Pressure adjusting method, device and equipment of homogenizer for dairy product preparation

Abstract

The invention provides a pressure adjusting method, device and equipment of a homogenizer for dairy product preparation, and relates to the technical field of dairy product processing and preparation. The method comprises the following steps: acquiring formula parameters of a dairy product to be produced; inputting the formula parameters into a pre-trained mixing-process parameter coupling model to obtain homogeneous parameters; wherein the homogenizing parameters comprise a plurality of pressures, and the mixing-process parameter coupling model is obtained by training according to a preset milk quality stability condition; and controlling the pressure of the homogenizer according to the plurality of pressures. The pressure of the homogenizer can be matched with formula parameters of the dairy product, so that the stability of the dairy product is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of dairy product processing and preparation, and particularly to a pressure regulation method, device and equipment for a homogenizer used in dairy product preparation. Background Art

[0002] In the field of dairy product processing and preparation, the homogenization process is an important process step, and its purpose is to disperse the fat globules in milk into a fine and evenly distributed state, thereby improving the stability of the product.

[0003] However, the prior art has the defect of fixed parameters. Most dairy product enterprises adopt fixed pressure parameters, such as the primary pressure / secondary pressure being equal to about 3, and fixing the primary pressure and secondary pressure at a relatively high level.

[0004] The inventors' research found that this practice leads to poor product stability, high energy consumption, and limited applicability. For example, for low-fat formulations (such as fat content ≤ 20%, based on total dry matter), if high pressure is fixedly used, it is easy to damage the milk protein interfacial film, resulting in fat globule reconstitution and aggregation (D50 > 0.7 μm). For high-fat or high-lactose formulations (such as fat > 20%, lactose ≥ 65%, based on total dry matter), conventional pressure cannot overcome the insufficient dispersion caused by high viscosity, resulting in a precipitation rate exceeding the 10% threshold. Summary of the Invention

[0005] Embodiments of the present invention provide a pressure regulation method, device and equipment for a homogenizer used in dairy product preparation to solve the problem of poor stability of dairy products caused by pressure mismatch.

[0006] In a first aspect, embodiments of the present invention provide a pressure regulation method for a homogenizer used in dairy product preparation, including:

[0007] Obtaining the formula parameters of the dairy product to be produced;

[0008] Inputting the formula parameters into a pre-trained mixing-process parameter coupling model to obtain homogenization parameters; wherein, the homogenization parameters include several pressures, and the mixing-process parameter coupling model is trained according to a preset milk quality stability condition;

[0009] Controlling the pressure of the homogenizer according to several pressures.

[0010] In a possible implementation manner, the preset milk quality stability condition includes:

[0011] The particle size D50 of the dairy product to be produced is less than or equal to a first threshold, and the Span value of the particle size distribution is less than or equal to a second threshold, and the precipitation rate is less than or equal to a third threshold.

[0012] In a possible implementation manner, the training process of the mixing-process parameter coupling model includes:

[0013] Obtain historical formulation parameters and historical homogenization parameters; among them, the historical homogenization parameters include several pressures.

[0014] Determine the corresponding historical response parameters according to each historical formulation parameter and historical homogenization parameter; among them, the historical response parameters include particle size, particle size distribution Span value, and precipitation rate.

[0015] Construct a historical data set according to each historical formulation parameter, the corresponding historical homogenization parameter, and the corresponding historical response parameter.

[0016] Train a mixture-process parameter coupling model according to the historical data set and the preset milk quality stability conditions.

[0017] In a possible implementation manner, the process of training a mixture-process parameter coupling model according to the historical data set and the preset milk quality stability conditions includes:

[0018] Use the historical formulation parameters and the corresponding historical homogenization parameters of the dairy product to be produced as process components, use the corresponding response parameters as response components, and perform step-by-step fitting using the preset milk quality stability conditions to obtain an initial mixture-process parameter coupling model.

[0019] Obtain a mixture-process parameter coupling model for pressure regulation of the homogenizer through residual analysis and rectification.

[0020] In a possible implementation manner, before performing residual analysis and rectification, the method further includes:

[0021] When the response component obtained through the initial mixture-process parameter coupling model does not meet the preset milk quality stability conditions, add data of new formulation parameters, the corresponding homogenization parameters, and the corresponding response parameters, and update the historical data set.

[0022] Retrain a mixture-process parameter coupling model according to the updated historical data set and the preset milk quality stability conditions.

[0023] In a possible implementation manner, determining the corresponding historical response parameters according to each historical formulation parameter and historical homogenization parameter includes:

[0024] Obtain the particle size D10, particle size D50, and particle size D90 of multiple dairy product samples to be produced; among them, each dairy product sample to be produced is prepared according to each historical formulation parameter and historical homogenization parameter, and the particle size D10, particle size D50, and particle size D90 are obtained by analyzing each dairy product sample to be produced with a Malvern particle size analyzer.

[0025] The Span values of the particle size distribution of each dairy product sample to be produced are calculated respectively according to the particle size D10, particle size D50, and particle size D90 of each dairy product sample to be produced;

[0026] Obtain the sediment dry weight of each dairy product sample to be produced; wherein, the sediment dry weight is obtained by centrifuging each dairy product sample to be produced with a centrifuge, collecting the sediment, and drying it;

[0027] The sedimentation rate of each dairy product sample to be produced is calculated according to the sediment dry weight and the total dry weight of the components of the formula of each dairy product sample to be produced.

[0028] In a possible implementation, the formula parameters include the ratios of protein, fat, and lactose; the several pressures include the primary pressure and the secondary pressure.

[0029] In a possible implementation, after controlling the pressure of the homogenizer according to the several pressures, the method further includes:

[0030] Obtain the particle size, Span value of the particle size distribution, and sedimentation rate of the dairy product to be produced;

[0031] When at least one of the particle size, Span value of the particle size distribution, and sedimentation rate does not meet the requirements, adjust the homogenization parameters and retrain the mixing-process parameter coupling model.

[0032] In a second aspect, an embodiment of the present invention provides a pressure regulating device for a homogenizer used in dairy product preparation, including:

[0033] A formula parameter acquisition module, configured to acquire the formula parameters of the dairy product to be produced;

[0034] A homogenization parameter determination module, configured to input the formula parameters into a pre-trained mixing-process parameter coupling model to obtain homogenization parameters; wherein, the homogenization parameters include several pressures, and the mixing-process parameter coupling model is trained according to preset milk quality stability conditions;

[0035] A pressure regulation module, configured to control the pressure of the homogenizer according to the several pressures

[0036] In a third aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the method in the first aspect above or any possible implementation manner of the first aspect.

[0037] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it implements the method in the first aspect above or any possible implementation manner of the first aspect.

[0038] In the embodiments of the present invention, by inputting the specific formula parameters of the dairy product to be produced, the homogenization parameters matching the formula can be accurately obtained. Among them, the homogenization parameters include several pressures, and the mixing - process parameter coupling model is trained according to the preset milk quality stability conditions. Control the pressure of the homogenizer according to the obtained accurate several pressures, and can accurately match the homogenization parameters according to different formulas, reduce the product quality fluctuations caused by unreasonable parameter settings, make different batches of dairy products highly consistent in terms of taste, texture, etc., and improve the overall quality of the product. Description of the Drawings

[0039] Figure 1 is the implementation flowchart of the pressure regulation method for the homogenizer used in the preparation of dairy products provided by the embodiments of the present invention;

[0040] Figure 2 is the logical block diagram of the pressure regulation method for the homogenizer used in the preparation of dairy products provided by the embodiments of the present invention;

[0041] Figure 3a is the fitting line graph of the predicted precipitation rate and the measured precipitation rate provided by the embodiments of the present invention;

[0042] Figure 3b is the fitting line graph of the predicted Span value and the measured Span value provided by the embodiments of the present invention;

[0043] Figure 4 is the structural schematic diagram of the pressure regulation device for the homogenizer used in the preparation of dairy products provided by the embodiments of the present invention;

[0044] Figure 5 is the schematic diagram of the electronic device provided by the embodiments of the present invention. Detailed Embodiments

[0045] The following will describe the embodiments of the present invention in detail with reference to the drawings.

[0046] See Figure 1 , which shows the implementation flowchart of the pressure regulation method for the homogenizer used in the preparation of dairy products provided by the embodiments of the present invention, and is described in detail as follows:

[0047] Step 101: Obtain the formula parameters of the dairy product to be produced.

[0048] In this embodiment, the formula parameters refer to the proportions or contents of various components in the dairy product. The formula parameters determine the characteristics and quality of the dairy product, such as taste, nutritional components, stability, and shelf life, etc. The formula parameters are usually determined by dairy product manufacturers or R & D departments according to product requirements and market positioning, and these parameters can come from laboratory tests, historical data, or industry standards.

[0049] Step 102: Input the formulation parameters into a pre-trained mixing-process parameter coupling model to obtain homogenization parameters; wherein, the homogenization parameters include a number of pressures, and the mixing-process parameter coupling model is trained according to a preset milk quality stability condition.

[0050] In this embodiment, the mixing-process parameter coupling model is a pre-trained model that can predict the optimal homogenization parameters based on the input formulation parameters. The model is trained according to preset milk quality stability conditions, which include quality indicators such as the particle size D50, particle size distribution Span value, and sedimentation rate of the product. The goal of the model is to ensure that these indicators meet the preset standards. In this embodiment, the specific constraint conditions are: D50 ≤ 0.5 μm, Span value IPD ≤ 0.8, and centrifugal sedimentation rate ≤ 10%.

[0051] Step 103: Control the pressure of the homogenizer according to a number of pressures.

[0052] In this embodiment, during the homogenization process of dairy products, according to a number of pressure values predicted by the model, the actual operating pressure of the homogenizer is adjusted to be consistent with the pressure recommended by the model. By precisely controlling the homogenization pressure, it is ensured that key indicators such as the particle size D50, particle size distribution Span value, and sedimentation rate of dairy products meet the preset quality standards. Product quality problems caused by improper pressure setting are avoided, such as too large fat globules, too high sedimentation rate, or too high energy consumption.

[0053] The present invention constructs a mixing-process parameter coupling model, predicts the optimal combination of a number of homogenization pressures through formulation components, and ensures that D50 ≤ 0.5 μm, Span value IPD ≤ 0.8, and sedimentation rate ≤ 10% for different formulations. For the first time, the mixing design and homogenization process variables are coupled through a composite model, covering the interaction effects of multiple components.

[0054] In the embodiment of the present invention, by inputting the specific formulation parameters of the dairy product to be produced, the homogenization parameters matching the formulation can be accurately obtained. Among them, the homogenization parameters include a number of pressures, and the mixing-process parameter coupling model is trained according to a preset milk quality stability condition. Control the pressure of the homogenizer according to the obtained accurate number of pressures, and can accurately match the homogenization parameters according to different formulations, reducing product quality fluctuations caused by unreasonable parameter settings, making dairy products of different batches highly consistent in terms of taste, texture, etc., and improving the overall quality of the product.

[0055] The above is an overall introduction to the solution of the present invention. Before using the model, it is necessary to train the model so that the model can accurately predict the homogenization parameters that meet the milk quality stability conditions. After the model training is completed, during the actual use process, due to reasons such as changes in formula ingredients, equipment differences, or environmental conditions changes, the homogenization parameters predicted by the model do not meet the requirements of actual production, and the model needs to be further adjusted. The following is an introduction to the dynamic correction of the model during the use process and the training process of the model before using the model.

[0056] In a possible implementation manner, after controlling the pressure of the pressure-controlled homogenizer according to a plurality of pressures, the method further includes:

[0057] Obtain the particle size, particle size distribution Span value, and precipitation rate of the dairy product to be produced;

[0058] When at least one of the particle size, particle size distribution Span value, and precipitation rate does not meet the requirements, adjust the homogenization parameters and retrain the mixture-process parameter coupling model.

[0059] In this embodiment, when using the mixture-process parameter coupling model, during the actual production process, after performing the homogenization process, evaluate whether the homogenization effect meets the quality standard through indicators such as particle size, particle size distribution Span value, and precipitation rate. If all indicators meet the preset milk quality stability conditions, it means that the current homogenization parameters are appropriate and the production process can continue. If at least one of the detected indicators does not meet the requirements, it means that the current homogenization parameters may need to adjust a plurality of pressures.

[0060] Then add the new formula parameters, adjusted homogenization parameters, and new test results to the historical dataset, and retrain the mixture-process parameter coupling model. The purpose of this step is to dynamically correct the model and improve the prediction accuracy of the model. Enable the model to better adapt to new formula and process conditions. Ensure that during future production processes, the model can provide more homogenization parameters that meet actual needs.

[0061] Figure 2 It is a logic block diagram of the pressure adjustment method for a homogenizer used for dairy product preparation provided by an embodiment of the present invention. It includes:

[0062] Ingredient input: Obtain the formula parameters of the dairy product to be produced according to step 101; the formula parameters refer to the proportions or contents of various ingredients in the dairy product, and the formula parameters determine the characteristics and quality of the dairy product, such as taste, nutritional components, stability, and shelf life, etc. The formula parameters are usually determined by dairy product manufacturers or R & D departments according to product requirements and market positioning, and these parameters can come from laboratory tests, historical data, or industry standards.

[0063] Model matching: Input the formulation parameters into the pre-trained mixing-process parameter coupling model according to Step 102 to obtain the homogenization parameters. Among them, the homogenization parameters include a number of pressures, and the mixing-process parameter coupling model is trained according to the preset milk quality stability conditions. The mixing-process parameter coupling model is a pre-trained model that can predict the optimal homogenization parameters based on the input formulation parameters. The model is trained according to the preset milk quality stability conditions, which include quality indicators such as the particle size D50, the particle size distribution Span value, and the sedimentation rate of the product. The goal of the model is to ensure that these indicators meet the preset standards.

[0064] Homogenization: Control the pressure of the homogenizer according to a number of pressures in Step 103. During the homogenization process of dairy products, according to the number of pressure values predicted by the model, adjust the actual operating pressure of the homogenizer to make it consistent with the pressure recommended by the model. By precisely controlling the homogenization pressure, ensure that key indicators such as the particle size D50, the particle size distribution Span value, and the sedimentation rate of dairy products meet the preset quality standards. Avoid product quality problems caused by improper pressure settings, such as too large fat globules, too high sedimentation rate, or too high energy consumption.

[0065] Detection indicators: Obtain the particle size D10, particle size D50, and particle size D90 of the dairy product sample to be produced. Among them, the dairy product sample to be produced is prepared according to the formulation parameters and homogenization parameters, and the particle size D10, particle size D50, and particle size D90 are obtained by analyzing the dairy product sample to be produced with a Malvern particle size analyzer; calculate the particle size distribution Span value of the dairy product sample to be produced based on the particle size D10, particle size D50, and particle size D90 of the dairy product sample to be produced; obtain the sediment dry weight of the dairy product sample to be produced. Among them, the sediment dry weight is obtained by centrifuging the dairy product sample to be produced with a centrifuge, collecting the sediment and drying it; calculate the sedimentation rate of the dairy product sample to be produced based on the sediment dry weight and the total dry weight of the components of the formulation of the dairy product sample to be produced. By preparing the dairy product sample, measuring the particle size, calculating the Span value, measuring the sedimentation rate and other steps, obtain D50, Span value, sedimentation rate. D50 refers to the volume median diameter of the fat globules in dairy products, that is, half of the fat globules have a volume smaller than this value and the other half is larger. D50 needs to be less than or equal to a set first threshold (for example, D50 ≤ 0.5 μm). A smaller D50 value means that the fat globules are finer, which helps to improve the stability and taste of the product. The Span value is used to describe the width of the particle size distribution, and the Span value needs to be less than or equal to a set second threshold (for example, Span value ≤ 0.8). A smaller Span value indicates a narrower particle size distribution, and a larger Span value indicates a wider particle size distribution. The sedimentation rate is used to evaluate the stability of dairy products. The sedimentation rate needs to be less than or equal to a set third threshold (for example, sedimentation rate ≤ 10%). A lower sedimentation rate indicates better product stability.

[0066] Data feedback and correction: When at least one of the particle size, the Span value of the particle size distribution, and the precipitation rate does not meet the requirements (i.e., exceeds the standard), adjust the homogenization parameters and retrain the coupled model of the mixing-process parameter. When using the coupled model of the mixing-process parameter, during the actual production process, after performing the homogenization process, evaluate whether the homogenization effect meets the quality standard through indicators such as the particle size, the Span value of the particle size distribution, and the precipitation rate. If all indicators meet the preset milk quality stability conditions (i.e., do not exceed the standard), it means that the current homogenization parameters are appropriate and the production process can continue. If at least one of the detected indicators does not meet the requirements, it means that the current homogenization parameters may need to be adjusted by a certain pressure.

[0067] In different embodiments, the setting of the preset milk quality stability conditions varies according to the actual product and requirements.

[0068] In one possible implementation, the preset milk quality stability conditions include:

[0069] The D50 of the dairy product to be produced is less than or equal to the first threshold, the Span value of the particle size distribution is less than or equal to the second threshold, and the precipitation rate is less than or equal to the third threshold.

[0070] In this embodiment, D50 refers to the volume median diameter of the fat globules in the dairy product, that is, half of the fat globules have a volume smaller than this value and the other half is larger. D50 needs to be less than or equal to a set first threshold (for example, D50 ≤ 0.5 μm). A smaller D50 value means finer fat globules, which helps to improve the stability and taste of the product. The Span value is an indicator measuring the width of the fat globule particle size distribution, and the calculation formula is Span value = (D90 - D10) / D50, where D10 is the value at which 10% of the fat globules have a volume smaller than this value, D10 represents the lower limit of the smaller particles, and reflects the proportion of small particles in the sample. If D10 is smaller, it means there are more small particles in the emulsion, which may have a positive impact on the stability of the product. D90 is the value at which 90% of the fat globules have a volume smaller than this value. D90 represents the upper limit of the larger particles and reflects the proportion of large particles in the sample. If D90 is larger, it means there are large particles in the emulsion, which may affect the taste and stability of the product. The Span value needs to be less than or equal to a set second threshold (for example, Span value ≤ 0.8). A lower Span value indicates a more uniform particle size distribution, which helps to improve the stability and consistency of the product. The precipitation rate refers to the proportion of the solid components precipitating after the dairy product is centrifuged or left standing. The precipitation rate needs to be less than or equal to a set third threshold (for example, precipitation rate ≤ 10%). A lower precipitation rate indicates a more uniform distribution of the solid components in the product, which helps to improve the stability and shelf life of the product.

[0071] In other possible implementation manners, one or a combination of multiple ones of particle size D50, particle size distribution Span value, precipitation rate, particle size D10, particle size D90, fat globule uniformity, protein stability, or lactose crystallinity are selected according to different products and requirements of dairy products to determine preset milk quality stability conditions.

[0072] In one possible implementation manner, the training process of the mixing-process parameter coupling model includes:

[0073] Obtain historical formula parameters and historical homogenization parameters; wherein, the historical homogenization parameters include a plurality of pressures;

[0074] Determine corresponding historical response parameters according to each historical formula parameter and historical homogenization parameter; wherein, the historical response parameters include particle size, particle size distribution Span value, and precipitation rate;

[0075] Construct a historical data set according to each historical formula parameter, the corresponding historical homogenization parameter, and the corresponding historical response parameter;

[0076] Train to obtain a mixing-process parameter coupling model according to the historical data set and preset milk quality stability conditions.

[0077] In this embodiment, the historical formula parameters refer to collecting past dairy product formula data. The historical homogenization parameters refer to collecting the homogenization parameters used in the past production of these dairy products, including a plurality of pressures. For each combination of historical formula and homogenization parameters, determine the corresponding historical response parameters, and these parameters include: particle size, which measures the size of fat globules; particle size distribution Span value, which measures the uniformity of fat globule particle size; precipitation rate, which measures the precipitation of solid components in dairy products. Integrate the historical formula parameters, historical homogenization parameters, and the corresponding historical response parameters together to form a historical data set for model training.

[0078] Exemplarily, the response surface method is a statistical modeling method that approximates complex implicit relationships through polynomial functions and is used to optimize and analyze the input-output relationships of multivariable systems. The regression method is a statistical analysis method that predicts the relationships between variables by establishing a mathematical model, including types such as linear regression and nonlinear regression. Use the response surface method or the regression method to analyze the historical data set and establish a model that can predict the relationship between homogenization parameters (a plurality of pressures) and response parameters (particle size, Span value, and precipitation rate). During the training process, the model will be optimized according to preset quality standards (such as D50≤0.5μm, Span value≤0.8, precipitation rate≤10%) to ensure that the model can predict homogenization parameters that meet these standards.

[0079] In one possible implementation manner, the process of training to obtain a mixing-process parameter coupling model according to the historical data set and preset milk quality stability conditions includes:

[0080] Taking each historical formula parameter and the corresponding historical homogenization parameter as process components, taking the corresponding historical response parameter as the response component, and using the preset milk quality stability condition for step-by-step fitting to obtain an initial mixture-process parameter coupling model;

[0081] Through residual analysis and correction, a mixture-process parameter coupling model for pressure regulation of the homogenizer is obtained.

[0082] In this embodiment, through step-by-step fitting and residual analysis, it is ensured that the model can accurately predict the key indicators (such as particle size, Span value, and precipitation rate) in the dairy product homogenization process, thereby optimizing the pressure parameters of the homogenizer. This process combines historical data and statistical analysis methods to ensure the reliability and practicality of the model. Using the response surface method or regression method, taking the formula parameters and homogenization parameters as input variables (process components), and taking the response parameters (such as D50, Span value, precipitation rate) as output variables (response components), a mathematical relationship between the input variables and output variables is established. This process continuously adjusts the model parameters to make the model better fit the historical data. During the model fitting process, the preset milk quality stability conditions (such as D50 ≤ 0.5 μm, Span value ≤ 0.8, precipitation rate ≤ 10%) are considered to ensure that the model output meets these conditions. After step-by-step fitting, an initial mixture-process parameter coupling model is obtained.

[0083] Calculate the difference (residual) between the model predicted value (response parameter) and the actual value, and analyze the distribution of these residuals. Residual analysis can help identify the deficiencies of the model, such as whether there are systematic biases or some variables are not fully considered. According to the results of the residual analysis, the model is adjusted and optimized. Such as adjusting the structure of the model (such as adding or deleting certain variables); correcting outliers or noise in the data; refitting the model to improve the prediction accuracy. The model after correction can more accurately predict the response parameters (such as D50, Span value, precipitation rate), thereby providing reliable guidance for the pressure regulation of the homogenizer.

[0084] Based on the sufficient sample data in the historical dataset, it is possible to train based on the pre-constructed historical dataset, taking the formula parameters and homogenization parameters as process components and the response parameters as response components. After residual analysis and correction, a trained mixture-process parameter coupling model is obtained. In some possible embodiments, when the amount of data in the historical dataset is small, it is necessary to expand the sample data in the historical dataset to ensure the prediction accuracy of the mixture-process parameter coupling model.

[0085] In a possible implementation manner, before residual analysis and correction, the method includes:

[0086] When the response components obtained through the initial mixture - process parameter coupling model do not meet the preset milk quality stability conditions, add the data of new formulation parameters, corresponding homogenization parameters, and corresponding response parameters, and update the historical data set.

[0087] According to the updated historical data set and the preset milk quality stability conditions, retrain the mixture - process parameter coupling model.

[0088] In this embodiment, the mixture - process parameter coupling model is optimized by updating the historical data set. When the response components (such as particle size D50, particle size distribution Span value, sedimentation rate, etc.) obtained through the mixture - process parameter coupling model do not meet the preset milk quality stability conditions (such as D50 ≤ 0.5μm, Span value ≤ 0.8, sedimentation rate ≤ 10%), it indicates that the prediction result of the current model does not meet the actual requirements, and the data set is considered insufficient. In this case, it is necessary to add the data of new formulation parameters, corresponding homogenization parameters, and corresponding response parameters. These new data can come from new experiments or new data in actual production. Add these new data to the historical data set to form an updated data set. Using the updated historical data set and the preset milk quality stability conditions, use the response surface method or regression method for model training again. Through retraining, the model can learn new data patterns, thereby improving the accuracy and reliability of the prediction. The retrained model needs to be verified to ensure that its prediction results meet the preset milk quality stability conditions. By continuously adding new data and retraining, the model can better adapt to different formulation and process conditions, improving the prediction accuracy. This dynamic update mechanism enables the model to continuously optimize with the change of production conditions, maintaining its applicability and effectiveness.

[0089] Optimize the model from the two aspects of the prediction accuracy and data volume of the model itself, so that the homogenization parameters obtained according to the model also meet the process condition requirements (that is, meet the preset milk quality stability conditions) when accurately matching with the formulation parameters.

[0090] In a possible implementation manner, determining the corresponding historical response parameters according to each historical formulation parameter and historical homogenization parameter includes:

[0091] Obtain the particle size D10, particle size D50, and particle size D90 of multiple dairy product samples to be produced; wherein, each dairy product sample to be produced is prepared according to each historical formulation parameter and historical homogenization parameter, and the particle size D10, particle size D50, and particle size D90 are obtained by analyzing each dairy product sample to be produced with a Malvern particle size analyzer;

[0092] Calculate the particle size distribution Span value of each dairy product sample to be produced respectively according to the particle size D10, particle size D50, and particle size D90 of each dairy product sample to be produced;

[0093] Obtain the sediment dry weight of each dairy product sample to be produced; wherein, the sediment dry weight is obtained by centrifuging each dairy product sample to be produced with a centrifuge, collecting the sediment and drying it.

[0094] Calculate the sedimentation rate of each dairy product sample to be produced based on the sediment dry weight and the total dry weight of the components in the formula of each dairy product sample to be produced.

[0095] In this embodiment, through steps such as preparing dairy product samples, measuring particle size, calculating the Span value, and measuring the sedimentation rate, historical response parameters (such as D50, Span value, sedimentation rate) are obtained. These response parameters will be used to construct and train a mixture-process parameter coupling model to optimize the dairy product homogenization process. Multiple dairy product samples to be produced are prepared according to these parameters, similar to producing dairy products according to established formulas and process parameters. The Span value is used to describe the width of the particle size distribution. A smaller Span value indicates a narrower particle size distribution, and a larger Span value indicates a wider particle size distribution. The sedimentation rate is used to evaluate the stability of dairy products. A lower sedimentation rate indicates better product stability. The measurement of the sedimentation rate includes: placing the dairy product sample in a centrifuge and centrifuging it at a speed of 4000 revolutions per minute for 10 minutes. After centrifugation, collect the sediment and perform a drying treatment. Weigh the dried sediment to obtain the sediment dry weight. Then calculate the sedimentation rate according to the formula.

[0096] In a possible implementation, the formula parameters include the ratios of protein, fat, and lactose; several pressures include the primary pressure and the secondary pressure.

[0097] In this embodiment, the percentages of fat (Z1), protein (Z2), and lactose (Z3) in the formula parameters (Z3 = 100 - Z1 - Z2). The homogenizer usually has two pressure stages, namely the primary pressure and the secondary pressure, which are used to control the pressure exerted by the homogenizer during the production process to ensure the quality of dairy products. The primary pressure is used for preliminary homogenization to break the fat globules in the milk into smaller particles. The secondary pressure is used for further refining the fat globules to ensure the uniform distribution and stability of the particles. These two pressure parameters are key variables in the homogenization process. According to the process requirements, the primary pressure and the secondary pressure can be adjusted. When the particle size is too large (D50 > 0.5 μm), increasing the primary pressure (X1) can more effectively disperse the fat globules, reduce the particle size, and increasing the secondary pressure (X2) can further refine the fat globules and improve the particle size distribution. When the particle size distribution is uneven (Span value > 0.8), optimize the ratio of the primary pressure and the secondary pressure to make it more coordinated, thereby improving the particle size distribution. When the sedimentation rate is too high (sedimentation rate > 10%), too high a pressure may cause excessive fragmentation of the fat globules and increase the sedimentation rate. Appropriately reducing the primary pressure can reduce sedimentation. Appropriately adjust the secondary pressure to ensure the uniform distribution of the fat globules and reduce sedimentation.

[0098] In other possible implementation manners, according to different production requirements, such as high-fat or high-lactose formulations, multi-component complex formulations, particle size and distribution control, etc., several pressures including a first pressure, a second pressure, a third pressure, etc. are set to meet the usage requirements of the above different working conditions. For example, for high-fat or high-lactose formulations, multi-stage pressures are required to fully disperse and refine the components. For multi-component complex formulations, multi-stage pressures are required to gradually disperse and refine different components. For particle size and distribution control, in order to ensure the uniformity of particle size and distribution, multi-stage pressures are required to gradually optimize.

[0099] The specific implementation manners of the solution are introduced above. In the actual process, experimental analysis is carried out based on specific embodiments to verify and illustrate the effects of the above solution.

[0100] Experimental design: A mixture design (protein 8.6%-19%, fat 16.8%-30%, lactose 51%-70%) is nested with homogenization process parameters (first stage 100-200 bar, second stage 10-60 bar) to generate 11 groups of basic formulations × 5 process combinations (a total of 55 data points).

[0101] Data measurement: The particle size and precipitation rate of each formulation × process combination are tested.

[0102] Among them, for particle size measurement, a Malvern particle size analyzer is used to obtain data of D10 (μm), D50 (μm) and D90 (μm);

[0103] For the particle size distribution Span value, IDP = (D90 - D10) ÷ D50.

[0104] A centrifuge is used at a speed of 4000 r for 10 min to collect the precipitate and dry it.

[0105] Precipitation rate (%) = precipitate dry weight (g) ÷ total dry weight of formulation (g) × 100%.

[0106] Model construction: Using protein, fat, and lactose as process components, and the first-stage pressure and the second-stage pressure as process components, D50, IDP, and precipitation rate are used as responses for stepwise fitting respectively. After inputting the formulation parameters, call the database that couples mixture design (D-optimal) and response surface, or use the regression method to match the optimal pressure parameters (X1, X2). After residual correction, the adjusted R2 values of the model are 96.6%, 92.07%, and 97.2% respectively. Accurate prediction can be achieved and recommended pressures can be given through the model.

[0107] Model verification data: Table 1 shows the model verification data.

[0108] Table 1 Model verification data

[0109]

[0110] Figure 3a It is the fitting line graph of the predicted precipitation rate and the measured precipitation rate provided by the embodiment of the present invention. The predicted precipitation rate = 0.3309 + 0.9682 × the measured precipitation rate. Among them, the scatter points are the ratio of the predicted precipitation rate to the measured precipitation rate, and the straight line is obtained by fitting these scatter points. In the obtained linear relationship, the value of 0.9682 is quite close to 1, indicating that the predicted precipitation rate and the measured precipitation rate are very close. Moreover, most of the scatter points are on the fitting straight line, indicating that the fitting effect is good. Therefore, the predicted precipitation rate under the homogenization parameters predicted by the model of the present invention is very close to the actual precipitation rate, indicating that the model of the present invention is very accurate. Figure 3b It is the fitting line graph of the predicted Span value and the measured Span value provided by the embodiment of the present invention. The predicted Span value = 0.04556 + 0.9677 × the measured Span value. Among them, the scatter points are the ratio of the predicted Span value to the measured Span value, and the straight line is obtained by fitting these scatter points. In the obtained linear relationship, the value of 0.9677 is quite close to 1, indicating that the predicted Span value and the measured Span value are very close. Moreover, most of the scatter points are on the fitting straight line, indicating that the fitting effect is good. Therefore, the predicted Span value under the homogenization parameters predicted by the model of the present invention is very close to the actual precipitation rate, indicating that the model of the present invention is very accurate.

[0111] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0112] The following is the device embodiment of the present invention. For the details not described in detail, reference can be made to the corresponding method embodiment above.

[0113] Figure 4 The structural schematic diagram of the pressure regulating device of the homogenizer for dairy product preparation provided by the embodiment of the present invention is shown. For the convenience of description, only the parts related to the embodiment of the present invention are shown and are described in detail as follows:

[0114] As Figure 4 shown, the pressure regulating device 4 of the homogenizer for dairy product preparation includes:

[0115] A formula parameter acquisition module 41, configured to acquire the formula parameters of the dairy product to be produced;

[0116] A homogenization parameter determination module 42, configured to input the formula parameters into a pre-trained mixing-process parameter coupling model to obtain homogenization parameters; wherein, the homogenization parameters include a plurality of pressures, and the mixing-process parameter coupling model is trained according to a preset milk quality stability condition;

[0117] A pressure regulation module 43 for regulating the pressure of a homogenizer according to a plurality of pressures.

[0118] In a possible implementation, the homogenization parameter determination module 42 is further configured to:

[0119] Obtain historical formula parameters and historical homogenization parameters; wherein, the historical homogenization parameters include a plurality of pressures;

[0120] Determine corresponding historical response parameters according to each historical formula parameter and historical homogenization parameter; wherein, the historical response parameters include particle size, particle size distribution Span value, and sedimentation rate;

[0121] Construct a historical data set according to each historical formula parameter, the corresponding historical homogenization parameter, and the corresponding historical response parameter;

[0122] Train a mixture-process parameter coupling model according to the historical data set and a preset milk quality stability condition.

[0123] In a possible implementation, the homogenization parameter determination module 42 is further configured to:

[0124] The process of training a mixture-process parameter coupling model according to the historical data set and a preset milk quality stability condition includes:

[0125] Using each historical formula parameter and the corresponding historical homogenization parameter as process components, using the corresponding historical response parameter as a response component, and performing step-by-step fitting using the preset milk quality stability condition to obtain an initial mixture-process parameter coupling model;

[0126] Obtain a mixture-process parameter coupling model for pressure regulation of the homogenizer through residual analysis and correction.

[0127] In a possible implementation, the homogenization parameter determination module 42 is further configured to:

[0128] When the response component obtained through the initial mixture-process parameter coupling model does not reach the preset milk quality stability condition, add data of new formula parameters, the corresponding homogenization parameters, and the corresponding response parameters, and update the historical data set;

[0129] Retrain a mixture-process parameter coupling model according to the updated historical data set and the preset milk quality stability condition.

[0130] In a possible implementation, the homogenization parameter determination module 42 is further configured to:

[0131] Obtain the particle sizes D10, D50, and D90 of multiple dairy product samples to be produced; wherein, each dairy product sample to be produced is prepared according to each historical formulation parameter and historical homogenization parameter, and the particle sizes D10, D50, and D90 are obtained by analyzing each dairy product sample to be produced with a Malvern particle size analyzer;

[0132] Calculate the particle size distribution Span value of each dairy product sample to be produced respectively according to the particle sizes D10, D50, and D90 of each dairy product sample to be produced;

[0133] Obtain the sediment dry weight of each dairy product sample to be produced; wherein, the sediment dry weight is obtained by centrifuging each dairy product sample to be produced with a centrifuge, collecting the sediment and drying it;

[0134] Calculate the sedimentation rate of each dairy product sample to be produced according to the sediment dry weight and the total dry weight of the components of the formulation of each dairy product sample to be produced.

[0135] In the embodiments of the present invention, by inputting the specific formulation parameters of the dairy product to be produced, the homogenization parameters matching the formulation can be accurately obtained. Among them, the homogenization parameters include several pressures, and the mixing - process parameter coupling model is trained according to the preset milk quality stability conditions. Control the pressure of the homogenizer according to the obtained accurate several pressures, and can accurately match the homogenization parameters according to different formulations, reduce the product quality fluctuations caused by unreasonable parameter settings, make different batches of dairy products highly consistent in terms of taste, texture, etc., and improve the overall quality of the product.

[0136] Figure 5 It is a schematic diagram of the electronic device provided by the embodiments of the present invention. As Figure 5 shown, the electronic device 5 of this embodiment includes: a processor 50 and a memory 51. The memory 51 stores a computer program 52. When the processor 50 executes the computer program 52, the steps in the above - mentioned various method embodiments are implemented. Or, when the processor 50 executes the computer program 52, the functions of each module / unit in the above - mentioned various device embodiments are implemented.

[0137] Exemplarily, the computer program 52 can be divided into one or more modules / units, and the one or more modules / units are stored in the memory 51 and executed by the processor 50 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of completing specific functions, and these instruction segments are used to describe the execution process of the computer program 52 in the electronic device 5.

[0138] The electronic device 5 may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art can understand, Figure 5This is only an example of the electronic device 5, which does not constitute a limitation on the electronic device 5. It may include more or fewer components than those shown in the figure, or combine certain components, or different components. For example, the electronic device 5 may also include input / output devices, network access devices, buses, etc.

[0139] The processor 50 may be a central processing unit (CPU), or may also be 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 may be a microprocessor or the processor may also be any conventional processor, etc.

[0140] The memory 51 may be an internal storage unit of the electronic device 5, such as a hard disk or memory of the electronic device 5. The memory 51 may also be an external storage device of the electronic device 5, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the electronic device 5. Further, the memory 51 may also include both the internal storage unit and the external storage device of the electronic device 5. The memory 51 is used to store the computer program 52 and other programs and data required by the electronic device 5. The memory 51 may also be used to temporarily store the data that has been output or will be output.

[0141] For the convenience and simplicity of description, only the above division of each functional module / unit is used as an example. In actual applications, the above functions may be assigned to different functional modules / units according to needs. The above modules / units may be implemented in the form of hardware, or in the form of software, or in the form of a combination of hardware and software.

[0142] The embodiment of the present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the methods in the above method embodiments are implemented.

[0143] The embodiment of the present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the methods in the above method embodiments are implemented.

[0144] Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0145] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments. Without special instructions and logical conflicts, the terms and / or descriptions among different embodiments are consistent and can be referenced mutually. The technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0146] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A pressure regulation method for a homogenizer used in dairy product preparation, characterized in that, Including: Obtain the formulation parameters of the dairy product to be produced; Input the formulation parameters into a pre-trained mixing-process parameter coupling model to obtain homogenization parameters; wherein, the homogenization parameters include a number of pressures, and the mixing-process parameter coupling model is trained according to preset milk quality stability conditions; Control the pressure of the homogenizer according to the number of pressures.

2. The pressure regulation method of the homogenizer for dairy product preparation according to claim 1, characterized in that, The preset milk quality stability conditions include: The particle size D50 of the dairy product to be produced is less than or equal to a first threshold, the particle size distribution Span value is less than or equal to a second threshold, and the sedimentation rate is less than or equal to a third threshold.

3. The pressure regulation method of the homogenizer for dairy product preparation according to claim 2, characterized in that, The training process of the mixing-process parameter coupling model includes: Obtain historical formulation parameters and historical homogenization parameters; wherein, the historical homogenization parameters include a number of pressures; Determine the corresponding historical response parameters according to each historical formulation parameter and historical homogenization parameter; wherein, the historical response parameters include particle size, particle size distribution Span value, and sedimentation rate; Construct a historical data set according to each historical formulation parameter, the corresponding historical homogenization parameter, and the corresponding historical response parameter; Train a mixing-process parameter coupling model according to the historical data set and the preset milk quality stability conditions.

4. The pressure adjustment method of the homogenizer for dairy product preparation according to claim 3, characterized in that, The process of training a mixing-process parameter coupling model according to the historical data set and the preset milk quality stability conditions includes: Use each historical formulation parameter and the corresponding historical homogenization parameter as process components, use the corresponding historical response parameter as the response component, and perform step-by-step fitting using the preset milk quality stability conditions to obtain an initial mixing-process parameter coupling model; Obtain a mixing-process parameter coupling model for pressure regulation of the homogenizer through residual analysis and correction.

5. The pressure regulation method of the homogenizer for dairy product preparation according to claim 4, characterized in that, Before performing residual analysis and correction, the method further includes: When the response component obtained through the initial mixing-process parameter coupling model does not meet the preset milk quality stability conditions, add data of new formulation parameters, the corresponding homogenization parameters, and the corresponding response parameters, and update the historical data set; Retrain a mixing-process parameter coupling model according to the updated historical data set and the preset milk quality stability conditions.

6. The pressure adjustment method of the homogenizer for dairy product preparation according to claim 3, characterized in that, Determining the corresponding historical response parameter according to each historical formulation parameter and historical homogenization parameter includes: Obtain the particle size D10, particle size D50, and particle size D90 of multiple dairy product samples to be produced; wherein, each dairy product sample to be produced is prepared according to each historical formulation parameter and historical homogenization parameter, and the particle size D10, particle size D50, and particle size D90 are obtained by analyzing each dairy product sample to be produced with a Malvern particle size analyzer; Calculate the particle size distribution Span value of each dairy product sample to be produced respectively according to the particle size D10, particle size D50, and particle size D90 of each dairy product sample to be produced; Obtain the sediment dry weight of each dairy product sample to be produced; wherein, the sediment dry weight is obtained by centrifuging each dairy product sample to be produced with a centrifuge, collecting the sediment, and drying; Calculate the sedimentation rate of each dairy product sample to be produced according to the sediment dry weight and the total dry weight of the components of the formulation of each dairy product sample to be produced.

7. The pressure regulation method of the homogenizer for dairy product preparation according to claim 1, characterized in that, The formulation parameters include the ratios of protein, fat, and lactose; the number of pressures includes a primary pressure and a secondary pressure.

8. A pressure regulating device for a homogenizer used in dairy product preparation, characterized in that, Including: A formula parameter acquisition module, configured to acquire the formula parameters of the dairy product to be produced; A homogenization parameter determination module, configured to input the formula parameters into a pre-trained mixture-process parameter coupling model to obtain homogenization parameters; wherein, the homogenization parameters include a plurality of pressures, and the mixture-process parameter coupling model is trained according to a preset milk quality stability condition; A pressure adjustment module, configured to control the pressure of a homogenizer according to the plurality of pressures.

9. An electronic device, characterized in that, Including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images