suitable for model training. After preprocessing, the dataset is divided into training and test sets. The training set is used for parameter fitting, optimization, and validation of the model, while the test set is used for subsequent model validation. The ratio of the two sets can be determined based on the size and distribution characteristics of the data to ensure the objectivity of model training and validation.

[0102] Based on the partitioned training set, the initial linear regression analysis model was fitted using the least squares method. By minimizing the sum of squares between the actual observed values ​​and the model's predicted values, the initial coefficients in the model were solved, enabling the initial model to initially reflect the linear correlation between osmotic pressure and intestinal motility promotion rate, laying the foundation for subsequent model optimization.

[0103] Five-fold cross-validation was used to fine-tune the parameters of the initial linear regression model. Specifically, the training set was randomly divided into five mutually exclusive subsets. Four subsets were selected sequentially as training data, and one subset was selected as validation data. This process was repeated five times, adjusting the model parameters with the goal of minimizing the root mean square error. Cross-validation effectively avoids overfitting and improves the model's generalization ability to unknown data, ultimately resulting in the optimized linear regression model.

[0104] The optimized model was validated using training and test sets, with mean deviation, maximum deviation, and coefficient of determination selected as core evaluation metrics. Mean deviation and maximum deviation measure the overall deviation and extreme deviation between the model's predicted values ​​and the actual measured values, while the coefficient of determination evaluates the model's ability to explain data trends. When all evaluation metrics meet the preset criteria, such as the mean deviation being within a reasonable range and the coefficient of determination reaching a threshold, it indicates that the model's fitting accuracy and stability meet the requirements, and the model can then be determined as a well-trained linear regression analysis model.

[0105] In this embodiment, the method prepares multiple sets of nutritional food samples with different formulations and repeatedly measures them to obtain a large amount of raw data on osmotic pressure and zebrafish intestinal motility promotion rate. This provides sufficient and diverse data sources for model training, avoiding model bias caused by single data. After preprocessing the raw data, it is divided into training and test sets to ensure data quality and dataset rationality. Based on the training set, the initial model is fitted using the least squares method to obtain initial coefficients, laying the foundation for model construction. Then, the model parameters are optimized through five-fold cross-validation with the goal of minimizing the root mean square error, improving the model fitting accuracy. Finally, the optimized model is validated using the training and test sets, and the mean deviation, maximum deviation, and coefficient of determination are used as evaluation indicators to ensure that the model has good fitting effect and generalization ability when meeting the discrimination criteria. This provides reliable model support for subsequent prediction of intestinal motility promotion rate of nutritional foods and guidance on optimization of nutritional food preparation processes, ensuring the accuracy and stability of nutritional food intestinal adaptability evaluation.

[0106] In one possible implementation, preprocessing of the n×m sets of original data includes:

[0107] Outliers are removed using Z-score standardization.

[0108] In the specific implementation process, for the osmotic pressure data and zebrafish intestinal peristalsis promotion rate data in the n×m groups of original data, their respective means and standard deviations were calculated. The mean is the average of each group of data, reflecting the central tendency of the data; s is the standard deviation, reflecting the dispersion of the data.

[0109] Each set of data is standardized using the Z-score formula, and the Z-score for each data point is calculated using the following formula:

[0110] Z=(x-μ) / s

[0111] Where x is a single data value, μ is the mean of the data set, and s is the standard deviation of the data set. The Z-score represents the distance between a single data point and the mean, expressed in standard deviations, and quantifies the degree to which a data point deviates from the overall distribution.

[0112] Outliers are identified and removed based on Z-scores. When the Z-score of a data point exceeds a preset range, that data point is identified as an outlier and removed. This process effectively eliminates the interference of outliers caused by random errors, operational deviations, or extreme experimental conditions on the overall dataset, making the preprocessed data more reflective of the true correlation between osmotic pressure and intestinal motility promotion rate in nutritional foods.

[0113] In this embodiment, Z-score standardization is used to remove outliers from the n×m sets of original data. This effectively eliminates the interference of extreme data caused by measurement errors, experimental operation deviations, and other factors on model training. Standardization makes the data distribution more consistent with the requirements of model training, avoids deviations in model fitting due to outliers, and improves the reliability and consistency of training data. This provides a high-quality data foundation for building an accurate linear regression analysis model, ensuring that the model accurately reflects the correlation between the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish, guaranteeing the accuracy of the model's prediction results, and providing a valid basis for optimizing milk powder preparation processes and evaluating intestinal adaptability.

[0114] In one possible implementation, outliers are values ​​that deviate from a set range of the mean; where the set range is 3 times the standard deviation.

[0115] In practice, for the osmotic pressure data and zebrafish intestinal peristalsis promotion rate data in the n×m sets of original data, they are filtered according to their respective Z scores: if the Z score of a data point is >3 or Z score <-3, it means that the deviation of the data point from the mean is more than 3 times the standard deviation, and it is judged as an outlier and removed from the dataset; the remaining data points with Z scores in the range of -3 to 3 are retained as valid data.

[0116] In this embodiment, outliers are defined as those deviating from the mean by three times the standard deviation. This allows for the accurate identification of extreme values ​​in the original data that significantly deviate from the normal data distribution. Retaining these extreme values ​​would interfere with the model's accurate learning of the relationship between osmotic pressure and intestinal motility promotion rate. By removing such outliers, the training data can be further purified, enabling the data to more accurately reflect the inherent laws between nutritional food formulations, processes, and intestinal adaptability. This improves the accuracy of model training and ensures that subsequent model-based predictions and process optimization schemes are more reliable, ultimately contributing to the production of low-osmotic nutritional foods with better intestinal adaptability.

[0117] In one possible implementation, n groups of nutritional food samples with different formulations are prepared, and each group of samples is measured m times repeatedly to obtain osmotic pressure and zebrafish intestinal motility promotion rate data, including:

[0118] According to the mixing ratio indicated on the nutritional food label, mix milk powder and solvent to prepare n sets of nutritional food samples with different formulations.

[0119] The osmotic pressure of n groups of nutritional food samples with different formulations was measured using the freezing point method, and the osmotic pressure data were recorded.

[0120] Female and male zebrafish were mated to select zebrafish with the swim bladder stage. A Nile Red working solution of a predetermined concentration was prepared, and the selected zebrafish were treated with the Nile Red working solution to construct an intestinal fluorescent labeling model. The concentration was set at 5 μg / L to 10 μg / L to ensure clear observation of the Nile Red filling in the zebrafish intestine under a fluorescence microscope. Optionally, the concentration could be set to 5 μg / L, 8 μg / L, or 10 μg / L.

[0121] Each group of samples was applied to the intestinal fluorescent labeling model and measured m times. After the application was completed, the zebrafish corresponding to each group of intestinal fluorescent labeling models were placed under a fluorescence microscope to take pictures and obtain intestinal fluorescence images. The intestinal peristalsis promotion rate of zebrafish was determined based on the intestinal fluorescence images.

[0122] In this study, for n groups of nutritional food products with different formulas that differed in raw material composition or proportions, milk powder and solvent were mixed according to the reconstitution ratio indicated on the labels of each group of nutritional food products. The mixture was stirred thoroughly to completely dissolve the milk powder, resulting in n groups of reconstituted milk powder samples. This operation simulated a real-world drinking scenario, ensuring that the sample state was consistent with the milk powder solution used by infants and other young children, thus making the subsequent osmotic pressure data more practically meaningful.

[0123] The osmotic pressure of the above n groups of reconstituted milk powder samples was measured using the freezing point method. Specifically, the osmotic pressure was calculated by detecting the freezing point depression of the samples, and the measurement results for each group were recorded. To reduce random errors, each group of samples was measured m times, resulting in m osmotic pressure data points, which were used as the osmotic pressure dataset for that group of samples.

[0124] Healthy female and male zebrafish were selected and mated. After embryonic development, zebrafish juveniles in the swim bladder stage were selected. Zebrafish have relatively complete intestinal development, making them suitable for intestinal function evaluation. A Nile Red working solution of a predetermined concentration was prepared, and the selected zebrafish juveniles were placed in the Nile Red working solution and cultured under suitable conditions for a period of time to allow the dye to fully label the intestinal contents, thus constructing a standardized intestinal fluorescent labeling model. The suitable conditions were those conducive to the survival of zebrafish juveniles and maintaining their intestinal activity.

[0125] n groups of nutritional food samples were applied to intestinal fluorescently labeled models, with each group being treated m times. A new batch of intestinal fluorescently labeled models was used for each treatment. After the preset treatment time, zebrafish from each group were placed under a fluorescence microscope, and intestinal fluorescence images were captured under the same imaging parameters. By analyzing the migration distance or distribution changes of fluorescent substances in the intestine within the fluorescence images, the intestinal peristalsis intensity was quantified, and the intestinal peristalsis promotion rate for each group of samples was calculated.

[0126] In this embodiment, samples of different formula nutritional foods were prepared according to the mixing ratios on the nutritional food labels to ensure that the mixed state of the samples was consistent with actual consumption, making the osmotic pressure measurement results more practically meaningful. The freezing point method was used to measure osmotic pressure to ensure the accuracy of the measurement data. Simultaneously, a standardized intestinal fluorescent labeling model was constructed through standardized zebrafish mating, screening, and Nile Red working solution processing procedures to reduce the impact of experimental operation differences on the model. Each group of samples was repeatedly applied to the model, and the intestinal motility promotion rate was determined by photographing and analyzing the fluorescence images using a fluorescence microscope, resulting in stable and reliable measurement results. The entire process, from sample preparation and osmotic pressure measurement to zebrafish model construction and intestinal motility promotion rate measurement, followed a standardized procedure, ensuring the authenticity, accuracy, and repeatability of the original data. This provides high-quality data for subsequent model training, supporting the model to accurately reflect the correlation between nutritional food osmotic pressure and intestinal adaptability.

[0127] In one possible implementation, n×m sets of original data are obtained, including:

[0128] The osmotic pressure data and zebrafish intestinal motility promotion rate data corresponding to a single measurement were used as a set of data;

[0129] Based on the data of each group of n different formula nutritional food samples, construct n×m sets of original data.

[0130] This method uses the osmotic pressure data from a single measurement and the corresponding zebrafish intestinal motility promotion rate data as one set of data. Then, it combines the repeated measurement results of n sets of nutritional food samples with different formulations to construct n×m sets of original data. This ensures that each set of data can accurately correspond to the osmotic pressure and intestinal adaptability-related indicators in a single experiment, thus avoiding data confusion.

[0131] In the embodiments of this application, n ≥ 3 and m ≥ 10, optionally.

[0132] In the specific implementation process, considering that the model is prone to underfitting or overfitting when the sample size is too small, the values ​​of n and m are set to ensure that enough effective data samples can be obtained during the model training phase, so that the regression analysis results of the final linear regression analysis model are more accurate and reliable, and can truly reflect the correlation between variables.

[0133] In this embodiment, the data construction method of constructing n×m sets of original data based on the data of each set of n sets of nutritional food samples with different formulations can completely retain the independent information of each measurement, so that the original data can more comprehensively cover the characteristics of different formulations of nutritional foods under different measurement scenarios. This provides a rich and structured data foundation for subsequent data preprocessing, model training and validation, which helps the model to learn more fully the intrinsic relationship between nutritional food formulation, osmotic pressure and intestinal motility promotion rate, improve the model fitting accuracy, and thus improve the accuracy of low-osmotic nutritional food preparation process optimization and intestinal adaptability evaluation.

[0134] In one possible implementation, the linear regression analysis model is as follows:

[0135] Y=β0+β1X

[0136] Where Y represents the zebrafish intestinal motility promotion rate, X represents the measured osmotic pressure, β0 is a constant term, and β1 is the regression coefficient of osmotic pressure. The zebrafish intestinal motility promotion rate is expressed as a percentage (%), and the osmotic pressure is expressed as mOsm / kg.

[0137] In this embodiment, based on a linear regression analysis model, the corresponding intestinal motility promotion rate can be quickly calculated simply by measuring the osmotic pressure of the nutritional food, eliminating the need for complex experimental procedures and significantly shortening the evaluation cycle. Simultaneously, the simple model structure facilitates rapid application in the preparation of low-osmotic nutritional foods. Staff can promptly assess the intestinal adaptability of the nutritional food under current process parameters based on the model's prediction results, allowing for targeted adjustments to process parameters, effectively controlling the osmotic pressure of low-osmotic nutritional foods, reducing the risk of intestinal discomfort in infants and young children, and ensuring the stability and reliability of product quality.

[0138] In summary, as described in the above embodiments, training the linear regression analysis model includes three stages: determination of intestinal motility promotion rate, determination of osmotic pressure of reconstituted milk powder, and data modeling and analysis. The specific steps are as follows:

[0139] (1) Determination of intestinal peristalsis promotion rate

[0140] a) Zebrafish model construction

[0141] Fertilization and hatching: Female and male zebrafish are mated, and the fertilized eggs obtained after spawning are collected. Healthy fertilized eggs with complete morphology and normal development are selected under a microscope and transferred to a suitable hatching environment for cultivation.

[0142] Sample selection: Once the zebrafish have developed to the stage of swim bladder formation, healthy individuals are selected under a microscope as experimental samples;

[0143] Modeling process: Prepare a Nile Red working solution of a set concentration, place the screened zebrafish in it for treatment, and construct an intestinal fluorescent labeling model.

[0144] b) Sample intervention and data acquisition

[0145] Sample preparation: The hypotonic nutritional food samples to be evaluated were processed according to the set plan and then applied to the zebrafish model, with an appropriate treatment time set.

[0146] Image acquisition: After processing, each group of zebrafish was photographed under a fluorescence microscope to obtain intestinal fluorescence images.

[0147] c) Data Analysis

[0148] The collected images were analyzed using professional image processing software to measure the fluorescence intensity of zebrafish intestines and to calculate the intestinal motility promotion rate according to a preset formula.

[0149] (2) Osmotic pressure determination of milk powder reconstituted solution

[0150] a) Sample preparation: Mix milk powder and solvent according to the mixing ratio indicated on the nutritional food label to prepare a milk powder reconstituted milk sample;

[0151] b) Osmotic pressure measurement: The osmotic pressure of the above-mentioned reconstituted emulsion sample was measured using the freezing point method, and the measurement data were recorded.

[0152] (3) Establishment and application of regression analysis model

[0153] a) Data collection: Summarize the osmotic pressure data and intestinal motility promotion rate data of different nutritional food samples;

[0154] b) Model building: Based on the collected data, a mathematical model of osmotic pressure and intestinal motility promotion rate was established using regression analysis.

[0155] c) Model application: By measuring the osmotic pressure of the milk powder reconstituted solution, the corresponding intestinal motility promotion rate is predicted using a regression analysis model, thereby assessing the adaptability of nutritional foods to the intestines.

[0156] The above embodiments have introduced the training process of the linear regression analysis model from aspects such as sample data acquisition and data preprocessing. The following describes the training process of the linear regression analysis model with a specific embodiment to facilitate understanding of the complete model training process:

[0157] (1) Data acquisition and preprocessing

[0158] Sample set construction: Ten groups of nutritional food samples with different formulations (covering different protein ratios, calcium-phosphorus ratios, and process parameters) were prepared. Each sample was measured three times to obtain osmotic pressure (using the freezing point method) and zebrafish intestinal motility promotion rate (fluorescence intensity analysis method), for a total of 30 valid data sets.

[0159] Data cleaning: Outliers deviating from the mean ± 3 standard deviations were removed by Z-score standardization. After screening, all 30 sets of data met the requirements.

[0160] Dataset partitioning: Following the principles of random sampling and stratified balance, the dataset was divided into a training set (21 groups) and a test set (9 groups) in a 7:3 ratio to ensure that the osmotic pressure distribution range of the two sets of data is consistent.

[0161] (2) Initial model fitting

[0162] The initial linear regression model was fitted on the training set using the least squares method, and the preliminary coefficients β0=-9.98 and β1=0.125 were obtained.

[0163] Intestinal motility promotion rate (%) = -9.98 + 0.125 × osmotic pressure (mOsmol / kg).

[0164] (3) Model tuning and model validation

[0165] Cross-validation optimization:

[0166] Five-fold cross-validation was employed, with the training set evenly divided into five subsets based on osmotic pressure (covering low-to-intermediate-to-low-infiltration, intermediate-to-high-infiltration, high-infiltration, and the full range). Each round used four subsets for training and one subset for validation, with the parameters being tuned to minimize the root mean squared error (RMSE). The optimal coefficients were ultimately determined to be β0 = -9.18 and β1 = 0.1208, reducing the RMSE of cross-validation from 1.5 to 1.2.

[0167] Optimal equation: Peristalsis promotion rate (%) = -9.18 + 0.1208 × osmotic pressure (mOsmol / kg)

[0168] Model validation:

[0169] Training set validation: Substituting the 21 sets of osmotic pressure (X) from the training set into the optimal equation, the corresponding predicted values ​​of intestinal motility promotion rate (Ypredicted) were obtained. The prediction bias for each sample was calculated to evaluate the model's fitting accuracy to the training data. Training set validation results: average bias = 1.1%; maximum bias = 2.3%. Training set coefficient of determination R0 2 =0.83, indicating a good fit.

[0170] Test set validation: Substituting the 9 sets of osmotic pressure (X) from the test set into the optimal equation, the corresponding predicted values ​​of intestinal motility promotion rate (Ypredicted) were obtained. The prediction bias for each sample was calculated to evaluate the model's fitting accuracy to the test data. Test set validation results: average bias = 1.2%; maximum bias = 3.1%. Test set coefficient of determination R0 2 =0.81, indicating that the model has good generalization ability.

[0171] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0172] The above embodiments mainly introduce how to determine the optimized preparation process based on the predicted value and early warning conditions of the intestinal motility promotion rate, and how to adjust the process parameters according to the optimized preparation process. In actual implementation, the predicted value of the intestinal motility promotion rate is obtained based on a linear regression analysis model to achieve rapid evaluation of the intestinal adaptability of low-osmotic nutritional foods. It can also be used in scenarios such as quality control and consumption of low-osmotic nutritional foods. The specific implementation plan is as follows:

[0173] (1) Used for quality control and standardized supervision

[0174] Rapid detection standard establishment: The model is incorporated into the nutritional food factory testing process, and intestinal adaptability is directly predicted by measuring osmotic pressure (completed in about 10 minutes), replacing the traditional time-consuming zebrafish experiment (the traditional method requires more than 72 hours).

[0175] Tiered regulatory scheme: Risk levels are classified based on the predicted rate of intestinal motility promotion.

[0176] Low risk (≤25%): Labeled as "Gut-friendly";

[0177] Medium risk (26%~35%): Recommended for infants and young children over 3 months old;

[0178] High risk (>35%): The formula should be rectified.

[0179] (2) Used for personalized feeding guidance and consumer services

[0180] Age-appropriate recommendations: Based on model results and infant age (gut maturity), suggestions are provided for parents:

[0181] 0-6 months (when the intestines are most sensitive): Recommended nutritional foods with an osmotic pressure ≤280mOsm / kg and a promotion rate ≤25% (e.g., milk powder).

[0182] 7-12 months: Acceptable nutritional foods with an osmotic pressure of 280-300 mOsm / kg and a promotion rate of 26%-30% (e.g., milk powder).

[0183] Discomfort risk warning: By linking the product packaging QR code to the model database, consumers can obtain the osmotic pressure and predicted promotion rate of the batch of products (e.g., milk powder) after scanning. If the infant has a history of diarrhea, the system will automatically prompt the selection of products with low promotion rate.

[0184] In practical implementation, it is necessary not only to ensure the efficiency of the linear regression analysis model in evaluating the intestinal adaptability of nutritional foods, but also to ensure the accuracy of the evaluation. To verify that the linear regression analysis model has good predictive performance, the linear regression analysis model provided in the above embodiments was validated.

[0185] Example 1: Formulation Preparation and Physical Property Data

[0186] experimental group

[0187] Formula (per 100g milk powder): 0.6g hydrolyzed whey protein, 0.16g hydrolyzed casein, 9.8g non-hydrolyzed whey protein, 0.028g immunoglobulin (IgG), 318mg calcium (of which exogenous calcium: milk-derived calcium = 1:1.17), 200mg phosphorus, with the remaining ingredients supplemented by raw milk, lactose, compound minerals, compound vitamins, blended vegetable oil, choline chloride, potassium citrate, sodium citrate, arachidonic acid oil powder, docosahexaenoic acid oil powder, GOS, 2'-FL, nucleotides, etc.

[0188] Preparation process: After mixing, homogenizing, sterilizing and concentrating the formulation, the semi-finished product is prepared by spray drying. The homogenization pressure is 165 bar, the spray drying temperature is 172℃ and the exhaust temperature is 80℃. The semi-finished product and immunoglobulin (IgG) are physically mixed to ensure the preservation of the activity of immunoglobulin in the product.

[0189] Control group 1

[0190] Formula (per 100g milk powder): 1.0g hydrolyzed whey protein, 0.6g hydrolyzed casein, 8.96g non-hydrolyzed whey protein, 0.028g immunoglobulin (IgG), 318mg calcium (of which exogenous calcium: milk-derived calcium = 1:2), 200mg phosphorus, with the remaining ingredients supplemented by raw milk, lactose, compound minerals, compound vitamins, blended vegetable oil, choline chloride, potassium citrate, sodium citrate, arachidonic acid oil powder, docosahexaenoic acid oil powder, GOS, 2'-FL, nucleotides, etc.

[0191] Preparation process: After mixing, homogenizing, sterilizing and concentrating the formula, the semi-finished product is prepared by spray drying. The homogenization pressure is 165 bar, the spray drying temperature is 170℃ and the exhaust temperature is 78℃. The semi-finished product and immunoglobulin (IgG) are physically mixed to ensure the preservation of the activity of immunoglobulin in the product.

[0192] Control group 2

[0193] Formula (per 100g milk powder): 0.6g hydrolyzed whey protein, 0.16g hydrolyzed casein, 9.8g non-hydrolyzed whey protein, 0.028g immunoglobulin (IgG), 318mg calcium (of which exogenous calcium: milk-derived calcium = 1:1.17), 200mg phosphorus, with the remaining ingredients supplemented by raw milk, lactose, compound minerals, compound vitamins, blended vegetable oil, choline chloride, potassium citrate, sodium citrate, arachidonic acid oil powder, docosahexaenoic acid oil powder, GOS, 2'-FL, nucleotides, etc.

[0194] Preparation process: The formula is mixed, homogenized, sterilized and concentrated, and then spray-dried to prepare a semi-finished product. The homogenization pressure is 220 bar, the spray drying temperature is 200℃ and the exhaust temperature is 100℃. Immunoglobulin (IgG) is added in the wet process.

[0195] The specific results of the implementation are shown in the table below:

[0196]

[0197] Example 2: Verification of the effect of promoting intestinal peristalsis in zebrafish

[0198] The intestinal peristalsis promotion rate of each group was determined using a zebrafish intestinal peristalsis model:

[0199]

[0200] Figure 4 The following are zebrafish intestinal fluorescence imaging images corresponding to the blank group, experimental group, control group 1, and control group 2. Figure 4 As shown, the average intestinal fluorescence signal intensity of the experimental group was compared with that of control groups 1 and 2. The results showed that there were significant differences between the experimental group and the two control groups (experimental group: control group 1, p<0.05; experimental group: control group 2, p<0.01). The intestinal peristalsis promotion rate of the experimental group was lower than that of the two control groups, indicating that the milk powder of this experimental group is more adaptable to the intestine.

[0201] Relationship between the measured rate of intestinal motility promotion and the rate of intestinal motility promotion predicted based on osmotic pressure:

[0202] Substituting the osmotic pressure into the optimal equation: Intestinal motility promotion rate (%) = -9.18 + 0.1208 × osmotic pressure (mOsmol / kg), the prediction results are shown in the table below:

[0203]

[0204] The measured rate of intestinal motility promotion was close to that predicted based on osmotic pressure, with a difference of ≤2% between the two groups, indicating that the model has a good predictive effect.

[0205] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0206] Figure 5 A schematic diagram of the control device for preparing low-osmotic nutritional food according to an embodiment of the present invention is shown. For ease of explanation, only the parts related to the embodiment of the present invention are shown, and are described in detail below:

[0207] like Figure 5 As shown, the control device 5 for preparing hypotonic nutritional foods includes:

[0208] The acquisition module 501 is used to acquire the measured osmotic pressure of the mixture of materials to be prepared into a low-osmotic nutritional food; wherein, the measured osmotic pressure is obtained by measuring the osmotic pressure of the material sample of the mixture using the freezing point method;

[0209] Prediction module 502 is used to input the measured osmotic pressure into a trained linear regression analysis model to obtain the predicted value of the intestinal peristalsis promotion rate; wherein, the linear regression analysis model is trained based on the osmotic pressure of milk powder reconstituted solution and the intestinal peristalsis promotion rate of zebrafish.

[0210] The control module 503 is used to determine the preparation process optimization scheme based on the predicted value of intestinal peristalsis promotion rate and the early warning condition of intestinal peristalsis promotion rate, so as to adjust the process parameters according to the preparation process optimization scheme;

[0211] The process parameters in the optimized preparation process include one or more of the following: the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature; the types of proteins include hydrolyzed whey protein, hydrolyzed casein, and non-hydrolyzed whey protein.

[0212] In this embodiment, the osmotic pressure of the mixture of ingredients for the low-osmotic nutritional food to be prepared is accurately measured using the freezing point method. This osmotic pressure is then input into a linear regression analysis model trained on the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish, rapidly obtaining a predicted value for the intestinal motility promotion rate without relying on traditional time-consuming animal experiments or clinical observations. Subsequently, based on this predicted value and early warning conditions, an optimized preparation process scheme is determined. Key process parameters such as the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature are adjusted accordingly. This allows for synergistic control of the osmotic pressure of the low-osmotic nutritional food to be prepared from both the formulation and preparation process perspectives, effectively reducing the osmotic pressure of its reconstituted solution and minimizing the risk of intestinal irritation, diarrhea, or vomiting in infants caused by high osmotic pressure. Simultaneously, it enables dynamic control of the milk powder preparation process, ensuring that the final product has good intestinal adaptability.

[0213] Figure 6 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Figure 6As shown, the electronic device 6 of this embodiment includes a processor 60 and a memory 61. The memory 61 stores a computer program 62. When the processor 60 executes the computer program 62, it implements the steps in the various method embodiments described above. Alternatively, when the processor 60 executes the computer program 62, it implements the functions of each module / unit in the various device embodiments described above.

[0214] For example, computer program 62 may be divided into one or more modules / units, which are stored in memory 61 and executed by processor 60 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 62 in electronic device 6.

[0215] Electronic device 6 may include, but is not limited to, processor 60 and memory 61. Those skilled in the art will understand that... Figure 6 This is merely an example of electronic device 6 and does not constitute a limitation on electronic device 6. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 6 may also include input / output devices, network access devices, buses, etc.

[0216] The processor 60 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0217] The memory 61 can be an internal storage unit of the electronic device 6, such as a hard disk or RAM. The memory 61 can also be an external storage device of the electronic device 6, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 61 can include both internal and external storage units of the electronic device 6. The memory 61 is used to store the computer program 62 and other programs and data required by the electronic device 6. The memory 61 can also be used to temporarily store data that has been output or will be output.

[0218] For the sake of simplicity and clarity, only the above-described functional modules / units are used as examples. In practical applications, the functions described above can be assigned to different functional modules / units as needed. These modules / units can be implemented in hardware, software, or a combination of both.

[0219] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the methods described in the above-described method embodiments.

[0220] This invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the methods described in the above-described method embodiments.

[0221] Computer programs include computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. Computer-readable media can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0222] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0223] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for controlling the preparation of low-osmotic nutritional foods, characterized in that, include: The osmotic pressure of the mixture of materials to be prepared into a low-osmotic nutritional food is obtained; wherein the osmotic pressure is determined by measuring the osmotic pressure of the material sample of the mixture using the freezing point method; The measured osmotic pressure is input into a trained linear regression analysis model to obtain a predicted value of the intestinal motility promotion rate; wherein, the linear regression analysis model is trained based on the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish; Based on the predicted value of the intestinal peristalsis promotion rate and the early warning condition of the intestinal peristalsis promotion rate, an optimization scheme for the preparation process is determined, and the process parameters are adjusted according to the optimization scheme. The process parameters in the optimized preparation process include one or more of the following: the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature; the types of proteins include hydrolyzed whey protein, hydrolyzed casein, and non-hydrolyzed whey protein. The process includes, before inputting the measured osmotic pressure into the trained linear regression analysis model, the following: Prepare n groups of nutritional food samples with different formulations. Each group of samples is measured m times to obtain data on osmotic pressure and zebrafish intestinal peristalsis promotion rate, and obtain n×m groups of raw data. Preprocess the n×m sets of original data, and divide the preprocessed data into training set and test set; Based on the training set, the initial coefficients are obtained by fitting the initial linear regression analysis model using the least squares method. Five-fold cross-validation was used to optimize the parameters of the initial linear regression analysis model by minimizing the root mean square error, resulting in the optimized linear regression analysis model. The optimized linear regression analysis model was validated based on the training set and test set, and the mean deviation, maximum deviation and coefficient of determination were selected as evaluation indicators. When all evaluation indicators meet the discrimination criteria, a well-trained linear regression analysis model is obtained.

2. The control method for preparing low-osmotic nutritional food according to claim 1, characterized in that, The preprocessing of the n×m sets of original data includes: Outliers are removed using Z-score standardization.

3. The control method for preparing low-osmotic nutritional food according to claim 2, characterized in that, The outlier value is a value that deviates from the set range of the mean; wherein the set range is 3 times the standard deviation.

4. The control method for preparing low-osmotic nutritional food according to claim 1, characterized in that, The preparation of n groups of nutritional food samples with different formulations, with each group of samples measured m times, yielded data on osmotic pressure and zebrafish intestinal motility promotion rate, including: According to the mixing ratio indicated on the nutritional food label, mix milk powder and solvent to prepare n sets of nutritional food samples with different formulations. The osmotic pressure of n groups of nutritional food samples with different formulations was measured using the freezing point method, and the osmotic pressure data were recorded. Female and male zebrafish were mated to screen for zebrafish with swim bladder stage. A Nile Red working solution of a set concentration was prepared, and the screened zebrafish were treated with the Nile Red working solution to construct an intestinal fluorescent labeling model. The set concentration was 5 μg / L to 10 μg / L. Each group of samples was applied to the intestinal fluorescent labeling model and measured m times. After the application was completed, the zebrafish corresponding to each group of intestinal fluorescent labeling models were placed under a fluorescence microscope to take pictures and obtain intestinal fluorescence images. The intestinal peristalsis promotion rate of zebrafish was determined based on the intestinal fluorescence images.

5. The control method for preparing low-osmotic nutritional food according to claim 4, characterized in that, The process of obtaining n×m sets of original data includes: The osmotic pressure data and zebrafish intestinal motility promotion rate data corresponding to a single measurement were used as a set of data; Based on the data of each group of n different formula nutritional food samples, construct n×m sets of original data.

6. The method for controlling the preparation of low-osmotic nutritional food according to claim 1, characterized in that, The linear regression analysis model is as follows: Y=β0+β1X Where Y is the zebrafish intestinal peristalsis promotion rate, X is the measured osmotic pressure, β0 is a constant term, and β1 is the regression coefficient of osmotic pressure.

7. A control device for preparing low-osmotic nutritional foods using the control method for preparing low-osmotic nutritional foods according to any one of claims 1 to 6, characterized in that, include: The acquisition module is used to acquire the measured osmotic pressure of the mixture of materials to be prepared into a low-osmotic nutritional food; wherein the measured osmotic pressure is obtained by measuring the osmotic pressure of the material sample of the mixture using the freezing point method; The prediction module is used to input the measured osmotic pressure into a trained linear regression analysis model to obtain a predicted value of the intestinal motility promotion rate; wherein, the linear regression analysis model is trained based on the osmotic pressure of the milk powder reconstituted solution and the intestinal motility promotion rate of zebrafish. The control module is used to determine the preparation process optimization scheme based on the predicted value of the intestinal peristalsis promotion rate and the early warning condition of the intestinal peristalsis promotion rate, so as to adjust the process parameters according to the preparation process optimization scheme; The process parameters in the optimized preparation process include one or more of the following: the content range of different types of proteins, the ratio of exogenous calcium to milk-derived calcium, homogenization pressure, and spray drying temperature; the types of proteins include hydrolyzed whey protein, hydrolyzed casein, and non-hydrolyzed whey protein.

8. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images