Intelligent regulation and control method for radio frequency vacuum drying of kiwi fruit slices based on numerical simulation optimization

Abstract

The invention discloses an intelligent regulation and control method for radio frequency vacuum drying of kiwi fruit slices based on numerical simulation optimization, which belongs to the field of fruit and vegetable drying processing and comprises the steps of sample pretreatment, actual drying experiment, construction of a radio frequency vacuum drying physical model, setting of a control equation and boundary conditions, model solving and verification, intelligent regulation and control method screening and the like. A physical model containing Joule heating and dilute substance transfer modules is constructed through COMSOL Multiphysics (v5.6) software, after the effectiveness of the model is verified through a decision coefficient (R2) and a root-mean-square error (RMSE), multiple intelligent regulation and control methods are implemented in the model, and an optimal scheme is screened by comparing a heating uniformity index () and energy consumption. According to the method, the temperature and moisture change of the kiwi fruit slices in the drying process can be accurately predicted, the optimal intelligent regulation and control method can be rapidly screened at low cost, the experiment cost is saved, the efficiency is improved, and efficient support is provided for research and development of a radio frequency vacuum drying intelligent regulation and control technology.

Description

Technical Field

[0001] This invention relates to the field of fruit and vegetable drying and processing technology, and in particular to a method for intelligent control of radio frequency vacuum drying of kiwifruit slices based on numerical simulation optimization. Background Technology

[0002] Kiwifruit is rich in various bioactive components, possessing both high nutritional value and health benefits. my country is the world's largest producer of kiwifruit. However, the high water content and susceptibility to spoilage of kiwifruit severely restrict the high-value development of the industry. Processing fresh fruit into dried fruit has become the mainstream approach to extend shelf life and increase added value.

[0003] Traditional drying methods, such as hot air drying and sun drying, generally suffer from drawbacks such as high energy consumption, long drying cycles, and easy deterioration of product quality. Radio frequency (RF) vacuum drying, as a novel electromagnetic heating technology, heats the internal volume of materials through an RF electric field, and combines this with vacuum conditions to reduce the latent heat of vaporization, enabling rapid and high-quality drying. However, existing RF vacuum drying methods often use fixed operating parameters, neglecting the changes in material properties with temperature and moisture content during the drying process, affecting heating uniformity and making it difficult to guarantee optimal heating performance throughout the entire cycle.

[0004] Intelligent control technology can dynamically regulate materials by monitoring changes in material indicators in real time. However, existing research is limited by experimental costs, making it impossible to verify and screen control methods one by one. This results in intelligent control methods being mostly based on fixed parameter assumptions, which deviate from actual working conditions. Numerical simulation technology can analyze, visualize, and accurately predict the temperature, moisture, and electromagnetic field distributions during the drying process in real time. By quantifying the intelligent control logic and integrating it into the simulation model, it can effectively solve the problems of difficulty in selecting intelligent control methods, high experimental costs, and limited applicability, providing efficient support for the research and development of intelligent control technology for radio frequency vacuum drying. Summary of the Invention

[0005] This invention aims to solve the problems of high difficulty in selecting intelligent control methods for radio frequency vacuum drying of kiwifruit slices, high experimental costs, and limited applicability. It provides an intelligent control strategy for radio frequency vacuum drying of kiwifruit slices that can achieve optimal heating uniformity and enables precise selection of intelligent control methods.

[0006] Technical Solution: To solve the above-mentioned technical problems, according to one aspect of the present invention, more specifically, a method for intelligent control of radio frequency vacuum drying of kiwifruit slices based on numerical simulation optimization, comprising the following steps: S1, Sample Pretreatment Fresh, undamaged "Xu Xiang" kiwifruit (initial wet basis (wb) moisture content (%) 83.30 ± 0.01, soluble solids content (°Brix) 14.83 ± 0.49) were peeled using a peeler and sliced ​​into thin slices with a thickness of 7.60 ± 0.20 mm, a major axis of 5.25 ± 0.30 cm, and a minor axis of 4.17 ± 0.24 cm. Twenty-four kiwifruit slices (total mass 384.74 ± 1.20 g) were evenly laid out in a rectangular tray, with a center-to-center distance of 65 mm between adjacent samples. The tray dimensions were 400 mm (length) × 270 mm (width) × 20 mm (height). A schematic diagram of the kiwifruit slices' placement within the RF vacuum drying chamber is shown below. Figure 1 As shown.

[0007] S2, Actual Drying Experiment A tray was placed at the center of the lower electrode of a 6 kW RF vacuum drying apparatus (GJS-3-27-JY, Hebei Huashijiyuan High Frequency Equipment Co., Ltd.). Three fiber optic temperature sensors (HQ-FTS-D120, Xi'an Heqi Optoelectronic Technology Co., Ltd.) were inserted into the kiwi slices in the corners, edges, and center of the tray, with an insertion depth of 23 mm and parallel to the radial direction of the kiwi slices. The kiwi slices were dried at a drying temperature of 60 ± 0.5 °C, a vacuum degree of 0.023 ± 0.003 MPa, and an electrode spacing of 75 mm. Drying was completed when the dry basis (db) moisture content of the kiwi slices was below 0.18 g / g, and the actual experimental results were obtained. The weight of the kiwi slices was measured every 30 minutes by observing the front panel of the RF vacuum drying apparatus, and the moisture content of the kiwi slices was calculated.

[0008] The formula for calculating the moisture content of kiwi slices is: Formula 1: Formula 2: In the formula, and Kiwi slices Moisture content on a wet or dry basis at any given time; for The quality of the kiwi slices at all times; This refers to the absolute dry quality of the kiwi slices.

[0009] After confirming the current operating parameters, place three fiber optic temperature sensors 5 mm to the left of the kiwi slices at three different locations to measure the temperature change of water vapor inside the cavity during the evaporation process.

[0010] S3, Constructing a Physical Model Based on the finite element method and using COMSOL Multiphysics® (v5.6) software, a physical model for kiwifruit slice radio frequency vacuum drying was established according to the dimensions of the kiwifruit slices and the radio frequency vacuum drying equipment, as well as the control logic of the radio frequency vacuum drying equipment. The physical model includes Joule heating and rare matter transfer modules.

[0011] The assumptions made before establishing the model were as follows: the air temperature inside the radio frequency cavity is constant during radio frequency heating; the solid, liquid water, and gas phases are continuous; all fluid phases share the gas pressure; the solid, liquid, and gas phases inside the kiwi slice maintain local thermal equilibrium; the kiwi slice is an isotropic porous medium; the radio frequency vacuum drying of the kiwi slice is mainly volumetric evaporation, and there is no surface mass flux during the simulation; the thermal expansion and shrinkage of the kiwi during the drying process are ignored; the influence of gravity is ignored, and the dry basis mass of the material remains unchanged during the shrinkage process; the influence of the polypropylene container on the sample during drying is ignored; the influence of vacuum fluctuations on the sample during drying is ignored; and the response time of the sensor used in the intelligent control method is ignored.

[0012] S4. Constructing the geometric model Using COMSOL Multiphysics® (v5.6) software, a geometric model was constructed based on the actual dimensions of the kiwi slices and the RF vacuum equipment. The constructed physical model of the kiwi slices for RF vacuum drying is as follows: Figure 2 As shown.

[0013] S5. Define the governing equations and boundary conditions. Electromagnetic field related equations and boundary conditions: The electromagnetic field distribution in the kiwi slices was determined by solving Maxwell's equations. Since the radio frequency wavelength (11 m) is much larger than the maximum size of the material, the influence of the magnetic field was ignored, and Maxwell's equations were simplified to Laplace's equations. The governing equations of the electromagnetic field are: Formula 3: In the formula, This indicates the electrical conductivity of the material (S / m). Indicates the imaginary part (-1). 0.5 ; The radio frequency is 27.12MHz. The vacuum permittivity is 8.854 × 10⁻⁶. -12 F / m); This represents the voltage (V) between the two plates.

[0014] Electromagnetic energy provided by electromagnetic fields ( W / m 3 The calculation formula is: Formula 4: In the formula, The imaginary part (F / m) represents the complex absolute permittivity. Represents electric field The modulus is obtained by solving formula (3).

[0015] The boundary conditions for the electromagnetic field are: Formula 5: Formula 6: Formula 7: Formula (5) indicates that the RF external cavity wall is electrically insulated, and in Formula (6) With formula (7) These represent the actual voltage of the upper plate and the grounding of the lower plate, respectively.

[0016] In the physical model for control logic for: Formula 8: In the formula, This represents the maximum temperature returned by the three fiber optic temperature sensors, obtained by constructing three point probes and taking the maximum value. In the software, this is expressed as: if((max(point1,max(point2,point3))<=332.65[K]),ec.Qh,if((max(point1,max(point2,point3))>=333.65[K]),0,ec.Qh)).

[0017] The dielectric properties of kiwi slices were calculated using the Landau and Lifshitz, Looyenga equation (LLLE), as follows: Formula 9: In the formula, Represents the volume fraction of phase i; The dielectric properties of phase i are represented by s, w, and v, which represent the solid, liquid, and gas phases, respectively.

[0018] Mass transfer control equation: Calculate the concentrations of liquid water and water vapor using the law of conservation of mass (Fick's Law). (There is no mass transfer convection term, and the mass transfer governing equation is:) Formula 10: Formula 11: Formula 12: Formula 13: Formula 14: Formula 15: In the formula, and These represent the concentrations of liquid water and water vapor, respectively (mol / m³). 3 ); and These represent the effective diffusion coefficients (m) of liquid water and water vapor, respectively. 2 / s); Evaporation rate (kg / (m³)) 3 ·s)); The density of liquid water (kg / m³) 3 ); The equilibrium vapor pressure (Pa) of water vapor in the material; Vacuum pressure (Pa); The saturated vapor pressure (Pa) of water vapor in the material is calculated using the Antoni equation; The water activity of the sample; Let be the evaporation rate constant (1 / s). It is related to the dry basis moisture content The relevant piecewise functions, The size of the transition region is an empirical constant.

[0019] Heat transfer and energy conservation governing equations: Without heat transfer and convection terms, the governing equations combining heat transfer and energy conservation laws are: Formula 16: Equation 17: Formula 18: Formula 19: Formula 20: Equation 21: In the formula, Indicates the effective density of kiwi slices (kg / m³) 3 ); This indicates the effective specific heat capacity (J / (kg·K)). Indicates the effective thermal conductivity (W / (m·K)); Indicates total energy exchange; Indicates latent heat of vaporization (J / kg); , , These represent the density, specific heat capacity, and thermal conductivity of the solid phase, respectively. , , These represent the relevant parameters of liquid water; , These represent relevant air parameters; , , The mass (kg) of the solid phase, liquid phase and gas phase are respectively represented. and These represent the mass fractions of water vapor and air, respectively. and These are the liquid phase and gas phase saturation, respectively. Porosity.

[0020] The boundary conditions for the heat transfer physical field are: Equation 22: Equation 23: Formula 24: Formula 25: In the formula, This represents the boundary flux (W / (m²)) on the surface of the kiwifruit slice. 2 ·s)); The ambient temperature (K) is the average of the water vapor temperature in the three cavity locations. The convective heat transfer coefficient (W / (m) 2 ·K)); Indicates the equivalent diameter (m) of the kiwi slice; Represents the Reynolds number; Represent Prandtl numbers; Indicates air velocity (m / s); Aerodynamic viscosity (Pa·m).

[0021] The remaining parameters of the model are attached. Figure 6 With appendix Figure 7 .

[0022] S6. Mesh Independence Check: Solve the model sequentially with ultra-coarsened, coarsened, refined, relatively refined, ultra-refined, and extremely refined tetrahedral meshes. When the difference between the simulated water content and electric field intensity is <0.1%, determine the optimal mesh density and select the ultra-refined tetrahedral mesh for this invention. S7. Model Solving A transient solver, a fully coupled and iterative solution mode, and the iterative solver was FGMRES. Mesh independence was checked during the simulation, and the model was solved sequentially under different mesh densities. The optimal mesh density was determined when the difference between the simulated moisture content and electric field intensity was <0.1%. This study selected an ultra-fine tetrahedral mesh with a time step of 1 min. The mesh generation diagram of the kiwifruit slice radio frequency vacuum drying physical model is shown below. Figure 3 As shown. To balance computational accuracy and time, the relative tolerance is set to 1. Each run takes approximately 3 hours. The workstation configuration for running the software is: Dell T3660 processor, Intel CPU i9-13900K 3.00GHz, 128 GB of memory, and Windows Server 2019 R2 Standard 64-bit operating system.

[0023] S8, Model Validation The results of the radio frequency vacuum drying simulation of kiwi fruit slices were obtained, including the average dry basis moisture content after simulation and the temperature change curves of the corners, edges, and center. The coefficient of determination (COP) was used. ) and root mean square error ( The similarity between simulation and experimental results is assessed using the following formula: Equation 26: Equation 27: in, This represents the total number of experimental indicators over time. and These represent the experimental and predicted values, respectively. This is the average of the experimental values. When the average dry basis moisture content is... >0.95、 <1, three-point temperature The model is usable when the values ​​are >0.80 and the RMSE is <5 °C.

[0024] S9, Screening of Intelligent Control Methods The selected intelligent control method was implemented in the validated model, and the heating uniformity index before and after implementation was compared. Based on the energy consumption (kW·h), the optimal intelligent control method is selected. The formula for calculating heating uniformity is: Formula 28: In the formula, This indicates the average temperature (°C) of all the kiwi slices. This represents the total volume of all kiwi slices. Calculated in the software. The expression is abs( [1 / K]-273.15- ) / ( - ) / (a single kiwi slice) The expression for calculating energy consumption is ) / 1[m^3] / 24; In the software, the statement ` / 60000` sums the results after removing the first dot. Energy consumption is only compared in terms of numerical value.

[0025] The beneficial effects of the intelligent control method for radio frequency vacuum drying of kiwifruit slices based on numerical simulation optimization in this invention are as follows: This invention accurately predicts the average moisture content and temperature change curves at three points during the radio frequency vacuum drying process of kiwi slices. The verified model can guide the actual drying process. The software implements multiple methods for intelligently controlling the radio frequency vacuum drying of kiwifruit slices, eliminating the need for extensive actual experiments and saving experimental costs. By comparing the TUI and energy consumption of various intelligent control methods, the optimal intelligent control method can be quickly selected, thus improving the efficiency of intelligent control technology research and development. Attached Figure Description

[0026] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0027] Figure 1 : A diagram showing the arrangement of kiwi slices (where A: center; B: edge; C: corner). Figure 2 Physical model of radio frequency vacuum drying of kiwi fruit slices; Figure 3 Mesh generation diagram of the physical model for radio frequency vacuum drying of kiwi fruit slices; Figure 4 This is a comparison chart of TUI results in an embodiment of the present invention; Figure 5 This is a graph showing the energy consumption comparison results of an embodiment of the present invention; Figure 6 Input parameters and reference values ​​for the radio frequency vacuum drying model of kiwi slices in this invention. Figure 1 ; Figure 7 Input parameters and reference values ​​for the radio frequency vacuum drying model of kiwi slices in this invention. Figure 2 . Detailed Implementation

[0028] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0029] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] Example 1 Reference Figures 1-7 A method for intelligent control of radio frequency vacuum drying of kiwifruit slices based on numerical simulation optimization includes the following steps: S1, Sample Pretreatment Fresh, undamaged "Xu Xiang" kiwifruit (initial wet basis (wb) moisture content (%) 83.30 ± 0.01, soluble solids content (°Brix) 14.83 ± 0.49) were peeled using a peeler and sliced ​​into thin slices with a thickness of 7.60 ± 0.20 mm, a major axis of 5.25 ± 0.30 cm, and a minor axis of 4.17 ± 0.24 cm. Twenty-four kiwifruit slices (total mass 384.74 ± 1.20 g) were evenly laid out in a rectangular tray, with a center-to-center distance of 65 mm between adjacent samples. The tray dimensions were 400 mm (length) × 270 mm (width) × 20 mm (height). A schematic diagram of the kiwifruit slices' placement within the RF vacuum drying chamber is shown below. Figure 1 As shown.

[0031] S2, Actual Drying Experiment A tray was placed at the center of the lower electrode of a 6 kW RF vacuum drying apparatus (GJS-3-27-JY, Hebei Huashijiyuan High Frequency Equipment Co., Ltd.). Three fiber optic temperature sensors (HQ-FTS-D120, Xi'an Heqi Optoelectronic Technology Co., Ltd.) were inserted into the kiwi slices in the corners, edges, and center of the tray, with an insertion depth of 23 mm and parallel to the radial direction of the kiwi slices. The kiwi slices were dried at a drying temperature of 60 ± 0.5 °C, a vacuum degree of 0.023 ± 0.003 MPa, and an electrode spacing of 75 mm. Drying was completed when the dry basis (db) moisture content of the kiwi slices was below 0.18 g / g, and the actual experimental results were obtained. The weight of the kiwi slices was measured every 30 minutes by observing the front panel of the RF vacuum drying apparatus, and the moisture content of the kiwi slices was calculated.

[0032] The formula for calculating the moisture content of kiwi slices is: Formula 1: Formula 2: In the formula, and Kiwi slices Moisture content on a wet or dry basis at any given time; for The quality of the kiwi slices at all times; This refers to the absolute dry quality of the kiwi slices.

[0033] After confirming the current operating parameters, place three fiber optic temperature sensors 5 mm to the left of the kiwi slices at three different locations to measure the temperature change of water vapor inside the cavity during the evaporation process.

[0034] S3, Constructing a Physical Model Based on the finite element method and using COMSOL Multiphysics® (v5.6) software, a physical model for kiwifruit slice radio frequency vacuum drying was established according to the dimensions of the kiwifruit slices and the radio frequency vacuum drying equipment, as well as the control logic of the radio frequency vacuum drying equipment. The physical model includes Joule heating and rare matter transfer modules.

[0035] The assumptions made before establishing the model were as follows: the air temperature inside the radio frequency cavity is constant during radio frequency heating; the solid, liquid water, and gas phases are continuous; all fluid phases share the gas pressure; the solid, liquid, and gas phases inside the kiwi slice maintain local thermal equilibrium; the kiwi slice is an isotropic porous medium; the radio frequency vacuum drying of the kiwi slice is mainly volumetric evaporation, and there is no surface mass flux during the simulation; the thermal expansion and shrinkage of the kiwi during the drying process are ignored; the influence of gravity is ignored, and the dry basis mass of the material remains unchanged during the shrinkage process; the influence of the polypropylene container on the sample during drying is ignored; the influence of vacuum fluctuations on the sample during drying is ignored; and the response time of the sensor used in the intelligent control method is ignored.

[0036] S4. Constructing the geometric model Using COMSOL Multiphysics® (v5.6) software, a geometric model was constructed based on the actual dimensions of the kiwi slices and the RF vacuum equipment. The constructed physical model of the kiwi slices for RF vacuum drying is as follows: Figure 2 As shown.

[0037] S5. Define the governing equations and boundary conditions. Electromagnetic field related equations and boundary conditions: The electromagnetic field distribution in the kiwi slices was determined by solving Maxwell's equations. Since the radio frequency wavelength (11 m) is much larger than the maximum size of the material, the influence of the magnetic field was ignored, and Maxwell's equations were simplified to Laplace's equations. The governing equations of the electromagnetic field are: Formula 3: In the formula, This indicates the electrical conductivity of the material (S / m). Indicates the imaginary part (-1). 0.5 ; The radio frequency is 27.12MHz. The vacuum permittivity is 8.854 × 10⁻⁶. -12 F / m); This represents the voltage (V) between the two plates.

[0038] Electromagnetic energy provided by electromagnetic fields ( W / m 3 The calculation formula is: Formula 4: In the formula, The imaginary part (F / m) represents the complex absolute permittivity. Represents electric field The modulus is obtained by solving formula (3).

[0039] The boundary conditions for the electromagnetic field are: Formula 5: Formula 6: Formula 7: Formula (5) indicates that the RF external cavity wall is electrically insulated, and in Formula (6) With formula (7) These represent the actual voltage of the upper plate and the grounding of the lower plate, respectively.

[0040] In the physical model for control logic for: Formula 8: In the formula, This represents the maximum temperature returned by the three fiber optic temperature sensors, obtained by constructing three point probes and taking the maximum value. In the software, this is expressed as: if((max(point1,max(point2,point3))<=332.65[K]),ec.Qh,if((max(point1,max(point2,point3))>=333.65[K]),0,ec.Qh)).

[0041] The dielectric properties of kiwi slices were calculated using the Landau and Lifshitz, Looyenga equation (LLLE), as follows: Formula 9: In the formula, Represents the volume fraction of phase i; The dielectric properties of phase i are represented by s, w, and v, which represent the solid, liquid, and gas phases, respectively.

[0042] Mass transfer control equation: Calculate the concentrations of liquid water and water vapor using the law of conservation of mass (Fick's Law). (There is no mass transfer convection term, and the mass transfer governing equation is:) Formula 10: Formula 11: Formula 12: Formula 13: Formula 14: Formula 15: In the formula, and These represent the concentrations of liquid water and water vapor, respectively (mol / m³). 3 ); and These represent the effective diffusion coefficients (m) of liquid water and water vapor, respectively. 2 / s); Evaporation rate (kg / (m³)) 3 ·s)); The density of liquid water (kg / m³) 3 ); The equilibrium vapor pressure (Pa) of water vapor in the material; Vacuum pressure (Pa); The saturated vapor pressure (Pa) of water vapor in the material is calculated using the Antoni equation; The water activity of the sample; Let be the evaporation rate constant (1 / s). It is related to the dry basis moisture content The relevant piecewise functions, The size of the transition region is an empirical constant.

[0043] Heat transfer and energy conservation governing equations: Without heat transfer and convection terms, the governing equations combining heat transfer and energy conservation laws are: Formula 16: Equation 17: Formula 18: Formula 19: Formula 20: Equation 21: In the formula, Indicates the effective density of kiwi slices (kg / m³) 3 ); This indicates the effective specific heat capacity (J / (kg·K)). Indicates the effective thermal conductivity (W / (m·K)); Indicates total energy exchange; Indicates latent heat of vaporization (J / kg); , , These represent the density, specific heat capacity, and thermal conductivity of the solid phase, respectively. , , These represent the relevant parameters of liquid water; , These represent relevant air parameters; , , The mass (kg) of the solid phase, liquid phase and gas phase are respectively represented. and These represent the mass fractions of water vapor and air, respectively. and These are the liquid phase and gas phase saturation, respectively. Porosity.

[0044] The boundary conditions for the heat transfer physical field are: Equation 22: Equation 23: Formula 24: Formula 25: In the formula, This represents the boundary flux (W / (m²)) on the surface of the kiwifruit slice. 2 ·s)); The ambient temperature (K) is the average of the water vapor temperature in the three cavity locations. The convective heat transfer coefficient (W / (m) 2 ·K)); Indicates the equivalent diameter (m) of the kiwi slice; Represents the Reynolds number; Represent Prandtl numbers; Indicates air velocity (m / s); Aerodynamic viscosity (Pa·m).

[0045] S6. Mesh Independence Check: Solve the model sequentially with ultra-coarsened, coarsened, refined, relatively refined, ultra-refined, and extremely refined tetrahedral meshes. When the difference between the simulated water content and electric field intensity is <0.1%, determine the optimal mesh density and select the ultra-refined tetrahedral mesh for this invention. S7. Model Solving A transient solver, a fully coupled and iterative solution mode, and the iterative solver was FGMRES. Mesh independence was checked during the simulation, and the model was solved sequentially under different mesh densities. The optimal mesh density was determined when the difference between the simulated moisture content and electric field intensity was <0.1%. This study selected an ultra-fine tetrahedral mesh with a time step of 1 min. The mesh generation diagram of the kiwifruit slice radio frequency vacuum drying physical model is shown below. Figure 3As shown. To balance computational accuracy and time, the relative tolerance is set to 1. Each run takes approximately 3 hours. The workstation configuration for running the software is: Dell T3660 processor, Intel CPU i9-13900K 3.00GHz, 128 GB of memory, and Windows Server 2019 R2 Standard 64-bit operating system.

[0046] S8, Model Validation The results of the radio frequency vacuum drying simulation of kiwi fruit slices were obtained, including the average dry basis moisture content after simulation and the temperature change curves of the corners, edges, and center. The coefficient of determination (COP) was used. ) and root mean square error ( The similarity between simulation and experimental results is assessed using the following formula: Equation 26: Equation 27: in, This represents the total number of experimental indicators over time. and These represent the experimental and predicted values, respectively. This is the average of the experimental values. When the average dry basis moisture content is... >0.95、 <1, three-point temperature The model is usable when the values ​​are >0.80 and the RMSE is <5 °C.

[0047] S9, Screening of Intelligent Control Methods The selected intelligent control method was implemented in the validated model, and the heating uniformity index before and after implementation was compared. Based on the energy consumption (kW·h), the optimal intelligent control method is selected. The formula for calculating heating uniformity is: Formula 28: In the formula, This indicates the average temperature (°C) of all the kiwi slices. This represents the total volume of all kiwi slices. Calculated in the software. The expression is abs( [1 / K]-273.15- ) / ( - ) / (a single kiwi slice) The expression for calculating energy consumption is ) / 1[m^3] / 24; In the software, the statement ` / 60000` sums the results after removing the first dot. Energy consumption is only compared in terms of numerical value; During the drying process, the actual and predicted values ​​of the average dry basis moisture content and temperature changes at three points (corner, edge, and center) of the material were compared. The values ​​are 0.98, 0.86, 0.85, and 0.88 respectively. The values ​​were 0.21 °C, 3.91 °C, 4.41 °C, and 3.89 °C, respectively, indicating that the model is usable.

[0048] S10. Based on the validated model, kiwi slices were dried under conditions of a vacuum of 0.015 MPa and a sample center-to-center distance of 66.75 mm as a control group.

[0049] S11. The intelligent control method for the experimental group is as follows: when the lowest material temperature is below the set target temperature range (59.50 °C), the radio frequency heating system starts to provide a constant voltage; when the highest material temperature exceeds the target temperature range (60.50 °C), the radio frequency heating system stops. The physical model is designed for... control logic for: In the formula, The minimum temperature of all kiwi slice samples is represented by the minop operator, which is obtained by constructing a nonlocal coupling. The maximum temperature of all kiwi slice samples is represented by the maxop operator constructed through nonlocal coupling. In software, this is expressed as: if(minop1( )<=332.65[K],ec.Qh,if(maxop1( )>=333.65[K],0,ec.Qh)).

[0050] S12. Calculate the control group and experimental group. And energy consumption.

[0051] Example 2 The difference from Example 1 lies in the intelligent control method in step eleven: when the maximum temperature of all 24 centers of the material is lower than the set target temperature range (59.50 °C), the radio frequency heating system starts to provide a constant voltage; when it exceeds the target temperature range (60.50 °C), the radio frequency heating system stops. The physical model is designed for... control logic for: In the formula, The maximum temperature value is represented by the 24 centers of all kiwi fruit samples. It is obtained by constructing the maxop operator through nonlocal coupling and selecting 24 center points. In software, this is expressed as: if(maxop1( )<=332.65[K],ec.Qh,if(maxop1( )>=333.65[K],0,ec.Qh)).

[0052] Example 3 The difference from Example 1 lies in the intelligent control method in step eleven: when the lowest material temperature is lower than the set target temperature range (59.50 °C) and the maximum temperature at the center of the material at the three points is lower than the set target temperature range (59.50 °C), the radio frequency heating system starts to provide a constant voltage; when the highest material temperature exceeds the target temperature range (60.50 °C) and the maximum temperature at the center of the material at the three points exceeds the set target temperature range (60.50 °C), the radio frequency heating system stops. The physical model is designed for... control logic for: In software, this is expressed as: if((minop1( )<=332.65[K])&&(max(point1,max(point2,point3))<=332.65[K]),ec.Qh,if((maxop1( )>=333.65[K])&&(max(point1,max(point2,point3))>=333.65[K]),0,ec.Qh)).

[0053] Comparative Example Calculate the mathematical model according to step 10 of Example 1.

[0054] Test case The determination of Examples 1-3 and Comparative Examples And energy consumption, the results are shown in Figure 4 ( (Comparison results) and Figure 5 (Energy consumption comparison results).

[0055] Depend on Figure 4It can be seen that, from the perspective of heating uniformity (Examples 1, 2, and 3), the newly proposed intelligent control method for the maximum temperature of all 24 centers of the samples (Example 2) exhibits the best heating uniformity. Compared with the comparative example, the newly proposed intelligent control method for the minimum / maximum temperature of all samples (Example 1) has poor heating uniformity, and this method is not suitable for intelligent control of RF vacuum drying of kiwi slices. Compared with the comparative example, the proposed intelligent control method for the minimum / maximum temperature of all samples and the maximum temperature at three points (Example 3) maintains the same heating uniformity, making it unnecessary to adopt this intelligent control method for RF vacuum drying of kiwi slices. Figure 5 It can be seen that, from the perspective of energy consumption (Examples 1, 2, and 3), the newly proposed method for intelligently controlling the minimum / maximum temperatures of all samples (Example 1) has the lowest energy consumption. Compared with the comparative example, the newly proposed method for intelligently controlling the maximum temperature of all 24 centers of all samples (Example 2) and the method for controlling the minimum / maximum temperatures of all samples and the maximum temperature at three points (Example 3) have the same energy consumption as the comparative example. In summary... In terms of energy consumption, the intelligent control method for radio frequency vacuum drying of kiwi slices selected in Example 2 is the most suitable.

[0056] In summary, the intelligent control method for radio frequency vacuum drying of kiwifruit slices based on numerical simulation optimization of this invention utilizes COMSOL Multiphysics® (v5.6) software to construct a multiphysics numerical model incorporating Joule heating and rarefaction mass transfer. Simulations and experiments show that the average dry basis moisture content and temperature changes at key monitoring points (corners, edges, and center) of the kiwifruit slices during the radio frequency vacuum drying process meet the requirements of the average dry basis moisture content. >0.95、 <1, three-point temperature >0.80 respectively With temperatures below 5 °C, this model can effectively guide actual drying processes. Examples of the three intelligent control methods can be quantified and implemented in the model, with the optimal intelligent control method being the maximum temperature at all 24 centers across all samples. This invention enables rapid and low-cost screening of the optimal intelligent control method, effectively saving costs and improving efficiency.

[0057] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for intelligent control of radio frequency vacuum drying of kiwifruit slices based on numerical simulation optimization, characterized in that, Includes the following steps: S1. Sample pretreatment: Select fresh, undamaged "Xuxiang" kiwifruit, peel them, and cut them into thin slices with a thickness of 7.60 ± 0.20 mm, a major axis of 5.25 ± 0.30 cm, and a minor axis of 4.17 ± 0.24 cm. Take 24 slices and lay them evenly in a rectangular tray of 400 mm × 270 mm × 20 mm, with a center-to-center distance of 65 mm between adjacent samples. S2. Actual drying experiment: The tray was placed in the center of the lower plate of a 6 kW radio frequency vacuum drying device. Three fiber optic temperature sensors were inserted in the corners, edges and center of the kiwi slices. Drying was carried out at a drying temperature of 60 ± 0.5 °C, a vacuum degree of 0.023 ± 0.003 MPa and a plate spacing of 75 mm. Drying was completed when the dry basis moisture content was lower than 0.18 g / g. The weight was measured and the moisture content was calculated every 30 minutes. At the same time, the temperature change of water vapor in the chamber was measured. S3. Constructing a physical model: Based on the finite element method, using COMSOL Multiphysics® (v5.6) software, a physical model of radio frequency vacuum drying, including Joule heating and rare mass transfer modules, is established according to the dimensions and control logic of the kiwi slices and drying equipment, and model assumptions are set. S4. Constructing a geometric model: Based on the actual dimensions of the kiwi slices and the radio frequency vacuum drying equipment, a geometric model consistent with the actual working conditions is constructed using COMSOL Multiphysics® (v5.6) software; S5. Define governing equations and boundary conditions: Establish governing equations for electromagnetic field, mass transfer, heat transfer, and energy conservation; define corresponding boundary conditions; and define electromagnetic energy control logic. ; S6. Mesh independence test: Solve the model under different mesh densities in turn. When the difference between the simulated values ​​of water content and electric field strength is <0.1%, determine the optimal mesh density. S7. Model Solving: A transient solver, fully coupled and iterative solving mode are adopted, and an ultra-fine tetrahedral mesh is selected. The model is solved with a time step of 1 minute. S8. Model Validation: Using the coefficient of determination R... 2 The root mean square error (RMSE) is used to assess the similarity between simulation and experimental results, when the average dry basis moisture content R... 2 >0.95, RMSE<1, three-point temperature R 2 The model is usable when the values ​​are >0.80 and RMSE are <5 °C, respectively. S9. Screening of intelligent control methods: Implement multiple intelligent control methods in the validated model, compare the heating uniformity index TUI and energy consumption before and after implementation, and select the optimal intelligent control method.

2. The method according to claim 1, characterized in that, In step S1, the initial wet basis moisture content of "Xu Xiang" kiwifruit was 83.30 ± 0.01%, and the soluble solids content was 14.83 ± 0.49 °Brix.

3. The method according to claim 1, characterized in that, The moisture content in step S2 is calculated using the following formula: , ,in This is the moisture content on a wet basis. Moisture content on a dry basis. for The quality of kiwi slices at all times This is for absolute dry material quality.

4. The method according to claim 1, characterized in that, The model assumptions in step S3 include: the air temperature inside the radio frequency cavity is constant; the solid, liquid, and gas phases are continuous and share gas pressure; the three phases inside the kiwi slice are in local thermal equilibrium and are isotropic porous media; volume evaporation is the main process and there is no mass flux on the surface; thermal expansion and contraction, gravity, container and vacuum fluctuations, and sensor response time are ignored.

5. The method according to claim 1, characterized in that, In step S4, the governing equations for the electromagnetic field are simplified Laplace's equations, and the boundary conditions include electrical insulation of the RF external cavity wall, an upper plate voltage of 5400 V, and a lower plate grounded; electromagnetic energy control logic. The on / off state of electromagnetic energy is controlled by a temperature threshold.

6. The method according to claim 1, characterized in that, In step S4, the mass transfer control equation is based on the law of conservation of mass, while the heat transfer and energy conservation control equations consider effective density, effective specific heat capacity, effective thermal conductivity, and total energy exchange.

7. The method according to claim 1, characterized in that, Heating uniformity index in step S7 Through formula Calculations show that energy consumption is determined by volume fraction. Statement calculation.

8. The method according to claim 1, characterized in that, The intelligent control method in step S7 includes control logic based on the maximum temperature at three points, the minimum / maximum temperature of all samples, the maximum temperature of 24 centers of all samples, and the combination of the minimum / maximum temperature of all samples and the maximum temperature at three points.

Images