Optimization treatment method and device for dairy product heat treatment process

Abstract

The invention provides a dairy product heat treatment process optimization treatment method and device. The method comprises the following steps: combining different process parameters according to interval numerical values corresponding to a plurality of predetermined variable factors associated with a dairy product heat treatment processing process to obtain an initial process parameter combination; performing predictive analysis on the initial process parameter combination based on a preset heat treatment model to obtain influence weights of a plurality of variable factors output by the heat treatment model on each product index parameter of the dairy product; and determining a corresponding target process parameter combination according to the influence weights of the plurality of variable factors on each product index parameter of the dairy product and the initial process parameter combination. According to the optimization treatment method for the heat treatment process of the dairy products, the technological parameter combination in the heat treatment process of the dairy products can be rapidly predicted, the optimization treatment efficiency and accuracy of the heat treatment process of the dairy products are effectively improved, and therefore the number of times of optimization implementation tests can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of heat treatment of dairy products, and in particular to a method and device for optimizing the heat treatment process of dairy products. In addition, the present invention also relates to an electronic device and a processor-readable storage medium. Background Art

[0002] Dairy products are gradually becoming the main leisure food for consumers of different age groups due to their rich nutritional value. The core content of the indirect heat treatment process of dairy products usually includes two parts. One is to achieve precise control of the temperature of dairy products in the heat treatment tubular sterilizer (i.e., dairy heat processor) by heat exchange with the heat exchange medium. The other is to achieve the inactivation of harmful microorganisms such as pathogenic bacteria during the process of dairy products undergoing different temperature changes, so as to ensure product safety and shelf life stability, while maximizing the nutritional value of the product. The entire dairy heat treatment process involves multi-variable and multi-index optimization problems. The existing implementation process generally involves data research and analysis through a large number of implementations. However, due to the limitations of implementation costs and variable flexibility adjustment, it is difficult to accurately find the intrinsic relationship between multiple variable factors and product index parameters, and the implementation efficiency and accuracy are poor. Therefore, how to design a more efficient dairy heat treatment process optimization treatment scheme has become an important issue that needs to be solved by technicians in this field. Summary of the invention

[0003] To this end, the present invention provides a method for optimizing the heat treatment process of dairy products to solve the problem that the optimization scheme of the heat treatment process of dairy products in the prior art has high limitations, resulting in poor optimization efficiency and accuracy of the heat treatment process of dairy products.

[0004] In a first aspect, the present invention provides a method for optimizing a heat treatment process of a dairy product, comprising:

[0005] Combining different process parameters according to interval values ​​corresponding to a plurality of variable factors associated with the dairy product heat treatment process, to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the dairy product heat treatment process;

[0006] Based on a preset heat treatment model, the initial process parameter combination is predicted and analyzed to obtain the influence weights of multiple variable factors output by the heat treatment model on various product index parameters of dairy products; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and actual measured values;

[0007] According to the weights of the influence of the multiple variable factors on the index parameters of each dairy product and the initial process parameter combination, the corresponding target process parameter combination is determined.

[0008] Furthermore, the heat treatment model includes a heat transfer model and a reaction kinetics model; the heat transfer model is obtained by training based on the first sample data, the predicted value of the milk temperature during the heat treatment process of the milk corresponding to the first sample data, and the actual measured value of the milk temperature; the reaction kinetics model is obtained by training based on the second sample data, the predicted value of the product index parameter during the heat treatment process of the milk corresponding to the second sample data, and the actual measured value of the product index parameter;

[0009] The heat treatment model is used to predict the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products; the reaction kinetics model is used to predict the concentration change rate of the reaction products corresponding to the index parameters of each dairy product under different process parameter combinations of multiple variable factors during the heat treatment process of dairy products; wherein the multiple variable factors include the dairy product reaction temperature variable, and the dairy product reaction temperature variable includes the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline; the concentration change rate of the reaction product corresponds to the reaction rate of the reaction product;

[0010] The method of predicting and analyzing the initial process parameter combination based on the preset heat treatment model to obtain the influence weights of multiple variable factors output by the heat treatment model on various product index parameters of dairy products includes:

[0011] Inputting the target values ​​of multiple variable factors corresponding to the initial process parameter combination into the heat treatment model to predict the heat treatment temperature value, and obtaining the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products;

[0012] The temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process are input into the reaction kinetics model to predict the product index concentration, and the concentration change rate of the reaction product corresponding to each dairy product index parameter under different process parameter combinations of multiple variable factors during the dairy product heat treatment process is obtained; the concentration change rate is the speed at which the concentration of the reaction product changes over time;

[0013] Based on the concentration change rate of the reaction product, the influence weights of multiple variable factors on each product index parameter of the dairy product are determined; wherein the concentration change rate and the influence weight are in direct proportion.

[0014] Furthermore, before predicting and analyzing the initial process parameter combination based on the preset heat treatment model to obtain the influence weights of the multiple variable factors output by the heat treatment model on the various product index parameters of the dairy product, it also includes:

[0015] Construct a heat transfer model to predict the temperature of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process; and,

[0016] A reaction kinetic model is constructed to predict the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors during the heat treatment process of dairy products.

[0017] Furthermore, the heat transfer model for predicting the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products is constructed, specifically including:

[0018] Obtaining pre-collected first sample data; wherein the first sample data includes component content data of the dairy product, temperature data of the heating medium side in the dairy product heat processor, and flow data of the heating medium side;

[0019] Obtaining initial values ​​of input parameters of the heat transfer model to be determined, and obtaining a corresponding initial heat transfer model based on the initial values ​​of the parameters of the heat transfer model to be determined;

[0020] The initial heat transfer model is iteratively trained based on the first sample data to obtain a heat transfer model that meets preset conditions; wherein different heat transfer models correspond to different stages in the process of heat treating the dairy product using the dairy product heat processor.

[0021] Furthermore, the reaction kinetic model constructed for predicting the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors in the dairy product heat treatment process specifically includes:

[0022] Obtaining collected second sample data; wherein the second sample data includes previously predicted temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process;

[0023] Obtaining initial values ​​of input parameters of the reaction kinetic model to be determined, and obtaining a corresponding initial reaction kinetic model based on the initial values ​​of the parameters of the reaction kinetic model to be determined;

[0024] Iteratively training the initial heat transfer model based on the second sample data to obtain a reaction kinetics model that meets preset conditions;

[0025] Wherein, different stages in the process of heat treating the dairy product by using the dairy product heat processor correspond to different reaction kinetic models.

[0026] Furthermore, after determining the influence weights of multiple variable factors on each dairy product index parameter based on the concentration change rate of the reaction product, it also includes: determining the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter based on each dairy product temperature value at the pipeline inlet and outlet nodes during the dairy product heat treatment process; determining the corresponding target process parameter combination according to the size of the influence weights of the multiple variable factors on each dairy product heat treatment processing intensity index parameter and the initial process parameter combination.

[0027] Furthermore, after determining the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter, it also includes:

[0028] A comprehensive comparison and analysis is performed based on the influence weights of the multiple variable factors on each dairy heat treatment processing intensity index parameter and the influence weights of the multiple variable factors on each dairy product index parameter, so as to determine the corresponding final process parameter combination from the initial process parameter combination.

[0029] In a second aspect, the present invention further provides a device for optimizing a heat treatment process of dairy products, comprising:

[0030] An initial process parameter determination unit is used to combine different process parameters according to interval values ​​corresponding to a plurality of variable factors associated with the dairy product heat treatment process, so as to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the dairy product heat treatment process;

[0031] A heat treatment influence weight prediction unit is used to predict and analyze the initial process parameter combination based on a preset heat treatment model, and obtain the influence weights of multiple variable factors output by the heat treatment model on each product index parameter of the dairy product; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and actual measured values;

[0032] The target process parameter determination unit is used to determine the corresponding target process parameter combination according to the influence weights of the multiple variable factors on each dairy product index parameter and the initial process parameter combination.

[0033] Furthermore, the heat treatment model includes a heat transfer model and a reaction kinetics model; the heat transfer model is obtained by training based on the first sample data, the predicted value of the milk temperature during the heat treatment process of the milk corresponding to the first sample data, and the actual measured value of the milk temperature; the reaction kinetics model is obtained by training based on the second sample data, the predicted value of the product index parameter during the heat treatment process of the milk corresponding to the second sample data, and the actual measured value of the product index parameter;

[0034] The heat treatment model is used to predict the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products; the reaction kinetics model is used to predict the concentration change rate of the reaction products corresponding to the index parameters of each dairy product under different process parameter combinations of multiple variable factors during the heat treatment process of dairy products; wherein the multiple variable factors include the dairy product reaction temperature variable, and the dairy product reaction temperature variable includes the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline; the concentration change rate of the reaction product corresponds to the reaction rate of the reaction product;

[0035] The heat treatment influence weight prediction unit is specifically used for:

[0036] Inputting the target values ​​of multiple variable factors corresponding to the initial process parameter combination into the heat treatment model to predict the heat treatment temperature value, and obtaining the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products;

[0037] The temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process are input into the reaction kinetics model to predict the product index concentration, and the concentration change rate of the reaction product corresponding to each dairy product index parameter under different process parameter combinations of multiple variable factors during the dairy product heat treatment process is obtained; the concentration change rate is the speed at which the concentration of the reaction product changes over time;

[0038] Based on the concentration change rate of the reaction product, the influence weights of multiple variable factors on each product index parameter of the dairy product are determined; wherein the concentration change rate and the influence weight are in direct proportion.

[0039] Furthermore, before predicting and analyzing the initial process parameter combination based on the preset heat treatment model to obtain the influence weights of the multiple variable factors output by the heat treatment model on the various product index parameters of the dairy product, it also includes: a model building unit for:

[0040] Construct a heat transfer model to predict the temperature of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process; and,

[0041] A reaction kinetic model is constructed to predict the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors during the heat treatment process of dairy products.

[0042] Furthermore, the heat transfer model for predicting the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products is constructed, specifically including:

[0043] Obtaining pre-collected first sample data; wherein the first sample data includes component content data of the dairy product, temperature data of the heating medium side in the dairy product heat processor, and flow data of the heating medium side;

[0044] Obtaining initial values ​​of input parameters of the heat transfer model to be determined, and obtaining a corresponding initial heat transfer model based on the initial values ​​of the parameters of the heat transfer model to be determined;

[0045] The initial heat transfer model is iteratively trained based on the first sample data to obtain a heat transfer model that meets preset conditions; wherein different heat transfer models correspond to different stages in the process of heat treating the dairy product using the dairy product heat processor.

[0046] Furthermore, the reaction kinetic model constructed for predicting the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors in the dairy product heat treatment process specifically includes:

[0047] Acquire the collected second sample data; wherein the second sample data includes the previously predicted temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process;

[0048] Obtaining initial values ​​of input parameters of the reaction kinetic model to be determined, and obtaining a corresponding initial reaction kinetic model based on the initial values ​​of the parameters of the reaction kinetic model to be determined;

[0049] Iteratively training the initial heat transfer model based on the second sample data to obtain a reaction kinetics model that meets preset conditions;

[0050] Wherein, different stages in the process of heat treating the dairy product by using the dairy product heat processor correspond to different reaction kinetic models.

[0051] Furthermore, after determining the influence weights of multiple variable factors on each dairy product index parameter based on the concentration change rate of the reaction product, it also includes: a second target process parameter determination unit, which is used to determine the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter based on each dairy product temperature value at the pipeline inlet and outlet nodes during the dairy product heat treatment process; determine the corresponding target process parameter combination according to the size of the influence weights of the multiple variable factors on each dairy product heat treatment processing intensity index parameter and the initial process parameter combination.

[0052] Furthermore, after determining the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter, it also includes:

[0053] The third target process parameter determination unit is used for:

[0054] A comprehensive comparison and analysis is performed based on the influence weights of the multiple variable factors on each dairy heat treatment processing intensity index parameter and the influence weights of the multiple variable factors on each dairy product index parameter, so as to determine the corresponding final process parameter combination from the initial process parameter combination.

[0055] In a third aspect, the present invention further provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method for optimizing the heat treatment process of dairy products as described in any one of the above items are implemented.

[0056] In a fourth aspect, the present invention further provides a processor-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for optimizing the heat treatment process of dairy products as described in any one of the above items are implemented.

[0057] The method for optimizing the heat treatment process of dairy products provided by the present invention combines different process parameters by using interval values ​​corresponding to a plurality of variable factors associated with the heat treatment process of dairy products in advance, obtains an initial process parameter combination, performs a prediction analysis on the initial process parameter combination based on a preset heat treatment model, obtains the influence weights of the plurality of variable factors output by the heat treatment model on the index parameters of each dairy product, and determines the corresponding target process parameter combination according to the size of the influence weights of the plurality of variable factors on the index parameters of each dairy product and the initial process parameter combination. The method can quickly predict the process parameter combination in the heat treatment process of dairy products, effectively improves the optimization processing efficiency and accuracy of the heat treatment process of dairy products, and thus helps to reduce the number of optimization implementation tests. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0059] Figure 1 A schematic diagram of a process flow of a method for optimizing a heat treatment process of dairy products provided by an embodiment of the present invention;

[0060] Figure 2 A schematic diagram of a traditional dairy product temperature rise and fall curve provided by an embodiment of the present invention;

[0061] Figure 3A schematic diagram of a temperature rise and fall curve of dairy products calculated by a model provided in an embodiment of the present invention;

[0062] Figure 4 A schematic diagram for comparing the actual lactulose value and the model predicted value provided by an embodiment of the present invention;

[0063] Figure 5 A schematic diagram of a lactulose concentration prediction calculation curve provided by an embodiment of the present invention;

[0064] Figure 6 A schematic diagram of the structure of a device for optimizing the heat treatment process of dairy products provided in an embodiment of the present invention;

[0065] Figure 7 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar users, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0068] Heat treatment tubular sterilizers are generally divided into three stages of heating and three stages of cooling. The production site usually only arranges thermocouples at the inlet and outlet of the six heat exchange sections for temperature monitoring, which makes it difficult to truly describe the real-time temperature gradient changes inside the pipeline. This will lead to large calculation errors in the description of product indicators such as the sterilization intensity F0 value, B value, and C value. At the same time, the actual chemical reaction of heat-sensitive substances inside the pipeline is difficult to accurately calculate, which brings certain difficulties to product process optimization. In addition to F0 value, B value and C value, the product indicator parameters for evaluating the intensity of heat treatment processing of dairy products may also include degradation, denaturation and inactivation of heat-sensitive components (such as whey protein and enzymes) and products generated during heat treatment (such as lactulose, furosine, hydroxymethylfurfural) and other product indicator parameters. In the entire dairy heat treatment process, there are many variable factors that affect these product indicator parameters. At present, there is a lack of an effective method to quantitatively analyze these heat treatment influencing factors to obtain the influence weights of different variable factors on the product indicator parameters of dairy heat treatment intensity. As mentioned above, the factors that affect the heat treatment process of dairy products include dairy components (fat, protein, lactose, and total solids, etc.), dairy flow, water-side temperature of the heat exchange medium, flow, and pipeline layout, pipe diameter, pipe length, metal pipe material, etc. At the same time, for the six-stage heat exchange section and sterilization section of the sterilizer, the complexity of these influencing variable factors is further superimposed, making it difficult for R&D technicians to conduct multi-variable and multi-index parameter influence law research, and also increasing the difficulty of implementing the design. Only one or two key variables can be selected based on experience for single factor analysis and implementation testing, lacking systematic and scientific research.

[0069] The present invention can associate variables of all influencing factors affecting heat treatment intensity according to the equipment operation principle and mechanism formula, so that the model can perform sensitivity analysis and calculation according to the interval range values ​​of these influencing factors during use, thereby assisting R&D and design personnel to find the key influencing factors of multi-variable and multi-objective optimization problems more quickly and improve R&D efficiency.

[0070] Before using the model for calculation, the sterilizer pipeline is firstly processed into a one-dimensional grid with the help of software tools, and then the temperature of the dairy product is solved and calculated according to the model of the present invention on each differential pipeline unit, so as to obtain the data value of the actual temperature gradient change of the dairy product along the flow direction of the pipeline. That is, the present invention can associate all the variable factors affecting the heat treatment process, and with the help of the powerful solution and calculation capabilities of the software tools, perform a comprehensive impact analysis and calculation of multiple variables and multiple targets, and obtain the data distribution of the influence of different variables on different calculation target values. At the same time, the product index parameters can be predicted and calculated based on the known process parameters, so as to assist R&D technicians to conduct scientific analysis faster and better, significantly reduce the number of tests, and improve R&D efficiency.

[0071] The following is a detailed description of an embodiment of the method for optimizing the heat treatment process of dairy products according to the present invention. Figure 1 As shown, it is a schematic diagram of the process flow of the method for optimizing the heat treatment process of dairy products provided by an embodiment of the present invention, and the specific implementation process includes the following steps:

[0072] Step 101: combining different process parameters according to predetermined interval values ​​corresponding to a plurality of variable factors associated with the dairy product heat treatment process to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the dairy product heat treatment process.

[0073] Specifically, according to the interval values ​​corresponding to the multiple variable factors associated with the heat treatment process of dairy products, different values ​​can be selected from them to combine different process parameters to obtain the initial process parameter combination. For example: the numerical interval of the water temperature at the inlet of the second heating section is 95℃-155℃; the numerical interval of the water temperature at the inlet of the first heating section is 80℃-100℃; the numerical interval of the length of the sterilization section includes 3.52m, 8.84m, 13.26m, 22.1m; the numerical interval of the length of the protein holding section includes 26.5m, 53.1m, 79.6m, 83.8m. Among them, the water temperature and the length of the tube are different multiple variable factors, and the reaction time variables corresponding to different tube lengths are also different. One set of initial process parameter combinations can be: the water temperature at the inlet of the first heating section is 80℃, the water temperature at the inlet of the second heating section is 95℃, the length of the sterilization section is 3.52m, the length of the protein holding section is 26.5m, etc. The initial process parameter combination may include multiple sets of data, which are not specifically limited here.

[0074] Step 102: Predictively analyze the initial process parameter combination based on a preset heat treatment model to obtain the influence weights of multiple variable factors output by the heat treatment model on various product index parameters of dairy products; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and actual measured values.

[0075] Specifically, the heat treatment model includes a heat transfer model and a reaction kinetics model; the heat transfer model is obtained by training based on the first sample data, the predicted value of the temperature of the dairy product during the heat treatment process of the dairy product corresponding to the first sample data, and the actual measured value of the temperature of the dairy product; the reaction kinetics model is obtained by training based on the second sample data, the predicted value of the product index parameter during the heat treatment process of the dairy product corresponding to the second sample data, and the actual measured value of the product index parameter. The heat treatment model is used to predict the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of the dairy product; the reaction kinetics model is used to predict the concentration change rate of the reaction product corresponding to each product index parameter of the dairy product under different process parameter combinations under multiple variable factors during the heat treatment process of the dairy product; wherein the multiple variable factors include the dairy reaction temperature variable, and the dairy reaction temperature variable includes the various dairy temperature values ​​at the inlet and outlet nodes of the pipeline; the concentration change rate of the reaction product corresponds to the reaction rate of the reaction product.

[0076] In an embodiment of the present invention, the initial process parameter combination is predicted and analyzed based on a preset heat treatment model to obtain the influence weights of multiple variable factors output by the heat treatment model on each product index parameter of the dairy product. The corresponding specific implementation process may include: inputting the target values ​​of the multiple variable factors corresponding to the initial process parameter combination into the heat treatment model to predict the heat treatment temperature value, obtaining each dairy temperature value at the inlet and outlet nodes of the pipeline during the heat treatment process of the dairy product, inputting each dairy temperature value at the inlet and outlet nodes of the pipeline during the heat treatment process of the dairy product into the reaction kinetics model to predict the product index concentration, and obtaining the concentration change rate of the reaction products corresponding to each product index parameter of the dairy product under different process parameter combinations of the multiple variable factors during the heat treatment process of the dairy product; the concentration change rate is the speed at which the concentration of the reaction product changes with time; based on the concentration change rate of the reaction product, determining the influence weights of the multiple variable factors on each product index parameter of the dairy product; wherein the concentration change rate is proportional to the influence weight.

[0077] It should be noted that, in the embodiment of the present invention, before executing this step, it is necessary to construct a heat transfer model for predicting the temperature values ​​of various dairy products at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products; and to construct a reaction kinetic model for predicting the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations of multiple variable factors during the heat treatment process of dairy products.

[0078] Among them, the heat transfer model constructed for predicting the temperature values ​​of various dairy products at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products, and the corresponding specific implementation process include: obtaining pre-collected first sample data; wherein the first sample data includes the component content data of the dairy product, the temperature data on the heating medium side in the dairy heat processor, and the flow data on the heating medium side; obtaining the initial assignment of the input parameters of the heat transfer model to be determined, and obtaining the corresponding initial heat transfer model based on the initial assignment of the parameters of the heat transfer model to be determined; iteratively training the initial heat transfer model based on the first sample data to obtain a heat transfer model that meets preset conditions; wherein different heat transfer models correspond to different stages in the heat treatment process of the dairy product using the dairy heat processor.

[0079] Specifically, obtaining the first sample data collected in advance includes collecting the first sample data and predicting and calculating the physical property parameters of the milk. The physical property parameters of the milk include the thermal conductivity of the milk, the kinematic viscosity of the milk, and the dynamic viscosity of the milk, wherein the dynamic viscosity of the milk is divided by the density of the milk ρ m,T The kinematic viscosity of dairy products can be obtained. Four parts of data need to be collected in the early stage of heat transfer model building:

[0080] The first part of the data collection is the physical properties of different components of dairy products (i.e., the component content data of dairy products), including the molar mass (g / mol), density (kg / m3), thermal conductivity (W / (mk)) of fat, protein, lactose and water in dairy products, as well as the specific volume (m3 / kg) of whey protein, casein and fat required for calculating dairy viscosity. These data are mainly derived from test values ​​and existing data, and are not specifically limited here. The physical properties of dairy products depend on the material temperature and component content, and the density of dairy products is calculated using the following formula:

[0081]

[0082] In the formula, f c is a correction coefficient, which can be corrected by implementing data in the heat transfer model of the present invention; ρ i,T is the density value of each component of the dairy product at 20°C, i represents the different components of the dairy product, for example, 10 components, w i is the density of water (i.e., heat medium), which is a function of temperature. The physical parameters of thermal conductivity and viscosity of dairy products (including kinematic viscosity of dairy products and dynamic viscosity of dairy products) are calculated using a method similar to the above component mixing method. The specific heat capacity of dairy products varies slightly within the heat treatment temperature range, and a preset constant value can be used. It should be noted that, when the component content data of the above-mentioned dairy products are known, the data required to construct the heat transfer model, such as thermal conductivity of dairy products, kinematic viscosity of dairy products, dynamic viscosity of dairy products, etc., can be calculated using the corresponding mature algorithms of the prior art, which will not be described in detail here.

[0083] The second part of the data collection is the process parameters of the heat treatment process of dairy products. During the implementation process, the heat treatment process of the sterilizer is collected, including: preheating section, first heating section, second heating section, sterilization section, first cooling section, second cooling section and third cooling section. The water side flow (L / h), temperature (℃) data, and dairy side temperature (℃) data. The inlet and outlet dairy temperatures of the seven heat exchange sections are numbered in sequence: T1, T2, T3, T4, T5, T6, T7, T8. Except for the inlet dairy temperature T1 of the sterilizer as a known value, the remaining 7 dairy temperature values ​​are used as the target dairy temperature values ​​for heat transfer model prediction and calculation. Since the internal structure of each unit pipeline of the 6 heat exchange sections of the sterilizer is the same, that is, the heat transfer correlation equation of each heat exchange pipeline has the same form. At the same time, for each batch of feed, it is also necessary to collect concentration data of different components. Among them, T1 is the temperature of the dairy product at the entrance of the preheating section (known), T2 is the temperature of the dairy product at the exit of the first heating section, T3 is the temperature of the dairy product at the exit of the second heating section, T4 is the temperature of the dairy product at the exit of the third heating stage, T5 is the temperature of the dairy product at the exit of the sterilization stage, T6 is the temperature of the dairy product at the exit of the first cooling stage, T7 is the temperature of the dairy product at the exit of the second cooling stage, and T8 is the temperature of the dairy product at the exit of the third cooling stage.

[0084] The third part of the data collection is the content of dairy components and heat-sensitive components detected at the import and export of the sterilizer, including the concentration data values ​​of fat, protein, lactose, ash, lactulose, furosine, α-lactalbumin and β-lactoglobulin.

[0085] The fourth part of the collected data is the structural dimension data of the sterilizer equipment, as shown in Table 1 below, including the number of tube groups in each heat exchange section of the shell-and-tube heat exchanger, the outer tube diameter-tube length, the inner tube diameter-tube length, the holding section diameter-tube length, and the metal tube material and thermal conductivity (W / (mk)).

[0086]

[0087]

[0088] Table 1

[0089] In the process of building the heat transfer model, the convective heat transfer coefficient h on the dairy side of the tube m The Nusselt number correlation is used for calculation:

[0090]

[0091] Reynolds number

[0092] Prandtl number

[0093] In the formula, u is the flow rate of milk in the pipe (m / s), D is the pipe diameter (m), υ is the kinematic viscosity of milk (m2 / s), c p is the specific heat capacity of dairy products (J / (kg.K)), η is the dynamic viscosity of dairy products (Pa.s), λ is the thermal conductivity of dairy products (W / (mk)), a, b, c are the parameters of the heat transfer model to be determined in the heat transfer model, and in the implementation process of the initial heat transfer model of the present invention, a, b, and c are initially assigned values, a=0.02, b=0.8, and c=0.3. Secondly, the component content data and the temperature-flow data of the sterilizer water side (i.e., the temperature data of the heating medium side and the flow data of the heating medium side in the dairy heat treatment device) are input into the initial heat transfer model, and the heat transfer model can realize the prediction and calculation of the dairy temperature value of each heat exchange section (respectively, the preheating section, the first heating section, the second heating section, the sterilization section, the first cooling section, the second cooling section, and the third cooling section), wherein the dairy physical property parameter values ​​required for the dairy heat transfer calculation are calculated by the component mixing formula in the prior art, which will not be described in detail here. At the same time, the entire heat treatment model also needs to accurately characterize the internal pipeline structure of the heat treatment shell-and-tube sterilizer so that the heat transfer model can accurately associate the pipeline heat exchange area and the flow direction of the dairy product during the calculation process. During the heat transfer model verification process, the predicted values ​​of the dairy temperature values ​​T1, T2, T3, T4, T5, T6, T7, and T8 calculated by the heat transfer model under the initial a, b, and c values ​​are generally larger than the actual measured values ​​of the dairy temperature. Therefore, it is necessary to use an iterative calculation method to continuously adjust the a, b, and c values ​​until the calculation results meet the F-test (i.e., the preset conditions of the joint hypothesis test). Since the internal structures of the six heat exchange unit pipelines of the sterilizer (i.e., the heat treatment shell-and-tube sterilizer) are the same, that is, the heat transfer correlation equation of each heat exchange pipeline has the same form, the heat transfer model parameters a, b, and c are corrected through the implementation data of the three-stage heating section to obtain a determined heat transfer model, which is applied to the dairy temperature prediction calculation of the entire sterilizer pipeline. It should be noted that a, b, and c obtained by modeling at different stages are also different, that is, there may be multiple heat transfer models in a dairy heat treatment process to predict the temperature values ​​of dairy products at different stages (such as the seven stages mentioned above).

[0094] The specific implementation process is as follows: first, the multi-variable factor multi-batch implementation is designed with the dairy component content, dairy flow rate, and different temperatures and flows on the water side of each heat exchange section of the sterilizer as multiple variable factors, and the feed component content, feed flow rate, temperature data of the dairy side at the inlet and outlet of the heating section, water side temperature, and flow data of the sterilizer are recorded for each batch, see Appendix Table 2:

[0095]

[0096] Table 2: Heat transfer model verification implementation data table

[0097] Using batch implementation data as input parameters, an iterative calculation method is used to continuously adjust the a, b, and c values ​​to perform dairy temperature prediction calculations, and finally the model prediction values ​​(i.e., dairy temperature prediction values) corresponding to the batch dairy temperature measured values ​​(T2, T3, T4, T5, T6, T7, and T8) are obtained. An F-test is performed on the two sets of data, the measured values ​​and the model prediction values. At a significance level of α = 0.05, the calculated F value is less than the F table value, indicating that there is no significant difference between the model prediction value and the measured value, that is, the heat transfer model satisfies the engineering calculation. The measured values ​​corresponding to the heat transfer model are compared with the dairy temperature prediction values. Specifically, Figure 2 and 3 The heat treatment model of the present invention not only realizes the calculation of the real temperature rise and fall curve that is more consistent with the temperature gradient change of the sterilizer milk, but also compares the predicted value of the milk temperature at the inlet and outlet nodes with the measured value of the milk temperature, and its calculation accuracy can reach more than 98%. It should be noted that, by using the algorithm corresponding to the heat transfer model (i.e., the Nusselt number correlation formula), the convective heat transfer coefficient h on the milk side of the tube is calculated. m Then, based on the known heat transfer coefficient of water and the solid thermal conductivity of the metal pipe wall, combined with the convection heat transfer coefficient on the dairy side of the pipe and the algorithm of the existing technology, the temperature value of the dairy product can be predicted and calculated.

[0098] Specifically, the core content of the indirect heat treatment process of dairy products includes two parts: the change of heat exchange temperature of dairy products and the physical and chemical reactions carried out at different temperatures. The heat exchange part mainly realizes the temperature change on the dairy side by dynamically adjusting the temperature and flow of the heat exchange medium (the medium side is generally water, and its temperature adjustment is achieved by steam-water heat exchange). Generally, dairy products and heat exchange medium are heat exchanged through 6 heat exchange sections, namely the preheating section, the first heating section, the second heating section, the sterilization section, the first cooling section, the second cooling section and the third cooling section. Since the temperature range of dairy products is different in each heat exchange section, the intensity of thermochemical reactions of dairy products in different heat exchange sections is also different. Thermal processing of dairy products can effectively kill various pathogenic microorganisms. As the heat treatment temperature of dairy products increases, a series of physical and chemical reactions will occur in dairy products, such as whey protein denaturation and coagulation, lactose isomerization degradation and Maillard reaction. As these reactions proceed, the reduction of active ingredients in dairy products (such as alkaline phosphatase, whey protein) or the increase of reaction products (such as lactulose, furosine) are used as product indicator parameters of thermal processing intensity. Inside the shell-and-tube heat exchanger, the heat exchange process between dairy products and the medium (heat medium) is shown in the following formula. Taking the heating process of dairy products as an example, the heat is first transferred from the high-temperature medium side to the outside of the metal tube, then transferred to the inside of the metal tube by solid heat conduction, and finally transferred to the dairy side by convection heat transfer to achieve precise control of the dairy temperature. The algorithm of the existing technology used is as follows:

[0099]

[0100] Where, T 介质 is the medium temperature on the shell side of the pipeline, which can be detected; T 乳品 is the temperature of the milk on the pipeline side, which can be detected; T 管外 is the outer wall temperature of the metal pipe, which can be detected; T 管内 is the inner wall temperature of the metal pipe, which can be detected; T 乳品 is the final predicted temperature value of the dairy product, q is the transferred heat, such as the heat transferred from water to the pipe wall or from the pipe wall to the dairy product, and heat loss is not considered temporarily, that is, the heat transferred between the two is equal; R1 is the convection heat transfer resistance on the pipeline heat exchange medium (such as water) side (that is, the known heat transfer coefficient corresponding to water, which is known information and can be used directly), R2 is the thermal conductivity resistance of the metal pipe wall (that is, the known solid thermal conductivity of the metal pipe wall, which is known information and can be used directly), R3 is the convection heat transfer resistance on the dairy side (that is, the convection heat transfer coefficient h corresponding to the dairy side in the pipe). m ), see the following formula for details.

[0101] Since water has been maturely studied as a conventional fluid in traditional industries, there are very complete calculation formulas for its physical parameters and heat transfer calculations under different heat exchange forms, which can be directly applied in the invented model. However, the convective heat transfer thermal resistance on the dairy side is affected by the dairy components and non-Newtonian fluid characteristics, and further research is needed.

[0102] The calculation formula of the convective heat transfer resistance R3 on the dairy side is:

[0103] In the formula, h m A is the convective heat transfer coefficient on the dairy side, which is mainly determined by the physical properties of dairy products, equipment structure, material flow rate and operating pressure; i is the heat transfer area in the tube, which can be calculated by knowing the tube diameter and tube length, which can be directly input. Therefore, the dimensionless Nusselt number is introduced to indirectly calculate the heat transfer coefficient:

[0104]

[0105] Where λ is the thermal conductivity of the dairy side. This expression characterizes the dimensionless temperature gradient of the fluid on the wall, which is mainly related to the Reynolds number characterizing the flow state and the Prandtl number characterizing the flow boundary layer and the thermal boundary layer. The calculation formula is as follows:

[0106]

[0107] Where: Re and Pr are Reynolds number and Prandtl number respectively, a, b, c are unknown parameters to be determined.

[0108] In summary, the main factors affecting the heat transfer calculation of dairy products include dairy density, viscosity, specific heat capacity, thermal conductivity, pipeline layout, pipe diameter, pipe length, etc.

[0109] The above analytical models are usually used for macro-design or verification calculation of equipment, but fail to consider the temperature gradient change of materials inside the equipment, making it difficult to accurately characterize the actual distribution of dairy temperature inside the equipment. Some scholars have used CFD computational fluid dynamics to perform three-dimensional modeling calculations on sterilizer equipment. Although the dairy temperature can be accurately calculated, the modeling cycle is long and the cost is high. At the same time, the model is difficult to flexibly perform comprehensive impact analysis and calculations on multiple variables, which is not convenient for engineering calculations.

[0110] With the development of industrial digitalization and the improvement of computer computing power, the traditional industrial field has greatly improved production efficiency through digital models. The dairy processing technology involves multiple disciplines, and most of the problems are related to multi-variable and multi-objective optimization. The present invention proposes a heat treatment model of a data + mechanism coupling model, and uses sample data to perform parameter estimation and F-test on the model, comprehensively considering the complexity and accuracy of the model to ensure the reliability of the model. At the same time, the model of the present invention uses software tools to perform one-dimensional grid processing on the heat exchange pipeline of the tube sterilizer, and accurately captures the gradient change of the dairy temperature inside the pipeline by solving a large number of partial differential equations. By associating the complex heat exchange and chemical reaction inside the sterilizer with the product indicators, it can perform multi-variable and multi-objective optimization analysis, which can significantly reduce the number of implementations and achieve scientific decision-making analysis.

[0111] In addition, the reaction kinetic model constructed for predicting the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors in the dairy product heat treatment process, the corresponding specific implementation process includes: obtaining the collected second sample data; wherein the second sample data includes the previously predicted various dairy temperature values ​​at the pipeline inlet and outlet nodes in the dairy product heat treatment process; obtaining the initial assignment of the input reaction kinetic model parameters to be determined, and obtaining the corresponding initial reaction kinetic model based on the initial assignment of the reaction kinetic model parameters to be determined; iteratively training the initial heat transfer model based on the second sample data to obtain a reaction kinetic model that meets preset conditions; wherein different reaction kinetic models correspond to different stages in the dairy product heat treatment process using the dairy product heat processor.

[0112] In the process of building the reaction kinetics model, the algorithm corresponding to the reaction kinetics model is the Arrhenius equation, which is suitable for describing complex reactions in emulsion systems. Its calculation formula is as follows:

[0113]

[0114] Where T is the local milk temperature of the sterilizer internal pipeline, which is accurately calculated by the above heat transfer model, R is the gas constant, which is 8.314 J / (mol·K); K is the reaction rate used to describe the complex reaction of the emulsion system, that is, the reaction rate of the reaction product. 0 , activation energy Ea and reference temperature T 0 is the model parameter to be determined of the reaction kinetics model. The chemical reaction chain involved in the heat treatment process of dairy products is very complex. Temperature is an important influencing factor of complex reactions. The heat treatment model of the present invention combined with the heat transfer model can realize the prediction and calculation of dairy temperature values ​​for various reactions. Taking the isomerization reaction of lactose to produce lactulose as an example, lactose is defined as the reaction substrate and lactulose is defined as the reaction product in the reaction kinetics model. The concentration of the reaction product is used as the calculation target in the model. The concentration is related to the reaction rate and reaction time, respectively. The reaction time can be accurately calculated based on the dairy flow rate and the sterilizer pipeline structure. It will not be described in detail here. The reaction rate is calculated by the above-mentioned Arrhenius equation. Before starting the calculation, the pre-exponential factor K related to the lactose isomerization reaction is preliminarily set based on past experience. 0 =0.0001, activation energy Ea=100, reference temperature T 0 =110℃, and then the data was corrected later, and finally a reaction kinetic model for the isomerization of lactose to produce lactulose that met the engineering calculation was obtained.

[0115] In the process of reaction kinetic model verification, there is a set of corresponding reaction kinetic model parameters for each reaction product, that is, there are multiple reaction kinetic models, and corresponding to the above heat transfer model, by implementing data to verify the reaction kinetic model parameters (i.e., pre-exponential factor K 0 , activation energy Ea and reference temperature T 0 ) is used for verification to obtain a reaction kinetic model that meets engineering calculation requirements. Taking the lactose isomerization reaction to generate lactulose as an example, the specific model verification implementation process is as follows: Like the heat transfer model verification process, it is first necessary to design a multi-factor and multi-level implementation, see Appendix Table 3, with the milk temperature in the sterilization section of the sterilizer as the reaction temperature variable, and the different pipe lengths in the sterilization section as the reaction time variable. Each group of implementations is repeated twice to reduce the error in the implementation data detection. Implementation 1, Implementation 3 and Implementation 5 use a sterilization section pipe length of 3.52m, and Implementation 2, Implementation 4 and Implementation 6 use a pipe length of 22.1m. The implementation data for the lactulose reaction kinetic model verification are shown in Table 3 below:

[0116]

[0117] Table 3

[0118] Taking batch implementation data as input parameters, the iterative calculation method was used to continuously adjust the pre-exponential factor K0, activation energy Ea and reference temperature T0 to predict the lactulose concentration. Finally, the predicted value of the lactulose concentration at the sterilizer outlet was obtained. An F-test was performed on the two sets of data, the measured values ​​and the predicted values ​​of the product index parameters of the reaction kinetics model. At the significance level α=0.05, the calculated F value was smaller than the F table value, indicating that there was no significant difference between the predicted values ​​of the product index parameters of the reaction kinetics model and the measured values, that is, the reaction kinetics model met the engineering calculation conditions.

[0119] For example, Figure 4 As shown, the comparison between the measured values ​​corresponding to the reaction kinetics model and the predicted values ​​of the product index parameters is as follows:

[0120]

[0121] Table 4

[0122] It can be seen from the above figure and data calculation table 4 that the prediction and calculation of the thermosensitive indicator lactulose by the heat treatment model of the present invention meets the engineering calculation requirements. The prediction and calculation of other thermosensitive product index parameters including microbial inactivation, enzyme passivation and active protein denaturation can also be predicted and calculated by the heat treatment model of the present invention, which will not be repeated here.

[0123] In addition, after determining the weights of influence of multiple variable factors on each dairy product index parameter based on the concentration change rate of the reaction product, it also includes: determining the weights of influence of multiple variable factors on each dairy heat treatment processing intensity index parameter based on each dairy temperature value at the pipeline inlet and outlet nodes during the dairy heat treatment process; determining the corresponding target process parameter combination according to the size of the weights of influence of the multiple variable factors on each dairy heat treatment processing intensity index parameter and the initial process parameter combination. In addition, after determining the weights of influence of multiple variable factors on each dairy heat treatment processing intensity index parameter, a comprehensive comparison and analysis can be performed based on the weights of influence of the multiple variable factors on each dairy heat treatment processing intensity index parameter and the weights of influence of the multiple variable factors on each dairy product index parameter, and the corresponding final process parameter combination can be determined from the initial process parameter combination, which will not be described in detail here.

[0124] In the application stage of the heat treatment model, during the sensitivity analysis process, after completing the verification of the heat transfer model and the reaction kinetics model, a heat treatment model that conforms to engineering calculations is obtained. The model of the present invention can firstly use the powerful computing power of the computer to perform multivariate sensitivity analysis calculations. The specific implementation process is to set parameters in the model according to the actual operating range of each variable factor, such as the numerical range of the inlet water temperature of the second heating section is 95°C-155°C; the numerical range of the inlet water temperature of the first heating section is 80°C-100°C; the numerical range of the sterilization section pipe length includes 3.52m, 8.84m, 13.26m, 22.1m; the numerical range of the protein holding section pipe length includes 26.5m, 53.1m, 79.6m, 83.8m. The model can perform different process parameter combinations according to the interval values ​​corresponding to each variable factor, and then solve and calculate each process combination separately to obtain a large number of implementation data values, and finally obtain the influence weight of each variable on multiple calculation target values. Taking lactulose, furosine, bactericidal intensity F0 value, B value, and C value as calculation targets, the multivariate and multi-target sensitivity analysis calculation table is as follows:

[0125]

[0126]

[0127] Table 5

[0128] As shown in Table 5 above, the water temperature of the second heating section and the length of the sterilization section have the greatest influence on lactulose, furosine, sterilization intensity F0 value, B value, and C value. This is mainly because the water temperature of the second heating section determines the sterilization section dairy inlet temperature (for UHT milk, this temperature value is the upper limit temperature of heat treatment), and the length of the sterilization section determines the chemical reaction time of the dairy in the pipeline. The model of the present invention can perform model prediction calculations based on microbial inactivation, enzyme passivation, and protein denaturation, and obtain sensitivity analysis of more variables and calculation targets. It should be noted that in the above table, furosine and lactulose are the index parameters of each dairy product; the sterilization intensity F0 value, B value, and C value are the index parameters of dairy heat treatment processing intensity; the specific values ​​in the table are the corresponding influence weights under the process parameters corresponding to the different influencing factors of the output.

[0129] B value calculation formula

[0130] C value calculation formula

[0131] F0 value calculation formula

[0132] Among them, T can be the temperature value of the dairy products at each stage predicted by the heat transfer model.

[0133] In the process of product index prediction calculation and process optimization, the model of the present invention can calculate the corresponding product index parameters after inputting a set of process parameters. On this basis, a comprehensive impact analysis and calculation can be performed on multiple variables to obtain the optimal operating process parameters (i.e., the final process parameter combination) of these variables, and in the past, this part of the work often required a large amount of on-site repeated verification and implementation. For example, the sterilization section pipe length, the second heating section inlet water temperature, the protein holding section pipe length, and the first heating section inlet water temperature are input variables (i.e., the initial process parameter combination), and the influence trend of the four variable factors on the sterilizer outlet lactulose concentration is analyzed. The input parameters of each variable are shown in the table below, wherein the sterilization section and the protein holding section have four specifications of pipe length respectively, and the inlet water temperatures of the two heating sections are uniformly valued within the interval, and the water temperature interval determines the different heat treatment temperatures of the dairy products in the protein holding section and the sterilization section.

[0134] Variable factors Interval value The inlet water temperature of the second heating section 95℃-155℃ The first heating section inlet water temperature 80℃-100℃ Sterilization tube length 3.52m, 8.84m, 13.26m, 22.1m Protein retention tube length 26.5m, 53.1m, 79.6m, 83.8m

[0135] Table 6

[0136] As shown in Table 6 above, the above variable factors take different values ​​in the corresponding interval values ​​to obtain the initial process parameter combination, and the initial process parameter combination is predicted and analyzed based on the heat treatment model to obtain the lactulose concentration prediction calculation curve, as follows Figure 5 shown.

[0137] Step 103: Determine a corresponding target process parameter combination according to the weights of the influence of the plurality of variable factors on each of the dairy product index parameters and the initial process parameter combination.

[0138] Specifically, the influence weights of the multiple variable factors on the index parameters of each dairy product can be calculated respectively to obtain the mean of the influence weights of the multiple variable factors on the index parameters of each dairy product; and the corresponding target process parameter combination can be determined according to the mean of the influence weights of the multiple variable factors on the index parameters of each dairy product.

[0139] In an embodiment of the present invention, the heat transfer model describes the factors affecting the heat exchange of dairy products inside the shell-and-tube sterilizer. In the traditional electric power and chemical industries, the calculation correlation equations for the convective heat transfer of conventional fluids in round tubes have been very maturely applied. However, in the dairy field, since the material is a Newtonian fluid, a non-Newtonian fluid, or something in between, and the thermosensitivity of dairy products requires gentle heating and rapid cooling, these sensitive factors lead to a certain uncertainty in the heat exchange correlation equations for dairy products in the tube. Therefore, the heat treatment model of the present invention verifies the key core parameters of the heat transfer model through pilot implementation data, or for complex formula material systems and different forms of shell-and-tube sterilizers, the heat transfer model is modified and confirmed based on the implementation data. In addition, the key core parameters of the reaction kinetics model include K 0, Ea, T 0 For different reaction products, the Arrhenius equation describing the reaction rate has different parameters. The heat treatment model of the present invention describes the product indicators such as lactulose, furosine, active protein denaturation, microbial inactivation and enzyme passivation during the heat treatment of dairy products, but is not limited to the key core parameters of the above reactants, the pre-exponential factor K 0 , activation energy Ea and reference temperature T 0As a protection point. Multivariable and multi-target data association during heat treatment. The heat treatment process of dairy products involves multidisciplinary and multi-scale optimization problems, that is, the microscopic components of dairy products will affect the macroscopic heat transfer temperature distribution of dairy products, and the temperature distribution will further affect the concentration value of the reaction products. At the same time, the entire heat treatment process involves heat transfer, fluid mechanics, microbial chemistry, equipment mechanical design, etc. Therefore, the heat treatment model of the present invention is logically divided into a heat transfer model and a reaction kinetics model from the internal mechanism process of heat treatment, and the factors affecting the heat treatment process are digitally associated with the target value, thereby realizing scientific heat treatment process research. For other dairy heat treatment reactions, such as microbial inactivation reactions, enzyme passivation reactions, complex multi-step reactions, etc., the heat treatment model of the present invention can be continuously optimized and corrected based on the implementation data, and finally a reliable model that meets engineering calculations is obtained. In the existing heat treatment product process optimization analysis, generally only the inlet and outlet temperatures of each heat exchange section of the sterilizer can be monitored, and the temperature distribution along the pipe cannot be obtained, resulting in large errors in the calculated values ​​such as the sterilization intensity F0 value, and the reliability of the research cannot be guaranteed. The heat treatment model of the present invention can accurately calculate the real-time temperature distribution on the water side, metal pipe wall and dairy side of the internal flow direction of the sterilizer through the sub-heat transfer model, and guide R&D designers to conduct scientific analysis of the heat treatment process. Multivariable multi-objective optimization analysis. The existing technology generally conducts research on heat treatment processing technology through implementation. Due to the low flexibility of the process parameters of the on-site sterilizer and the high implementation cost, it is difficult to conduct a systematic study on all influencing variables. The heat treatment model of the present invention can deeply understand the coupling mechanism between product indicators and equipment structure and process parameters, improve R&D capabilities, reduce a large number of repetitive implementations such as small tests, pilot tests and trial production in the product development process, and realize efficient optimization analysis of dairy heat treatment product processes. The existing technology of data asset management generally implements research on single variables, and the variable data concerned is limited, making it difficult to think globally. This heat treatment invention model will be associated with different types of data such as dairy components, equipment structure dimensions, process and product indicators. The modeling process is the process of classifying and reprocessing different types of data. Combined with systematic implementation data, the heat treatment invention model can deeply explore the logical relationship behind these data, thereby forming data assets with core competitiveness of the enterprise. The existing technology for predicting and calculating product index parameters is difficult to accurately calculate the temperature gradient change value inside the pipeline, so it is impossible to accurately predict and calculate the reaction products. The heat treatment model of the present invention can predict and calculate the thermosensitive index parameters of products such as lactulose and furosine under different process parameters, thereby improving research and development efficiency. Global optimization analysis and operation space exploration. The existing CFD simulation technology can perform three-dimensional modeling of the sterilizer, and then predict and calculate the temperature field and reaction component concentration field. However, the modeling process has a long time cycle and high calculation cost. It is difficult to quickly perform global optimization analysis on the complex variables of the dairy heat treatment system. Generally, this technology is mainly used in equipment design optimization.The present invention constructs a key mechanism model of dairy heat treatment and uses mature one-dimensional software tools in traditional fields such as chemical industry to successfully realize the global optimization analysis of complex process variables in the food field, assisting R&D and design personnel in exploring the optimal process parameters for dairy heat treatment processing.

[0140] The method for optimizing the heat treatment process of dairy products provided by the present invention combines different process parameters by using interval values ​​corresponding to a plurality of variable factors associated with the heat treatment process of dairy products in advance, obtains an initial process parameter combination, performs a prediction analysis on the initial process parameter combination based on a preset heat treatment model, obtains the influence weights of the plurality of variable factors output by the heat treatment model on the index parameters of each dairy product, and determines the corresponding target process parameter combination according to the size of the influence weights of the plurality of variable factors on the index parameters of each dairy product and the initial process parameter combination. The method can quickly predict the process parameter combination in the heat treatment process of dairy products, effectively improves the optimization processing efficiency and accuracy of the heat treatment process of dairy products, and thus helps to reduce the number of optimization implementation tests.

[0141] Corresponding to the above-mentioned method for optimizing the heat treatment process of dairy products, the present invention also provides a device for optimizing the heat treatment process of dairy products. Since the embodiment of the device is similar to the above-mentioned method embodiment, the description is relatively simple. For relevant details, please refer to the description of the above-mentioned method embodiment. The embodiment of the device for optimizing the heat treatment process of dairy products described below is only illustrative. Please refer to Figure 6 As shown, it is a structural schematic diagram of a device for optimizing the heat treatment process of dairy products provided by an embodiment of the present invention.

[0142] The device for optimizing the heat treatment process of dairy products of the present invention specifically comprises the following parts:

[0143] The initial process parameter determination unit 601 is used to combine different process parameters according to the interval values ​​corresponding to the multiple variable factors associated with the dairy product heat treatment process, so as to obtain an initial process parameter combination; wherein the multiple variable factors are the influencing factors of the dairy product heat treatment process;

[0144] The heat treatment influence weight prediction unit 602 is used to predict and analyze the initial process parameter combination based on a preset heat treatment model, and obtain the influence weights of multiple variable factors output by the heat treatment model on each product index parameter of the dairy product; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and measured values;

[0145] The target process parameter determination unit 603 is used to determine the corresponding target process parameter combination according to the weights of the influence of the multiple variable factors on the index parameters of each dairy product and the initial process parameter combination.

[0146] Furthermore, the heat treatment model includes a heat transfer model and a reaction kinetics model; the heat transfer model is obtained by training based on the first sample data, the predicted value of the milk temperature during the heat treatment process of the milk corresponding to the first sample data, and the actual measured value of the milk temperature; the reaction kinetics model is obtained by training based on the second sample data, the predicted value of the product index parameter during the heat treatment process of the milk corresponding to the second sample data, and the actual measured value of the product index parameter;

[0147] The heat treatment model is used to predict the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products; the reaction kinetics model is used to predict the concentration change rate of the reaction products corresponding to the index parameters of each dairy product under different process parameter combinations of multiple variable factors during the heat treatment process of dairy products; wherein the multiple variable factors include the dairy product reaction temperature variable, and the dairy product reaction temperature variable includes the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline; the concentration change rate of the reaction product corresponds to the reaction rate of the reaction product;

[0148] The heat treatment influence weight prediction unit is specifically used for:

[0149] Inputting the target values ​​of multiple variable factors corresponding to the initial process parameter combination into the heat treatment model to predict the heat treatment temperature value, and obtaining the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products;

[0150] The temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process are input into the reaction kinetics model to predict the product index concentration, and the concentration change rate of the reaction product corresponding to each dairy product index parameter under different process parameter combinations of multiple variable factors during the dairy product heat treatment process is obtained; the concentration change rate is the speed at which the concentration of the reaction product changes over time;

[0151] Based on the concentration change rate of the reaction product, the influence weights of multiple variable factors on each product index parameter of the dairy product are determined; wherein the concentration change rate and the influence weight are in direct proportion.

[0152] Furthermore, before predicting and analyzing the initial process parameter combination based on the preset heat treatment model to obtain the influence weights of the multiple variable factors output by the heat treatment model on the various product index parameters of the dairy product, it also includes: a model building unit for:

[0153] Construct a heat transfer model to predict the temperature of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process; and,

[0154] A reaction kinetic model is constructed to predict the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors during the heat treatment process of dairy products.

[0155] Furthermore, the heat transfer model for predicting the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products is constructed, specifically including:

[0156] Obtaining pre-collected first sample data; wherein the first sample data includes component content data of the dairy product, temperature data of the heating medium side in the dairy product heat processor, and flow data of the heating medium side;

[0157] Obtaining initial values ​​of input parameters of the heat transfer model to be determined, and obtaining a corresponding initial heat transfer model based on the initial values ​​of the parameters of the heat transfer model to be determined;

[0158] The initial heat transfer model is iteratively trained based on the first sample data to obtain a heat transfer model that meets preset conditions; wherein different heat transfer models correspond to different stages in the process of heat treating the dairy product using the dairy product heat processor.

[0159] Furthermore, the reaction kinetic model constructed for predicting the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors in the dairy product heat treatment process specifically includes:

[0160] Obtaining collected second sample data; wherein the second sample data includes previously predicted temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process;

[0161] Obtaining initial values ​​of input parameters of the reaction kinetic model to be determined, and obtaining a corresponding initial reaction kinetic model based on the initial values ​​of the parameters of the reaction kinetic model to be determined;

[0162] Iteratively training the initial heat transfer model based on the second sample data to obtain a reaction kinetics model that meets preset conditions;

[0163] Wherein, different stages in the process of heat treating the dairy product by using the dairy product heat processor correspond to different reaction kinetic models.

[0164] Furthermore, after determining the influence weights of multiple variable factors on each dairy product index parameter based on the concentration change rate of the reaction product, it also includes: a second target process parameter determination unit, which is used to determine the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter based on each dairy product temperature value at the pipeline inlet and outlet nodes during the dairy product heat treatment process; determine the corresponding target process parameter combination according to the size of the influence weights of the multiple variable factors on each dairy product heat treatment processing intensity index parameter and the initial process parameter combination.

[0165] Furthermore, after determining the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter, it also includes:

[0166] The third target process parameter determination unit is used to conduct a comprehensive comparison and analysis based on the influence weights of the multiple variable factors on each dairy product heat treatment processing intensity index parameter and the influence weights of the multiple variable factors on each dairy product index parameter, so as to determine the corresponding final process parameter combination from the initial process parameter combination.

[0167] The device for optimizing the heat treatment process of dairy products provided by the present invention combines different process parameters by using interval values ​​corresponding to a plurality of variable factors associated with the heat treatment process of dairy products in advance, obtains an initial process parameter combination, performs a prediction analysis on the initial process parameter combination based on a preset heat treatment model, obtains the influence weights of the plurality of variable factors output by the heat treatment model on the index parameters of each dairy product, and determines the corresponding target process parameter combination according to the influence weights of the plurality of variable factors on the index parameters of each dairy product and the initial process parameter combination. The method can quickly predict the process parameter combination in the heat treatment process of dairy products, effectively improves the optimization processing efficiency and accuracy of the heat treatment process of dairy products, and thus helps to reduce the number of optimization implementation tests.

[0168] Corresponding to the above-mentioned method for optimizing the heat treatment process of dairy products, the present invention also provides an electronic device. Since the embodiment of the electronic device is similar to the above-mentioned method embodiment, the description is relatively simple. For the relevant parts, please refer to the description of the above-mentioned method embodiment. The electronic device described below is only for illustration. Figure 7As shown, it is a schematic diagram of the physical structure of an electronic device disclosed in an embodiment of the present invention. The electronic device may include: a processor 701, a memory 702 and a communication bus 703, wherein the processor 701 and the memory 702 communicate with each other through the communication bus 703 and communicate with the outside through the communication interface 704. The processor 701 can call the logic instructions in the memory 702 to execute a method for optimizing the heat treatment process of dairy products, the method comprising: combining different process parameters according to interval values ​​corresponding to a plurality of variable factors predetermined to be associated with the heat treatment process of dairy products to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the heat treatment process of dairy products; predicting and analyzing the initial process parameter combination based on a preset heat treatment model to obtain the influence weights of the plurality of variable factors output by the heat treatment model on each index parameter of dairy products; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and measured values; determining the corresponding target process parameter combination according to the size of the influence weights of the plurality of variable factors on each index parameter of dairy products and the initial process parameter combination.

[0169] In addition, the logic instructions in the above-mentioned memory 702 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: a storage chip, a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, and other media that can store program codes.

[0170] On the other hand, an embodiment of the present invention further provides a computer program product, which includes a computer program stored on a processor-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the dairy heat treatment process optimization processing method provided by the above-mentioned method embodiments, the method comprising: combining different process parameters according to interval values ​​corresponding to multiple variable factors predetermined to be associated with the dairy heat treatment process to obtain an initial process parameter combination; wherein the multiple variable factors are influencing factors of the dairy heat treatment process; predicting and analyzing the initial process parameter combination based on a preset heat treatment model to obtain the influence weights of the multiple variable factors output by the heat treatment model on each dairy product index parameter; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and measured values; determining the corresponding target process parameter combination according to the size of the influence weights of the multiple variable factors on each dairy product index parameter and the initial process parameter combination.

[0171] On the other hand, an embodiment of the present invention further provides a processor-readable storage medium, on which a computer program is stored. When the computer program is executed by the processor, it is implemented to execute the method for optimizing the heat treatment process of dairy products provided in the above embodiments, the method comprising: combining different process parameters according to interval values ​​corresponding to a plurality of variable factors predetermined to be associated with the heat treatment process of dairy products to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the heat treatment process of dairy products; predicting and analyzing the initial process parameter combination based on a preset heat treatment model to obtain the influence weights of the plurality of variable factors output by the heat treatment model on each index parameter of dairy products; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and measured values; determining the corresponding target process parameter combination according to the size of the influence weights of the plurality of variable factors on each index parameter of dairy products and the initial process parameter combination.

[0172] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD)), etc.

[0173] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0174] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0175] Finally, it should be noted that the above 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the heat treatment process of dairy products, characterized in that: include: Combining different process parameters according to interval values ​​corresponding to a plurality of variable factors associated with the dairy product heat treatment process, to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the dairy product heat treatment process; Based on a preset heat treatment model, the initial process parameter combination is predicted and analyzed to obtain the influence weights of multiple variable factors output by the heat treatment model on various product index parameters of dairy products; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and actual measured values; According to the weights of the influence of the multiple variable factors on the index parameters of each dairy product and the initial process parameter combination, the corresponding target process parameter combination is determined.

2. The method for optimizing the heat treatment process of dairy products according to claim 1, characterized in that: The heat treatment model includes a heat transfer model and a reaction kinetics model; the heat transfer model is obtained by training based on the first sample data, the predicted value of the milk temperature during the heat treatment process of the milk corresponding to the first sample data, and the actual measured value of the milk temperature; the reaction kinetics model is obtained by training based on the second sample data, the predicted value of the product index parameter during the heat treatment process of the milk corresponding to the second sample data, and the actual measured value of the product index parameter; The heat treatment model is used to predict the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products; the reaction kinetics model is used to predict the concentration change rate of the reaction products corresponding to the index parameters of each dairy product under different process parameter combinations of multiple variable factors during the heat treatment process of dairy products; wherein the multiple variable factors include the dairy product reaction temperature variable, and the dairy product reaction temperature variable includes the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline; the concentration change rate of the reaction product corresponds to the reaction rate of the reaction product; The method of predicting and analyzing the initial process parameter combination based on the preset heat treatment model to obtain the influence weights of multiple variable factors output by the heat treatment model on various product index parameters of dairy products includes: Inputting the target values ​​of multiple variable factors corresponding to the initial process parameter combination into the heat treatment model to predict the heat treatment temperature value, and obtaining the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products; The temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process are input into the reaction kinetics model to predict the product index concentration, and the concentration change rate of the reaction product corresponding to each dairy product index parameter under different process parameter combinations of multiple variable factors during the dairy product heat treatment process is obtained; the concentration change rate is the speed at which the concentration of the reaction product changes over time; Based on the concentration change rate of the reaction product, the influence weights of multiple variable factors on each product index parameter of the dairy product are determined; wherein the concentration change rate and the influence weight are in direct proportion.

3. The method for optimizing the heat treatment process of dairy products according to claim 1, characterized in that: Before predicting and analyzing the initial process parameter combination based on the preset heat treatment model to obtain the influence weights of multiple variable factors output by the heat treatment model on each product index parameter of the dairy product, the method further includes: Construct a heat transfer model to predict the temperature of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process; and, A reaction kinetic model is constructed to predict the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations under multiple variable factors during the heat treatment process of dairy products.

4. The method for optimizing the heat treatment process of dairy products according to claim 3, characterized in that: The heat transfer model for predicting the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the heat treatment process of dairy products specifically includes: Obtaining pre-collected first sample data; wherein the first sample data includes component content data of the dairy product, temperature data of the heating medium side in the dairy product heat processor, and flow data of the heating medium side; Obtaining initial values ​​of input parameters of the heat transfer model to be determined, and obtaining a corresponding initial heat transfer model based on the initial values ​​of the parameters of the heat transfer model to be determined; The initial heat transfer model is iteratively trained based on the first sample data to obtain a heat transfer model that meets preset conditions; wherein different heat transfer models correspond to different stages in the process of heat treating the dairy product using the dairy product heat processor.

5. The method for optimizing the heat treatment process of dairy products according to claim 3, characterized in that: The reaction kinetic model constructed for predicting the concentration change rate of reaction products corresponding to various product index parameters of dairy products under different process parameter combinations of multiple variable factors during the dairy product heat treatment process specifically includes: Obtaining collected second sample data; wherein the second sample data includes previously predicted temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process; Obtaining initial values ​​of input parameters of the reaction kinetic model to be determined, and obtaining a corresponding initial reaction kinetic model based on the initial values ​​of the parameters of the reaction kinetic model to be determined; Iteratively training the initial heat transfer model based on the second sample data to obtain a reaction kinetics model that meets preset conditions; Wherein, different stages in the process of heat treating the dairy product by using the dairy product heat processor correspond to different reaction kinetic models.

6. The method for optimizing the heat treatment process of dairy products according to claim 2, characterized in that: After determining the influence weights of multiple variable factors on each product index parameter of the dairy product based on the concentration change rate of the reaction product, the method further includes: Based on the temperature values ​​of each dairy product at the inlet and outlet nodes of the pipeline during the dairy product heat treatment process, the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter are determined; according to the size of the influence weights of the multiple variable factors on each dairy product heat treatment processing intensity index parameter and the initial process parameter combination, the corresponding target process parameter combination is determined.

7. The method for optimizing the heat treatment process of dairy products according to claim 6, characterized in that: After determining the influence weights of multiple variable factors on each dairy product heat treatment processing intensity index parameter, it also includes: A comprehensive comparison and analysis is performed based on the influence weights of the multiple variable factors on each dairy heat treatment processing intensity index parameter and the influence weights of the multiple variable factors on each dairy product index parameter, so as to determine the corresponding final process parameter combination from the initial process parameter combination.

8. A device for optimizing the heat treatment process of dairy products, characterized in that: include: An initial process parameter determination unit is used to combine different process parameters according to interval values ​​corresponding to a plurality of variable factors associated with the dairy product heat treatment process, so as to obtain an initial process parameter combination; wherein the plurality of variable factors are influencing factors of the dairy product heat treatment process; A heat treatment influence weight prediction unit is used to predict and analyze the initial process parameter combination based on a preset heat treatment model, and obtain the influence weights of multiple variable factors output by the heat treatment model on each product index parameter of the dairy product; wherein the heat treatment model is trained based on sample data, sample prediction values ​​corresponding to the sample data, and actual measured values; The target process parameter determination unit is used to determine the corresponding target process parameter combination according to the influence weights of the multiple variable factors on each dairy product index parameter and the initial process parameter combination.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method for optimizing the heat treatment process of dairy products as claimed in any one of claims 1 to 7 are implemented.

10. A processor-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for optimizing the heat treatment process of dairy products as claimed in any one of claims 1 to 7 are implemented.

Images