Multi-physical field coupling calculation and drying method considering random characteristics of materials

Abstract

The application claims a kind of multi-physical field coupling calculation and drying method considering random characteristics of material, belong to microwave heating simulation calculation technical field, it includes the following steps: step one: the particle size data of actual material is collected, and random particle geometric model based on Monte Carlo algorithm is constructed;Step two: establish multi-physical field coupling calculation model;Step three: the calculation domain is discretized using finite element method, and electromagnetic field is solved using frequency domain method;Step four: process parameter optimization and fuzzy PID control step based on random model are carried out.The application is based on Monte Carlo algorithm, combined with the dynamic relationship between dielectric properties and moisture content determined by experiment, a high coupling degree multi-physical field calculation model reflecting the influence of randomness of granular material on electromagnetic field, temperature field and moisture field is constructed, so as to accurately reproduce the localization hot spot, dry and wet patch and other phenomena in the drying process, and significantly improve the calculation accuracy.Finally, efficient and high-quality drying of material is realized.

Description

Technical Field

[0001] This invention belongs to the field of microwave heating simulation calculation technology, specifically to a multi-physics field coupling calculation and drying method that considers the random characteristics of materials. Background Technology

[0002] In actual production and processing, the size distribution, random stacking position, and initial humidity of particulate materials are usually random. Existing approximate homogeneous block models and regular uniform particle models cannot characterize the influence of the above randomness on the electromagnetic field, temperature field, and moisture field, resulting in a large deviation between the calculated and actual results.

[0003] This invention, based on the Monte Carlo algorithm and combined with experimentally determined dynamic relationships between dielectric properties and moisture content, constructs a highly coupled multiphysics computational model that reflects the influence of the randomness of particulate materials on electromagnetic, temperature, and moisture fields. This model accurately reproduces phenomena such as localized hot spots and wet / dry patches during the drying process, significantly improving computational accuracy. Furthermore, based on the computational results, a fuzzy PID algorithm is used to dynamically adjust the microwave-hot air combined action mode, ultimately achieving efficient and high-quality drying of the material.

[0004] A search revealed application CN118412050A, entitled "A Microwave Heating Simulation Calculation Method Based on Multiphysics Coupling Strength," belonging to the field of microwave heating simulation calculation technology. The method includes: establishing an electromagnetic field calculation model and a temperature field calculation model based on the geometric model and physical properties of the heated medium; solving for the coupling strength and relative rate of change of the electromagnetic and thermal fields; comparing the calculation results with a pre-set threshold; and optimizing the multi-timescale factors of the multiphysics calculation if the calculation results do not meet the threshold until the calculation results meet the threshold. This invention links the relative rate of change of coupling strength with multi-timescale factors, which can eliminate the problem of changes in the coupling relationship between multiphysics equations in microwave-heated chemical reactions caused by large changes in the dielectric properties of the chemical reaction process over time, improving the accuracy of multiphysics calculations and enhancing the scientific rigor and reliability of microwave-heated chemical reaction simulations. Summary of the Invention

[0005] This invention aims to solve the problems of the prior art mentioned above. It proposes a multiphysics coupling calculation and drying method that considers the random characteristics of materials. The technical solution of this invention is as follows:

[0006] A multiphysics coupling calculation and drying method considering the random properties of materials includes the following steps:

[0007] Step 1: Collect particle size data of actual materials and construct a random particle geometry model based on the Monte Carlo algorithm;

[0008] Step 2: Establish a multiphysics coupling calculation model;

[0009] Step 3: Discretize the computational domain using the finite element method. The computational domain refers to the entire spatial range where the electromagnetic field problem needs to be solved, including all medium regions, field sources, and boundaries where boundary conditions are applied. Solve the electromagnetic field using the frequency domain method.

[0010] Step 4: Optimize process parameters and implement fuzzy PID control based on a stochastic model.

[0011] Furthermore, step one, constructing a stochastic particle geometry model based on the Monte Carlo algorithm, specifically includes:

[0012] A1) Collect particle size data of actual materials, including major and minor axes, and statistically analyze their distribution patterns;

[0013] A2) A statistically representative population of particles is generated using the Monte Carlo random algorithm, with the particle morphology simplified to an approximate ellipsoid. The steps of the Monte Carlo random algorithm are as follows: First, the problem to be solved is transformed into the expected or probabilistic form of a certain random variable, and a suitable probability model (such as uniform distribution) is constructed within the domain. Then, a large number of random samples are generated independently, and the objective function value is calculated or judged for each sample. Then, an approximate solution to the problem is obtained through the sample mean or proportion. Finally, the confidence interval or relative error of the result is estimated using the sample variance.

[0014] A3) The random sequential adsorption algorithm (RSA) is used to generate the spatial position and orientation of particles in the loading area of ​​the drying chamber, and a collision detection algorithm is used to avoid overlapping between particles, simulating a real loose stacking state.

[0015] A4) The initial wet basis moisture content of each particle is randomly assigned according to a normal distribution within the measured range to reflect the natural fluctuation of the initial moisture content;

[0016] A5) At the same time, two traditional ideal models are established for comparison: the approximate homogeneous block model and the regular uniform particle model. The approximate homogeneous block model regards the entire material layer as a continuous, uniform, porous medium block; the regular uniform particle model is that the particles are of uniform size, regularly and tightly arranged, and have absolutely uniform initial temperature and moisture content.

[0017] Furthermore, step 3) utilizes the Random Sequential Adsorption (RSA) algorithm to generate the spatial position and orientation of particles within the loading area of ​​the drying chamber, and employs a collision detection algorithm to avoid particle overlap, simulating a real loosely packed state. Specifically, this includes:

[0018] The steps of the Random Sequential Adsorption (RSA) algorithm for generating the spatial position and orientation of particles within the loading area of ​​the drying chamber are as follows: First, a center position and spatial orientation (e.g., Euler angles) of a particle are randomly generated within the loading area. Then, it is checked whether the particle overlaps with all the particles already placed and the boundary of the area. If the non-overlap condition is met, the particle is adsorbed and fixed. Otherwise, the attempt is discarded and a new random position and orientation are regenerated. This process is repeated until the target number of particles is reached or no more particles can be placed.

[0019] The steps for simulating a real loosely packed state using a collision detection algorithm are as follows: First, each particle is assigned a random initial position and orientation (potentially penetrating each other). Then, the geometric overlap between all particle pairs and between particles and boundaries is iteratively detected. When an overlap is detected, the displacement or velocity correction amount that separates the particles is calculated based on physical laws (such as the contact force model of normal and tangential directions). The position and orientation of all particles are updated sequentially, and global collision detection and response updates are repeated until the overlap of the entire system is less than the allowable threshold or an equilibrium state is reached, thus obtaining a naturally loosely packed configuration.

[0020] Furthermore, step two: establishing a multiphysics coupling calculation model specifically includes:

[0021] The dielectric properties were first measured using an open coaxial probe reflection method; the moisture content of the sample was determined immediately after each measurement, and the measurement was repeated 5 times for each moisture content level.

[0022] Data was processed using Origin software to establish the real part of the relative permittivity. virtual part A fitting model between the dielectric constant and the moisture content W was developed. The dielectric properties of the material were measured and characterized.

[0023] Next, the generated random material geometry model is imported into the cavity environment, and a complete three-dimensional geometric model is established in COMSOL Multiphysics software based on the actual geometric dimensions of the microwave-hot air drying experimental system; the main physical equations are as follows:

[0024] B1) Electromagnetic field equations: Using Maxwell's equations in the frequency domain, the governing equations are as follows:

[0025] ;

[0026] Vector differential operator, Indicates electric field strength (V / m); It is the relative permeability; It is the complex relative permittivity, which is related to the moisture content and temperature of the material; Electrical conductivity (S / m); It is the free space wavenumber. It is the vacuum permittivity. ω is the vacuum permeability, ω=2πf is the angular frequency (rad / s), and f=2450MHz.

[0027] The TE10 mode excitation is set at the waveguide inlet, and the waveguide and the inner wall of the cavity are set as ideal electrical conductors (PEC).

[0028] B2) Heat transfer equation: Considering electromagnetic loss as a heat source, and combining energy conservation and Fourier's law, a transient heat conduction equation is established:

[0029] ;

[0030] in Indicates dielectric density (Kg / m³) 3 ); Specific heat capacity (J / (kg)) K)); Represents the thermal conductivity coefficient (W) m -1 K -1 ); Indicates the real-time temperature (K) of the medium; Represents the fluid velocity vector (m / s); This represents the heat source term (W / m²) generated by external sources such as electromagnetic fields. 3 ); The latent heat of vaporization of water (J / kg); Evaporation rate (kg) m -3 s -1 ).

[0031] The electromagnetic heat source Q is calculated from the electric field strength and the dielectric loss factor: In the formula It is microwave volumetric heat. It is a microwave frequency. It is the dielectric loss factor. Electrical conductivity (S / m);

[0032] B3) Mass transfer equations: Liquid water migration within porous media follows Darcy's law, water vapor diffusion follows Fick's law, and the total water transport equation is:

[0033] ;

[0034] In the formula Total moisture content; K is the Darcy velocity of liquid water (m / s), and K is the permeability. It is the viscosity of the liquid (Pa). s), It is the liquid pressure (Pa); Steam diffusion flux (kg) m -2 s -1 ); and The densities of liquid water and moist air, respectively (kg / m³). 3 ); It is the effective moisture diffusion coefficient (m 2 / s).

[0035] B4) Fluid flow equation: The air flow within the cavity satisfies the Navier-Stokes equation;

[0036] B5) Set boundary conditions: TE10 excitation at the wave inlet, ideal electrical conductor on the cavity wall; convective heat transfer boundary between material and air; no flux boundary at the bottom, and mass transfer boundary on the remaining surfaces.

[0037] Furthermore, step three: discretizing the computational domain using the finite element method and solving the electromagnetic field using the frequency domain method, specifically includes:

[0038] The computational domain is discretized using the finite element method. Local mesh refinement is performed on key areas including the material domain, waveguide, and air inlet / outlet, and boundary layer mesh is added to the wall surface.

[0039] The electromagnetic field is solved using the frequency domain method. The steady-state solver calculates the steady-state distribution of the electromagnetic field and the flow, while the transient solver calculates the evolution of the temperature and humidity fields over time.

[0040] The heat generation term calculated by electromagnetic waves is coupled to the heat transfer equation, and the temperature change is fed back to the dielectric properties and fluid properties.

[0041] After the calculations are completed, the temperature field, electric field distribution, moisture content distribution, and drying kinetic curves are extracted and compared with the experimental results. The root mean square error (RMSE) is calculated to verify the model accuracy.

[0042] Furthermore, the discretization of the computational domain using the finite element method specifically includes:

[0043] The loading area (computational domain) of the drying chamber is divided into a finite number of non-overlapping micro-cells (such as tetrahedral or hexahedral meshes). Within each cell, the unknown electric field is... Using the values ​​at the nodes and the shape function N i The vector wave equation is approximated, and then the weighted residual method is used to transform the vector wave equation into a discrete system of algebraic equations.

[0044] The electromagnetic field is solved using the frequency domain method. The steady-state solver calculates the steady-state distribution of the electromagnetic field and the flow, while the transient solver calculates the evolution of the temperature and humidity fields over time.

[0045] First, the electromagnetic field is solved using the frequency domain method. This involves assuming the time-harmonic electromagnetic field is in complex amplitude form and substituting it into Maxwell's equations to obtain the electric field... The frequency domain Helmholtz wave equation is discretized and the complex linear equations are solved using the finite element method to obtain the steady-state distribution of the electric field amplitude and phase in the computational domain. Then, the steady-state solver is activated to simultaneously solve the steady-state flow control equations (continuity equation, momentum equation, and energy equation) based on the known electromagnetic field distribution and the boundary conditions of the drying cavity, obtaining the steady-state distribution of the air velocity, pressure, and temperature fields. At this point, neither the electromagnetic field nor the flow field changes with time. Finally, the transient solver is invoked, using the microwave heat source term provided by the electromagnetic field. Using the convective heat transfer and mass transfer coefficients provided by the steady-state flow field as inputs, the transient heat conduction equation and moisture diffusion equation inside the material are solved step by step in the time domain, and the temperature field T is updated at each time step. ,t) and humidity field W( The electromagnetic field distribution is adjusted based on the dielectric parameters of the material, which are dependent on temperature and humidity, and this process is repeated until the preset drying time or moisture content target is reached.

[0046] Furthermore, step four: optimizing process parameters and implementing fuzzy PID control based on a stochastic model specifically includes:

[0047] C1) Process parameter optimization: Determine the optimal combination of process parameters through orthogonal experiments or response surface methodology; the stochastic model is a model with randomness in the size distribution of particulate materials, random stacking location, and initial humidity.

[0048] C2) Fuzzy PID Control: Based on the stochastic model, a fuzzy PID controller is designed for real-time control of the drying process. The specific method is as follows:

[0049] The deviation e between the highest internal temperature of the material and the target temperature and the rate of change of the deviation ec are used as input variables, and the adjustment of PID control parameters (Kp, Ki, Kd) are used as output variables.

[0050] A fuzzy rule base is established, and the real-time PID parameter correction is obtained through fuzzy inference and defuzzification;

[0051] A fuzzy PID controller is embedded in a multiphysics coupling model to simulate the drying process under closed-loop control.

[0052] A storage device includes a processor and a memory, the memory storing a computer program, the processor executing steps of a multiphysics coupling calculation and drying method considering the random properties of materials as described in any one of the claims by invoking the computer program stored in the memory.

[0053] A computer-readable storage medium for storing a computer program for blockchain-based movable property pledge financing, wherein the computer program, when run on a computer, performs the steps of the multiphysics coupling calculation and drying method considering the random characteristics of materials as described in any one of the claims.

[0054] The advantages and beneficial effects of this invention are as follows:

[0055] The innovation of this invention is mainly reflected in the steps or formulas corresponding to claims 1, 2, 3, 4, and 5. Its core breakthrough lies in the first systematic introduction of the triple randomness at the particle scale into a microwave hot air multiphysics coupling model, and the establishment of a dynamic dielectric property feedback and a general framework across drying methods. The following analyzes its innovativeness and non-obviousness:

[0056] 1. The hindrance of traditional technological inertia

[0057] Industry practice: The existing microwave drying simulation field has long adopted the homogenization assumption (treating the material as a continuous medium) or the regular arrangement of particles assumption (ignoring randomness to simplify calculations).

[0058] Cognitive bias: Engineers generally believe that "the randomness of particles has a negligible effect on the macroscopic temperature field" or "can be replaced by average parameters," lacking an understanding of the physical mechanism by which random stacking leads to localized resonance from the perspective of electromagnetic wave theory.

[0059] Computational cost concerns: stochastic modeling requires generating a large number of particles and solving multi-scale coupled problems. Those skilled in the art usually prefer to choose a low-cost, simplified model rather than actively increasing complexity.

[0060] 2. The complexity of the technical challenges

[0061] Deep coupling of multiphysics and stochastic geometry: The conventional approach is to separate the solution of stochastic geometry from that of physical fields (e.g., first build the stochastic model, then solve the electromagnetic field in a single step). This invention achieves real-time dynamic feedback: dielectric properties change with water content → electromagnetic field redistribution → heat source change → water evaporation → dielectric properties change again. This two-way coupling is computationally extremely expensive under stochastic particle geometry and requires an accurate dielectric-water content dynamic model. Furthermore, high-order fitting itself requires a large amount of experimental data and cannot be replaced by simple linear regression.

[0062] Prediction of localized hotspots: Traditional models assume that the electric field distribution can be smoothed using the average dielectric constant, failing to anticipate the localized electric field enhancement (hotspots) generated by multiple scattering and interference between random particles that far exceed the average value. This invention verifies this through experiments. Figure 4 This reveals a phenomenon that is easily overlooked in conventional theoretical derivations.

[0063] 3. The versatility across drying methods is not a conventional extension.

[0064] Those skilled in the art typically develop separate numerical models for each drying method (microwave, hot air, infrared, etc.) because the heat source forms differ greatly (volume heating vs. surface heating). This invention proposes a modular strategy that retains the random particle geometry and the core equations of heat and mass transfer, replacing only the heat source terms. This requires a deep understanding of the common structure of multi-physics coupling, rather than simply "replacing boundary conditions."

[0065] 4. The introduction of fuzzy PID control goes beyond the scope of conventional simulation.

[0066] Conventional drying simulations stop at open-loop process optimization (such as obtaining fixed parameters through orthogonal experiments). This invention further combines stochastic models with closed-loop fuzzy PID control to verify the control strategy's effectiveness in suppressing localized hotspots in a virtual environment. This integrated simulation-control approach is rare in the field of agricultural product drying and typically requires interdisciplinary knowledge of control theory and heat transfer.

[0067] This invention is not a simple improvement on existing models, but a full-chain innovation from physical cognition (randomness cannot be ignored) to modeling methods (dynamic coupling) and application boundaries (cross-mode universality, closed-loop control), which has significant creativity. Attached Figure Description

[0068] Figure 1 This is a flowchart of the construction process of a random particle geometry model according to a preferred embodiment of the present invention;

[0069] Figure 2 This is a schematic diagram of the dielectric properties measurement of mulberries;

[0070] Figure 3 This is a schematic diagram of a microwave-hot air drying system.

[0071] Figure 4 This is a temperature distribution diagram for microwave-hot air drying over 60 seconds. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0073] The technical solution of the present invention to solve the above-mentioned technical problems is:

[0074] This invention provides a multiphysics field coupling calculation and drying method that considers the random characteristics of materials, including the following steps:

[0075] Step 1: Construct a stochastic particle geometry model based on the Monte Carlo algorithm

[0076] 1) Collect particle size data (major axis, minor axis, etc.) of actual materials and statistically analyze their distribution patterns, which usually follow a log-normal distribution.

[0077] 2) A statistically representative particle population is generated using the Monte Carlo random algorithm, and the particle morphology is simplified to an approximate ellipsoid.

[0078] 3) The spatial position and orientation of particles in the loading area of ​​the drying chamber are generated using the random sequential adsorption algorithm (RSA), and a collision detection algorithm is used to avoid overlapping between particles, simulating a real loose stacking state.

[0079] 4) The initial wet basis moisture content of each particle is randomly assigned according to a normal distribution within the measured range to reflect the natural fluctuation of the initial moisture content.

[0080] 5) At the same time, two traditional ideal models are established for comparison: the approximate homogeneous block model (which regards the entire material layer as a continuous, uniform, porous medium block) and the regular uniform particle model (the particles are of uniform size, regularly and tightly arranged, and have absolutely uniform initial temperature and moisture content).

[0081] Step 2: Establish a multiphysics coupling calculation model

[0082] Using mulberry fruit as the research object, the dielectric properties (dielectric constant ε′ and loss factor ε″) were first measured using an open coaxial probe reflection method. The system consisted of a vector network analyzer (Agilent E8363C) and a dielectric probe kit (Keysight N1501A). Before measurement, the sample was calibrated sequentially using air and 25°C deionized water. The material was filled into a 50mL beaker, and the probe was fixed so that its end face was gently pressed vertically against the sample surface. Frequency sweep measurements were performed in the 2GHz–3GHz range, with a focus on recording data at 2450 MHz. The sample moisture content was determined immediately after each measurement, and each moisture content level was measured five times.

[0083] Data was processed using Origin software to establish the real part of the relative permittivity. virtual part A fitting model between the dielectric constant and the moisture content W was developed. The dielectric properties of the material were measured and characterized.

[0084] Next, the generated random material geometry model was imported into the cavity environment, and a complete three-dimensional geometric model was established in COMSOL Multiphysics software based on the actual geometry of the microwave-hot air drying experimental system (including the microwave resonant cavity, rectangular waveguide, and hot air inlet / outlet). The main physical equations are as follows:

[0085] 1) Electromagnetic field equations: Using Maxwell's equations in the frequency domain, the governing equations are as follows:

[0086] ;

[0087] 2) Set the TE10 mode excitation at the waveguide inlet, and set the waveguide and cavity inner wall as ideal electrical conductors (PEC).

[0088] 2) Heat transfer equation: Considering electromagnetic loss as a heat source term, and combining energy conservation and Fourier's law, a transient heat conduction equation is established:

[0089] ;

[0090] The electromagnetic heat source Q is calculated from the electric field strength and the dielectric loss factor:

[0091] ;

[0092] 3) Mass transfer equations: The migration of liquid water inside porous media follows Darcy's law, and the diffusion of water vapor follows Fick's law. The total water transport equation is:

[0093]

[0094] 4) Fluid flow equation: The air flow inside the cavity satisfies the Navier-Stokes equation.

[0095] 5) Set boundary conditions: TE10 excitation at the wave inlet, ideal electrical conductor on the cavity wall; convective heat transfer boundary between material and air; no flux boundary at the bottom, and mass transfer boundary on the remaining surfaces.

[0096] Step 3: Numerical Solution and Result Post-processing

[0097] The computational domain is discretized using the finite element method. Local mesh refinement is performed in key areas (material domain, waveguide, air inlet and outlet), and boundary layer meshes are added to the walls.

[0098] The electromagnetic field is solved using the frequency domain method. The steady-state solver calculates the steady-state distribution of the electromagnetic field and the flow, while the transient solver calculates the evolution of the temperature and humidity fields over time.

[0099] The heat generation term calculated by electromagnetic waves is coupled to the heat transfer equation, and the temperature change is fed back to the dielectric properties and fluid properties.

[0100] After the calculations are completed, the temperature field, electric field distribution, moisture content distribution, and drying kinetic curves are extracted and compared with the experimental results. The root mean square error (RMSE) is calculated to verify the model accuracy.

[0101] Step 4: Optimize process parameters and implement fuzzy PID control based on a stochastic model.

[0102] 1) Process parameter optimization: Investigate the effects of different microwave power, hot air temperature, wind speed, and other parameters on drying uniformity, drying rate, and energy consumption. Optimal process parameter combinations will be determined through orthogonal experiments or response surface methodology, providing theoretical guidance for actual production.

[0103] 2) Fuzzy PID Control: Based on the above stochastic model, a fuzzy PID controller is designed for real-time control of the drying process. The specific method is as follows:

[0104] 1. The deviation e between the highest internal temperature of the material and the target temperature and the rate of change of the deviation ec are used as input variables, and the adjustment of the PID control parameters (Kp, Ki, Kd) are used as output variables.

[0105] 2. Establish a fuzzy rule base (e.g., if e is positive and ec is positive, then Kp increases and Ki decreases, etc.), and obtain the real-time PID parameter correction through fuzzy inference and defuzzification.

[0106] 3. Embed the fuzzy PID controller into a multiphysics coupling model to simulate the drying process under closed-loop control.

[0107] 4. Compared with conventional PID control, fuzzy PID can suppress temperature overshoot more quickly, reduce the duration of localized hot spots, and improve drying uniformity.

[0108] 5. This control strategy can be verified through virtual simulation based on the random particle model of this invention, and then transferred to actual drying equipment.

[0109] Figure 1 This is a flowchart illustrating the process of generating a random particle model using the Monte Carlo algorithm. The diagram shows the sequence: inputting actual particle size statistics → generating a log-normal distribution → random sequential adsorption to generate particle positions → collision detection → outputting a random stacking model. This diagram explains the complete construction process from actual statistical patterns to a 3D geometric model.

[0110] Figure 4 Temperature distribution diagrams for microwave-hot air drying over 60 seconds: (a) Homogeneous block model (b) Temperature distribution of regular uniform model (c) Random model (d) Actual temperature distribution.

[0111] Figure 3The temperature distribution cloud maps calculated by the three models after 60 seconds of drying are shown and compared with experimental infrared thermograms. The figures show that the homogeneous block model exhibits a symmetrical gradual change; the regular uniform model shows a periodic alternation of strong and weak intensity; and the random model displays an alternating distribution of "hot spots" and "cold zones," with the hot spot locations consistent with experimental observations. This figure verifies the accuracy of the random model in temperature field simulation.

[0112] 1. This invention is the first to simultaneously incorporate the triple randomness of particle size distribution, random stacking location, and initial humidity fluctuation into microwave-hot air multiphysics field coupling calculation.

[0113] 2. By comparing three models (homogeneous block, regular uniform, and random), the mechanism by which random structures regulate electric field localization, temperature field non-uniformity, and moisture field patchiness was revealed.

[0114] 3. The method of this invention can be integrated into commercial software such as COMSOL Multiphysics to form a simulation module specifically for microwave-hot air drying of random materials.

[0115] 4. In step four, fuzzy PID control is introduced to perform intelligent control simulation and optimization of the drying process based on a stochastic model.

[0116] 5. This stochastic modeling and coupled computation method has versatility across drying methods. It is not only applicable to microwave-hot air combined drying, but can also be extended to other drying technologies such as pure microwave, pure hot air, microwave-vacuum, infrared-hot air, and radio frequency drying.

[0117] 1. A method for constructing a random particle geometry model, including random particle size distribution, random spatial location, and random initial moisture content.

[0118] 2. A numerical simulation method that fully couples the above-mentioned stochastic geometric model with electromagnetic fields, temperature fields, humidity fields, and flow fields.

[0119] 3. The drying process parameters are optimized using this coupled model, and the optimization design is achieved by adopting a fuzzy PID control strategy.

[0120] 4. The general application of the method in different drying methods (microwave-hot air, pure microwave, pure hot air, microwave-vacuum, infrared-hot air, radio frequency drying, etc.).

[0121] 1. It should be noted that alternatives to the random generation algorithm include: in addition to Monte Carlo + RSA, the Discrete Element Method (DEM) can be used to simulate particle packing, or a random perturbation method can be used to apply random displacement to regularly arranged particles.

[0122] 2. Other alternatives for particle shape include: ellipsoids can be replaced with spheres, polyhedra, or irregular convex hulls, as long as the statistical size distribution is maintained.

[0123] 3. Alternatives for multiphysics coupling software include: in addition to COMSOL Multiphysics, ANSYS Maxwell + Fluent co-simulation or open-source software (such as OpenFOAM + Moose) can be used to achieve similar coupling.

[0124] 4. Alternatives to randomizing initial conditions include: the initial moisture content can be randomly assigned according to the measured distribution, or a random field method (such as a Gaussian random field) can be used to describe the continuous spatial fluctuations.

[0125] 5. Alternatives to process optimization methods include: orthogonal experiments or response surface methodology can be replaced by intelligent optimization algorithms such as genetic algorithms and particle swarm optimization.

[0126] 6. Other alternative control strategies include: fuzzy PID can be replaced by adaptive PID, neural network PID, or model predictive control (MPC).

[0127] 7. In terms of application areas, alternative solutions also include applying the method of the present invention to other drying equipment configurations (such as continuous microwave dryers, rotary drum dryers, etc.), requiring only adjustments to the geometric model and boundary conditions.

[0128] This invention was experimentally verified using mulberries as an example. Compared with two traditional models, this model reduced the root mean square error (RMSE) of temperature by 36.1% and 53.5%, respectively, and the RMSE of humidity content by 76.9% and 52.0%, respectively, fully demonstrating its advantages. This method can be extended to other agricultural products (such as berries, grains, beans, etc.), requiring only the dielectric properties-moisture content relationship, particle size statistical distribution, and other relevant simulation parameters of the corresponding materials.

[0129] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0130] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0131] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0132] The above embodiments should be understood as illustrative only and not as limiting the scope of protection of the present invention. After reading the description of the present invention, those skilled in the art can make various alterations or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A multiphysics coupling calculation and drying method considering the random characteristics of materials, characterized in that, Includes the following steps: Step 1: Collect particle size data of actual materials and construct a random particle geometry model based on the Monte Carlo algorithm; Step 2: Establish a multiphysics coupling calculation model; Step 3: Discretize the computational domain using the finite element method. The computational domain refers to the entire spatial range in which the electromagnetic field problem needs to be solved, including all medium regions, field sources, and boundaries where boundary conditions are applied; solve the electromagnetic field using the frequency domain method. Step 4: Optimize process parameters and implement fuzzy PID control based on a stochastic model.

2. The multiphysics coupling calculation and drying method considering the random characteristics of materials according to claim 1, characterized in that, Step one: Constructing a random particle geometry model based on the Monte Carlo algorithm, specifically including: A1) Collect particle size data of actual materials, including major and minor axes, and statistically analyze their distribution patterns; A2) A statistically representative particle population is generated using the Monte Carlo random algorithm, and the particle morphology is simplified to an approximate ellipsoid. The steps of the Monte Carlo random algorithm are as follows: First, the problem to be solved is transformed into the expected or probabilistic form of a certain random variable, and a suitable probability model is constructed within the domain. Then, a large number of random samples are generated independently, the objective function value is calculated or a judgment is made for each sample, and then an approximate solution to the problem is obtained through the sample mean or proportion. Finally, the confidence interval or relative error of the result is estimated using the sample variance. A3) The random sequential adsorption algorithm (RSA) is used to generate the spatial position and orientation of particles in the loading area of ​​the drying chamber, and a collision detection algorithm is used to avoid overlapping between particles, simulating a real loose stacking state. A4) The initial wet basis moisture content of each particle is randomly assigned according to a normal distribution within the measured range to reflect the natural fluctuation of the initial moisture content; A5) At the same time, two traditional ideal models are established for comparison: the approximate homogeneous block model and the regular uniform particle model. The approximate homogeneous block model regards the entire material layer as a continuous, uniform, porous medium block; the regular uniform particle model is that the particles are of uniform size, regularly and tightly arranged, and have absolutely uniform initial temperature and moisture content.

3. The multiphysics coupling calculation and drying method considering the random characteristics of materials according to claim 2, characterized in that, The third step involves using the Random Sequential Adsorption (RSA) algorithm to generate the spatial position and orientation of particles within the loading area of ​​the drying chamber, and employing a collision detection algorithm to avoid particle overlap, simulating a real loosely packed state. Specifically, this includes: The random sequential adsorption algorithm (RSA) is used to generate the spatial position and orientation of particles within the loading area of ​​the drying chamber. Specifically, the center position and spatial orientation of a particle are randomly generated within the loading area. Then, it is checked whether the particle overlaps with all the particles already placed and the boundary of the area. If the non-overlap condition is met, the particle is adsorbed and fixed. Otherwise, the attempt is discarded and a new random position and orientation are generated. This process is repeated until the target number of particles is reached or no more particles can be placed. The steps for simulating a real loosely packed state using a collision detection algorithm are as follows: First, each particle is assigned a random initial position and orientation. Then, the geometric overlap between all particle pairs and between particles and boundaries is detected iteratively. When an overlap is detected, the displacement or velocity correction amount that separates the particles is calculated based on the contact force model of the normal and tangential directions according to physical laws. The position and orientation of all particles are updated sequentially. Global collision detection and response updates are repeated until the overlap of the entire system is less than the allowable threshold or an equilibrium state is reached, thus obtaining a naturally loosely packed configuration.

4. The multiphysics coupling calculation and drying method considering the random characteristics of materials according to claim 1, characterized in that, Step two, establishing a multiphysics coupling calculation model, specifically includes: The dielectric properties were first measured using an open coaxial probe reflection method; the moisture content of the sample was determined immediately after each measurement, and the measurement was repeated 5 times for each moisture content level. Data was processed using Origin software to establish the real part of the relative permittivity. virtual part A fitting model between the dielectric constant and the moisture content W was developed; the dielectric properties of the material were measured and characterized. Next, the generated random material geometry model is imported into the cavity environment, and a complete three-dimensional geometric model is established in COMSOL Multiphysics software based on the actual geometric dimensions of the microwave-hot air drying experimental system; the main physical equations are as follows: B1) Electromagnetic field equations: Using Maxwell's equations in the frequency domain, the governing equations are as follows: ； Vector differential operator, Indicates electric field strength (V / m); It is the relative permeability; It is the complex relative permittivity, which is related to the moisture content and temperature of the material; Electrical conductivity (S / m); It is the free space wavenumber. It is the vacuum permittivity. ω is the permeability of free space, ω=2πf is the angular frequency (rad / s), and f=2450MHz; The TE10 mode excitation is set at the waveguide inlet, and the waveguide and the inner wall of the cavity are set as ideal electrical conductors (PEC). B2) Heat transfer equation: Considering electromagnetic loss as a heat source, and combining energy conservation and Fourier's law, a transient heat conduction equation is established: ； in Indicates dielectric density (Kg / m³) 3 ); Specific heat capacity (J / (kg)) K)); Represents the thermal conductivity coefficient (W) m -1 K -1 ); Indicates the real-time temperature (K) of the medium; Represents the fluid velocity vector (m / s); This represents the heat source term (W / m²) generated by external sources such as electromagnetic fields. 3 ); The latent heat of vaporization of water (J / kg); Evaporation rate (kg) m -3 s -1 ); The electromagnetic heat source Q is calculated from the electric field strength and the dielectric loss factor: In the formula It is microwave volumetric heat. It is a microwave frequency. It is the dielectric loss factor. Electrical conductivity (S / m); B3) Mass transfer equations: Liquid water migration within porous media follows Darcy's law, water vapor diffusion follows Fick's law, and the total water transport equation is: ； In the formula Total moisture content; K is the Darcy velocity of liquid water (m / s), and K is the permeability. It is the viscosity of the liquid (Pa). s), It is the liquid pressure (Pa); Steam diffusion flux (kg) m -2 s -1 ); and The densities of liquid water and moist air, respectively (kg / m³). 3 ); It is the effective moisture diffusion coefficient (m 2 / s); B4) Fluid flow equation: The air flow within the cavity satisfies the Navier-Stokes equation; B5) Set boundary conditions: TE10 excitation at the wave inlet, ideal electrical conductor on the cavity wall; convective heat transfer boundary between material and air; no flux boundary at the bottom, and mass transfer boundary on the remaining surfaces.

5. The multiphysics coupling calculation and drying method considering the random characteristics of materials according to claim 1, characterized in that, Step three: Discretize the computational domain using the finite element method and solve the electromagnetic field using the frequency domain method, specifically including: The computational domain is discretized using the finite element method. Local mesh refinement is performed on key areas including the material domain, waveguide, and air inlet / outlet, and boundary layer mesh is added to the wall surface. The electromagnetic field is solved using the frequency domain method. The steady-state solver calculates the steady-state distribution of the electromagnetic field and the flow, while the transient solver calculates the evolution of the temperature and humidity fields over time. The heat generation term calculated by electromagnetic waves is coupled to the heat transfer equation, and the temperature change is fed back to the dielectric properties and fluid properties. After the calculations are completed, the temperature field, electric field distribution, moisture content distribution, and drying kinetic curves are extracted and compared with the experimental results. The root mean square error (RMSE) is calculated to verify the model accuracy.

6. The multiphysics coupling calculation and drying method considering the random characteristics of materials according to claim 5, characterized in that, The discretization of the computational domain using the finite element method specifically includes: The loading area of ​​the drying chamber, i.e., the computational domain, is divided into a finite number of non-overlapping micro-units. Within each unit, the unknown electric field is... Using the values ​​at the nodes and the shape function N i The vector wave equation is approximated, and then the weighted residual method is used to transform the vector wave equation into a discrete system of algebraic equations. The electromagnetic field is solved using the frequency domain method. The steady-state solver calculates the steady-state distribution of the electromagnetic field and the flow, while the transient solver calculates the evolution of the temperature and humidity fields over time. First, the electromagnetic field is solved using the frequency domain method. This involves assuming the time-harmonic electromagnetic field is in complex amplitude form and substituting it into Maxwell's equations to obtain the electric field... The frequency domain Helmholtz wave equation is discretized and the complex linear equations are solved using the finite element method to obtain the steady-state distribution of the electric field amplitude and phase in the computational domain. Then, the steady-state solver is activated to simultaneously solve the steady-state flow control equations based on the known electromagnetic field distribution and the boundary conditions of the drying cavity, obtaining the steady-state distributions of the air velocity, pressure, and temperature fields. At this point, neither the electromagnetic field nor the flow field changes with time. Finally, the transient solver is invoked, using the microwave heat source term provided by the electromagnetic field. Using the convective heat transfer and mass transfer coefficients provided by the steady-state flow field as inputs, the transient heat conduction equation and moisture diffusion equation inside the material are solved step by step in the time domain, and the temperature field T is updated at each time step. ,t) and humidity field W( The electromagnetic field distribution is adjusted based on the dielectric parameters of the material, which are dependent on temperature and humidity, and this process is repeated until the preset drying time or moisture content target is reached.

7. The multiphysics coupling calculation and drying method considering the random characteristics of materials according to claim 1, characterized in that, Step four, process parameter optimization and fuzzy PID control based on a stochastic model, specifically includes: C1) Process parameter optimization: Determine the optimal combination of process parameters through orthogonal experiments or response surface methodology; stochastic model refers to a model with randomness in the size distribution, random stacking position, and initial humidity of particulate materials; C2) Fuzzy PID Control: Based on the stochastic model, a fuzzy PID controller is designed for real-time control of the drying process. The specific method is as follows: The deviation e between the highest internal temperature of the material and the target temperature and the rate of change of the deviation ec are used as input variables, and the adjustment of PID control parameters (Kp, Ki, Kd) are used as output variables. A fuzzy rule base is established, and the real-time PID parameter correction is obtained through fuzzy inference and defuzzification; A fuzzy PID controller is embedded in a multiphysics coupling model to simulate the drying process under closed-loop control.

8. A storage device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the multiphysics coupling calculation and drying method considering the random properties of materials as described in any one of claims 1-7 by calling the computer program stored in the memory.

9. A computer-readable storage medium for storing a computer program for movable property pledge financing based on blockchain, characterized in that, The computer program, when running on a computer, performs the steps of the multiphysics coupling calculation and drying method considering the random characteristics of materials as described in any one of claims 1-7.

Images