Microwave heating acceleration calculation method, equipment and medium

Abstract

The invention discloses a microwave heating acceleration calculation method, microwave heating acceleration calculation equipment and a medium, and relates to the technical field of microwave heating numerical simulation, multi-physics coupling calculation and intelligent prediction algorithms. According to the method, the significant difference of the electromagnetic field and the temperature field on the time scale is taken as the theoretical basis, the number of numerical solution times of the temperature field is significantly reduced on the premise of ensuring the simulation precision by reconstructing the multi-physical field solution process and introducing a temperature field adaptive prediction mechanism, so that accelerated calculation of microwave heating multi-physical field simulation is realized.

Description

Technical Field

[0001] This invention relates to the fields of microwave heating numerical simulation, multiphysics coupling calculation and intelligent prediction algorithm technology, and particularly to a calculation method for accelerating microwave heating multiphysics simulation by decoupling the time scales of electromagnetic field and temperature field and introducing an adaptive prediction model for temperature field. Specifically, it relates to a microwave heating accelerated calculation method, equipment and medium. Background Technology

[0002] The statements in this section are provided only as background information in relation to this disclosure and may not constitute prior art.

[0003] Microwave heating is a heating method based on the interaction between electromagnetic waves and matter to achieve energy conversion. Its basic mechanism involves the microwave electromagnetic field acting on a dielectric material, causing polarization loss, conductivity loss, and other mechanisms within the medium to convert electromagnetic energy into heat energy, thus achieving bulk heating or selective heating. Due to its advantages such as rapid heating speed, high energy utilization rate, and flexible methods, microwave heating has become an important technology in food processing, chemical reactions, material sintering, biomedical processing, and microwave plasma applications. In practical engineering design and process optimization, computer numerical simulation has become a common method for analyzing the microwave heating process and its inherent physical laws. Simulation allows for quantitative analysis of the spatial distribution of the electromagnetic field within the microwave cavity, the evolution of the temperature field within the medium, and the dynamic coupling relationships between other physical fields, providing theoretical basis and decision support for equipment design, parameter optimization, and process control.

[0004] In typical simulation frameworks, microwave heating is a multiphysics coupled problem, requiring the joint solution of at least two core physical fields: the electromagnetic field and the temperature field. The electromagnetic field describes the propagation, reflection, standing wave formation, and energy distribution of microwaves within the heating cavity and medium, while the temperature field describes the heat generated within the medium due to the absorption of electromagnetic energy and the thermal diffusion process over time. In more complex real-world scenarios, it may be necessary to further consider physical processes such as changes in material electromagnetic parameters with temperature, phase transitions, fluid flow, and chemical reactions, making the overall simulation problem highly nonlinear and strongly coupled. Many engineering and research cases employ the finite element method, finite difference method, or similar numerical methods to establish multiphysics coupled models to analyze the heat, electricity, and mass transport laws during microwave heating.

[0005] Currently, multiphysics simulation of microwave heating faces significant computational efficiency pressures in engineering applications, especially in industrial microwave equipment, complex process systems, and large-scale material heating scenarios. Typical microwave heating simulations require jointly solving the coupled electromagnetic and temperature fields, closely linking them on both temporal and spatial scales. Microwaves, as high-frequency electromagnetic waves, act on the medium at GHz levels, and the electromagnetic field instantaneously establishes and forms a steady-state field distribution within the cavity and material. In contrast, the temperature field is controlled by thermal diffusion processes, with response times typically on the order of milliseconds or even longer. Due to this difference in physical mechanisms, repeatedly solving for the electromagnetic and temperature fields at the same simulation time step is extremely mismatched in terms of time scale and computational resource consumption, resulting in very long overall simulation times. Particularly in industrial microwave heating equipment and large-scale media, the high dielectric constant and loss factor of the materials necessitate the use of very fine three-dimensional meshes to accurately describe the field distribution and heat accumulation behavior in the electromagnetic-thermal coupling region, further amplifying the computational scale and resource consumption of the problem. Such high-dielectric materials not only increase the number of mesh elements, but also make it more difficult for the numerical solution iteration process to converge. They usually require a large amount of CPU / GPU resources and a long solution cycle, which greatly limits the availability of high-precision 3D simulation in the engineering design cycle.

[0006] Traditional microwave heating simulation methods typically employ sequential or weakly coupled computational strategies, alternating between solving the electromagnetic and temperature fields within a unified time step. First, the electromagnetic field is solved under known material parameters and geometry, yielding the spatial electric and magnetic field distributions and corresponding volume power loss. Then, the volume power loss is introduced as a heat source term into the heat conduction equation to solve for the temperature field. Material parameters are updated based on the temperature field as needed, and the process is repeated in the next time step. The core assumption of this method is that the electromagnetic and temperature fields can evolve synchronously with the same time step size. This approach is widely used for simple models or simulations requiring low precision due to its relatively direct implementation and clear workflow.

[0007] However, from a physical perspective, the time response characteristics of electromagnetic fields and temperature fields differ significantly. Microwaves, as high-frequency electromagnetic waves, typically establish a steady-state distribution within a cavity within nanoseconds, while the evolution of a temperature field, determined by thermal diffusion and material heat capacity, often has characteristic times on millisecond or even longer timescales. In other words, within a timeframe during which a significant temperature change occurs, the electromagnetic field may have already undergone a considerable amount of steady-state establishment. Therefore, repeatedly solving for the electromagnetic field within a uniform time step framework generates a large amount of redundant computation. This not only increases the computational burden but also leads to a significant decrease in overall simulation efficiency as model size and accuracy requirements increase.

[0008] To address the aforementioned issues, existing technologies have proposed several improvement strategies, such as increasing the solution time step to reduce the number of solutions, simplifying the electromagnetic model to reduce computational complexity, and employing parallel computing or hardware acceleration to improve the speed of a single solution. However, these methods essentially remain at the level of "speeding up the solution" and do not fundamentally reduce the number of repeated electromagnetic field solutions or effectively utilize the regularity of temperature field evolution to improve simulation efficiency. While increasing the time step can reduce the number of solutions, it can easily introduce numerical errors or instability during periods of drastic temperature changes; simplifying the model may sacrifice spatial resolution and physical realism; parallel or hardware acceleration increases costs without changing the overall solution framework. Therefore, existing methods still face performance bottlenecks when dealing with high-precision, large-scale three-dimensional microwave heating multiphysics coupled simulations.

[0009] On the other hand, the evolution of the temperature field during microwave heating typically exhibits significant temporal continuity and spatial regularity. Temperature changes between adjacent time steps are relatively gradual and are dominated by the distribution of the electromagnetic power field. It is precisely this physical continuity and regularity that makes the temperature field predictable within a certain time window; that is, its future evolution trend can be approximated using certain prediction methods without significant loss of accuracy. Furthermore, with the rapid development of artificial intelligence technology, deep neural networks, convolutional networks, and sequence models have achieved remarkable results in complex nonlinear mappings and time series prediction. By introducing physical constraints or combining physical information neural networks with physical equations, the physical consistency and accuracy of prediction results can be guaranteed to a certain extent, providing new ideas and technical foundations for predictive substitution of numerical solutions in multiphysics simulations.

[0010] In summary, while existing multiphysics simulation techniques for microwave heating have achieved widespread application in terms of process standardization and numerical solution methods, they still suffer from core problems such as underutilizing the time-scale differences between electromagnetic and temperature fields, failing to effectively reduce redundant solutions, and lacking simulation acceleration strategies based on temperature field prediction mechanisms. Therefore, there is an urgent need to propose a new multiphysics simulation method for microwave heating that can fully isolate the time-scale differences between electromagnetic and temperature fields, reduce unnecessary numerical solutions through adaptive prediction mechanisms, and ensure the accuracy and stability of the simulation process through verification and correction mechanisms. This would enable efficient, accurate, and engineering-usable microwave heating simulation calculations.

[0011] In the existing technology, there has been some development in invention patents related to microwave heating and its temperature field simulation and control, but these mainly focus on simulation process standardization, temperature field feedback control, or optimization of single physical quantities. Typical related patents include: (1) Invention patent “A Smart Monitoring Method for Microwave Heating Temperature Field Based on Online Learning” (Application No.: CN201811197619.0, corresponding publication / authorization No. CN109287021B), this technology proposes to improve temperature monitoring and heating control strategies by introducing an online learning model to predict and control the temperature field in real time during microwave heating, thereby achieving intelligent adjustment and uniformity control of the heating process. This scheme mainly focuses on the application of the online learning model in the actual heating process for temperature monitoring and control, rather than directly for accelerating simulation calculations; its purpose is to improve the control performance during the heating process, rather than to optimize the efficiency of multiphysics simulation solutions.

[0012] (2) The invention patent “A simulation method for microwave thermal ablation temperature field based on characteristic parameter feedback” (application number: CN201710719021.2, corresponding authorization number CN107526928B) analyzes the temperature field sensitivity parameters during the thermal ablation process, combines measured feedback data to obtain accurate characterization of the characteristic parameters, and adjusts the simulation model based on these parameters to improve the accuracy of temperature field simulation. This scheme proposes a feedback optimization mechanism to address parameter uncertainty in the microwave thermal ablation scenario, aiming to improve the consistency between simulation results and experimental measurements. However, its processing method is still based on the traditional numerical simulation process and does not fundamentally reduce the number of alternating solutions for electromagnetic field and temperature field or improve simulation efficiency.

[0013] (3) The invention patent "A Microwave Heating Simulation Analysis Method" (authorization number: CN113094955B, corresponding application number 202110388063.9) discloses a microwave heating simulation method based on finite element analysis, including steps such as geometric modeling, physical field setting, solving the electromagnetic field and heat conduction field, mesh generation and result analysis, etc., which systematically standardizes and refines the overall simulation process. This technology is applicable to general microwave heating simulation analysis, but it still uses a uniform time step to solve the electromagnetic field and heat conduction field, and does not propose an effective solution optimization strategy for the problem of mismatch between time scales of different physical fields.

[0014] A review of the aforementioned existing patents reveals that while they have explored microwave heating or microwave thermal ablation to varying degrees from the perspectives of intelligent monitoring, parameter feedback optimization, and simulation process standardization, none of them have proposed a systematic solution method to address the core issue of redundancy and inefficiency in microwave heating simulations caused by the difference in response time scales between electromagnetic and temperature fields. Furthermore, when introducing machine learning or online intelligent control, existing technologies focus on process control or real-time measurement and control, rather than incorporating predictive models as part of the simulation loop to replace numerical solutions, automatically adjust time steps, and establish correction mechanisms.

[0015] Therefore, there is still a significant technical need based on existing technologies, namely, how to design a microwave heating multiphysics simulation acceleration method that can balance simulation accuracy and computational efficiency from the perspective of multiphysics coupling mechanism and time scale difference. Summary of the Invention

[0016] Existing microwave heating multiphysics simulation technologies suffer from low computational efficiency, high resource consumption, and difficulty in meeting the needs of rapid engineering design and optimization. The technical problems that this invention aims to solve mainly include, but are not limited to, the following aspects: (1) The problem of low simulation efficiency caused by the severe mismatch between the time scales of electromagnetic field and temperature field Existing microwave heating simulation methods typically solve the electromagnetic field and temperature field alternately at a uniform time step, failing to distinguish the time scale difference between the nanosecond-level establishment of the electromagnetic field and the millisecond-level evolution of the temperature field. This results in the need for frequent and repetitive high-computation electromagnetic field solutions during the extremely slow-changing phase of the temperature field, leading to a large amount of redundant computation.

[0017] (2) The problem of rapid expansion of computational scale caused by large-scale, high-dielectric materials in industrial microwave heating scenarios In industrial microwave equipment simulation, the heating cavity is large, the material volume is large, and the dielectric constant and dielectric loss are high. In order to ensure the electromagnetic field distribution and power deposition resolution, fine three-dimensional meshes must be used for discrete modeling, resulting in a huge number of meshes, high memory consumption, and long solution time, making it difficult for traditional full numerical coupling simulation methods to be completed within the engineering cycle.

[0018] (3) The problem of over-reliance on full numerical solutions and lack of effective calculation reduction mechanisms in multiphysics coupled simulation. Existing technologies generally adopt a "full solution" strategy, which means that the electromagnetic field and temperature field equations are completely solved at each time step. This fails to take advantage of the temporal continuity and spatial regularity in the temperature field evolution process and lacks a simulation acceleration mechanism that can replace some numerical solution steps with prediction or approximation methods while ensuring accuracy.

[0019] (4) Existing simulation acceleration methods cannot simultaneously improve efficiency and maintain simulation accuracy and stability. Although existing technologies have attempted to improve simulation speed by increasing the time step, simplifying the model, parallel computing, or hardware acceleration, these methods either easily introduce numerical errors and instabilities or increase system costs, and none of them fundamentally reduce the number of repeated solutions between the electromagnetic field and the temperature field, making it difficult to achieve significant results in high-precision, large-scale simulations.

[0020] (5) The problem that artificial intelligence prediction methods have not been effectively integrated into the main loop of microwave multiphysics simulation. Existing technologies involving machine learning or online learning are mainly applied to temperature monitoring, process control, or simulation result correction. They do not use predictive models as a core component of the simulation calculation process to replace part of the numerical solution of temperature fields, dynamically adjust the solution rhythm, or participate in multiphysics coupling calculations, thus limiting their application value in simulation acceleration.

[0021] (6) The problem of lacking a mechanism for verifying prediction results and adaptive correction. When attempting to introduce approximate calculation or prediction methods, existing technologies typically lack systematic error assessment, verification, and correction mechanisms. This makes it difficult to dynamically assess the reliability of prediction results and trigger accurate solutions in a timely manner during the simulation process, thus hindering the long-term stability and engineering credibility of simulation results.

[0022] Based on the above-mentioned technical problems, the purpose of this invention is to propose a method for accelerating the calculation of multi-physics simulation of microwave heating by separating the time scale difference between electromagnetic field and temperature field, introducing an adaptive prediction model of temperature field, and combining electromagnetic field verification and correction mechanism. This method can significantly reduce redundant solution process in multi-physics simulation, reduce computing resource consumption, and improve the overall computing efficiency and engineering application value of industrial microwave heating simulation while ensuring simulation accuracy and physical consistency.

[0023] In summary, to address the problems of long computation time, high computational resource consumption, and difficulty in meeting engineering application efficiency requirements while maintaining simulation accuracy in existing microwave heating multiphysics simulations, this invention proposes an accelerated computation method for microwave heating multiphysics simulations by removing adaptive temperature field prediction. This technical solution is based on the significant time-scale differences between the electromagnetic field and the temperature field. By reconstructing the multiphysics solution process and introducing an adaptive temperature field prediction mechanism, the number of temperature field numerical solutions is significantly reduced while maintaining simulation accuracy, thereby accelerating the computation of microwave heating multiphysics simulations.

[0024] The technical solution of the present invention is as follows: A microwave heating acceleration calculation method includes: Step S1: Construct a microwave heating multiphysics simulation model, which includes a microwave cavity, a microwave excitation source, and a material to be heated, wherein the dielectric parameter of the material to be heated is set as a function of temperature; Step S2: During the simulation, the electromagnetic field is solved in the entire time domain using a numerical solution method to obtain the electromagnetic power loss density, and the electromagnetic power loss density is used as the heat source term for the temperature field evolution. Step S3: In the initial stage of the simulation, the temperature field is solved using numerical methods to obtain temperature field evolution data; Step S4: Establish a temperature field prediction model and train the temperature field prediction model using the temperature field evolution data; Step S5: In non-critical time intervals, use the trained temperature field prediction model to replace the numerical solution method and calculate the predicted temperature field distribution. Step S6: Monitor the simulation process to determine key time nodes. When a key time node is determined, perform a fully coupled numerical solution of the electromagnetic field and temperature field, and use the results of the fully coupled numerical solution to correct the temperature field prediction model.

[0025] Further, step S1 includes: Step S11: Establish a multiphysics simulation model for microwave heating, including cavity dimensions, excitation source location, boundary conditions, and the shape and location of the heated material; Step S12: Define the electromagnetic field control equation, which is used to describe the propagation and distribution of microwaves in the microwave cavity and the heated material, wherein the relative permittivity and dielectric loss factor of the heated material are set as functions of temperature. Step S13: Define the heat conduction equation, which describes the temperature evolution inside the heated material and uses the electromagnetic power loss density as the heat source term; Step S14: Set the thermal properties of the material to be heated, including density, specific heat capacity and thermal conductivity.

[0026] Furthermore, the specific expression of the electromagnetic field control equation is as follows:

[0027] in: Indicates the change in temperature of the heated material The varying relative permittivity; Represents the curl operator; Indicates the permeability of free space; Indicates angular frequency; Represents the vacuum permittivity; Represents the electric field vector; The specific expression for the heat conduction equation is as follows:

[0028] in: Indicates the density of the material being heated; This indicates the specific heat capacity of the material being heated; Indicates time; Indicates the thermal conductivity of the material being heated; This indicates the density of electromagnetic power loss inside the heated material. , Indicates the change in temperature of the heated material The varying dielectric loss factor.

[0029] Further, step S2 includes: Step S21: At each temperature update node, based on the relative permittivity of the heated material as a function of temperature... Solving for steady-state or quasi-steady-state electromagnetic fields ;in, Indicates the current time step The temperature of the material being heated; Indicates the current time step The electric field vector magnitude obtained by solving the following method; Step S22: At each temperature update node, calculate the electromagnetic power loss density inside the heated material:

[0030] in: Represents spatial location coordinates; Indicates the change in temperature of the heated material The varying dielectric loss factor; Step S23: Electromagnetic power loss density inside the material to be heated As a heat source for the evolution of the temperature field.

[0031] Further, step S3 includes: Step S31: In the initial stage of heating, select a continuous time step and use the finite element method or finite difference method to solve the temperature field in a fully coupled manner:

[0032] in: Indicates at time step Below, located at spatial coordinates The actual temperature value was obtained through fully coupled numerical solution. Step S32: Save sample data on the evolution of the temperature field over time; Step S33: Establish the mapping relationship between temperature field changes, electromagnetic power distribution, and historical temperature, and construct a sample dataset for training the temperature field prediction model.

[0033] Further, step S4 includes: Step S41: Construct a temperature field prediction model; its input: the current time step The temperature below Previous time step The temperature below Rate of change of relative permittivity Its output: the next time step Predicted temperature :

[0034] in: Indicates the first Index of rules; Indicates the first Rule #1 Fuzzy linguistic variables of input variables (where ); Indicates the first The rule outputs the predicted temperature for the next time step; Indicates the first Rule consequent weight; Step S42: Train the temperature field prediction model using historical full numerical solution data.

[0035] Further, step S5 includes: During non-critical time intervals, the temperature field is calculated using a temperature field prediction model:

[0036] in: For the first Rule activation level.

[0037] Further, step S6 includes: Step S61: Determine whether the current time step meets the key time node determination conditions, wherein the determination conditions are any one or more of the following: (1) Prediction error exceeds the threshold:

[0038] in, Represents the prediction error norm. This indicates the preset temperature error threshold; (2) The rate of change of the material's dielectric constant exceeds the threshold:

[0039] in: This represents the preset threshold for the rate of change of dielectric constant; (3) The temperature gradient or heating rate exceeds the threshold:

[0040] in, Indicates the current time step Next position Temperature gradient at that location, This indicates the preset temperature gradient threshold; (4) Reaching the maximum number of prediction steps: The number of consecutive steps using the temperature field prediction model reaches the preset maximum number of prediction steps. ; Step S62: If it is determined to be a critical time node, use the actual temperature value obtained by fully coupled numerical solution. Compared with predicted temperature value Error field between weights of the consequent Perform online corrections:

[0041] in: This represents the set of consequent parameters to be corrected, including ; This indicates the corrected parameter value; This indicates the parameter value before correction; This represents the learning rate, used to control the step size for parameter adjustment; This represents the partial derivative of the squared prediction error with respect to the consequent parameter; Step S63: After completing the online parameter correction, adaptively adjust the next prediction step size. : According to the prediction error Size dynamic adjustment When the prediction error If the value is less than the preset stability threshold, increase the next prediction step size. When the prediction error If the value is greater than the stability threshold, decrease the next prediction step size. .

[0042] The present invention also proposes an electronic device, comprising: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores instructions that can be executed by the at least one processor, and the at least one processor executes the instructions stored in the memory to perform the method described above.

[0043] The present invention also proposes a computer-readable storage medium for storing instructions that, when executed, cause the method described above to be implemented.

[0044] Compared with existing technologies, the advantages of this invention are: Compared with existing microwave heating multiphysics simulation methods, the fundamental difference of this invention lies in the reconstruction of the multiphysics simulation process itself. Existing technologies generally perform fully coupled numerical solutions for the electromagnetic field and temperature field simultaneously at each time step, resulting in the need to repeatedly perform computationally intensive temperature field solutions during periods of slow temperature change. In contrast, this invention, based on the physical fact that the electromagnetic field and temperature field have significantly different time scales, separates the temperature field from the unified time-step solution process. In non-critical time intervals, an adaptive prediction model is used to directly generate the temperature field distribution. Only when the prediction error exceeds the limit, the dielectric parameters of the material change drastically, or the temperature evolution characteristics change significantly, is a new electromagnetic-thermal fully coupled numerical solution node triggered, and the high-precision numerical results of this node are used to correct the prediction model. At the same time, by dynamically evaluating the length of the prediction interval, adaptive adjustment of the simulation segments is achieved. Thus, while ensuring the physical accuracy of the electromagnetic power distribution and the reliability of the temperature field calculation, the number of temperature field numerical solutions is significantly reduced, and the overall simulation calculation time and computational resource consumption are significantly reduced. This technical concept and implementation method have not yet been seen in existing microwave heating multiphysics simulation technologies. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in the embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0046] Figure 1 A flowchart of a microwave heating acceleration calculation method; Figure 2 The heating model constructed in Example 3; Figure 3 The implementation process of Example 3; Figure 4 This is a comparison chart of the predicted temperature and the actual simulated temperature when the input power is set to 200W in Example 3; Figure 5 This is a schematic diagram of the temperature distribution when the power is 500W in Example 3; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0047] Figure reference numerals: 1-microwave feed port, 2-heated load, 3-metal cavity. Detailed Implementation

[0048] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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 a process, method, article, or apparatus. Without further limitations, 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.

[0049] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0050] Example 1 Please see Figure 1 A microwave heating acceleration calculation method, comprising: Step S1: Construct a microwave heating multiphysics simulation model, which includes a microwave cavity, a microwave excitation source, and a material to be heated. The dielectric parameter of the material to be heated is set as a function of temperature. That is, establish the basic geometric model and physical parameters of the microwave heating system to provide a data foundation for subsequent simulations. Step S2: During the simulation, the electromagnetic field is solved in the entire time domain using a numerical solution method to obtain the electromagnetic power loss density, and the electromagnetic power loss density is used as the heat source term for the temperature field evolution; that is, to ensure the physical accuracy of the electromagnetic power deposition distribution and to provide a reliable driving source for the temperature field evolution. Step S3: In the initial stage of simulation, the temperature field is solved using numerical methods to obtain temperature field evolution data; that is, high-confidence training samples are generated to provide a data foundation for the prediction model. Step S4: Establish a temperature field prediction model and train the temperature field prediction model using the temperature field evolution data; Step S5: In non-critical time intervals, the trained temperature field prediction model is used to replace the numerical solution method to calculate the predicted temperature field distribution; that is, in non-critical time intervals, the temperature field prediction model is introduced to replace the numerical solution of the temperature field and the predicted temperature field distribution is obtained directly. Step S6: Monitor the simulation process to determine key time nodes. When a key time node is determined, perform a fully coupled numerical solution of the electromagnetic field and temperature field, and use the results of the fully coupled numerical solution to correct the temperature field prediction model. That is, in the early stage of the simulation and at key time nodes, the temperature field is solved using a fully numerical method to obtain high-precision temperature field evolution results; at the same time, through error evaluation and property change monitoring mechanisms, key time nodes are dynamically determined, and the fully coupled numerical solution of the electromagnetic field and temperature field is re-executed at the key time nodes.

[0051] In this embodiment, specifically, step S1 includes: Step S11: Establish a multiphysics simulation model for microwave heating, including cavity dimensions, excitation source location, boundary conditions, and the shape and location of the heated material; Step S12: Define the electromagnetic field control equation, which is used to describe the propagation and distribution of microwaves in the microwave cavity and the heated material, wherein the relative permittivity and dielectric loss factor of the heated material are set as functions of temperature. Step S13: Define the heat conduction equation, which describes the temperature evolution inside the heated material and uses the electromagnetic power loss density as the heat source term; Step S14: Set the thermal properties of the material to be heated, including density, specific heat capacity and thermal conductivity.

[0052] In this embodiment, the specific expression of the electromagnetic field control equation is as follows:

[0053] in: Indicates the change in temperature of the heated material The varying relative permittivity; Represents the curl operator; Indicates the permeability of free space; Indicates angular frequency; Represents the vacuum permittivity; Represents the electric field vector; The specific expression for the heat conduction equation is as follows:

[0054] in: Indicates the density of the material being heated; This indicates the specific heat capacity of the material being heated; Indicates time; Indicates the thermal conductivity of the material being heated; This indicates the density of electromagnetic power loss inside the heated material. , Indicates the change in temperature of the heated material The varying dielectric loss factor.

[0055] In this embodiment, specifically, step S2 includes: Step S21: At each temperature update node, based on the relative permittivity of the heated material as a function of temperature... Solving for steady-state or quasi-steady-state electromagnetic fields ;in, Indicates the current time step The temperature of the material being heated; Indicates the current time step The electric field vector magnitude obtained by solving the following method; Step S22: At each temperature update node, calculate the electromagnetic power loss density inside the heated material:

[0056] in: Represents spatial location coordinates; Indicates the change in temperature of the heated material The varying dielectric loss factor; Step S23: Electromagnetic power loss density inside the material to be heated As a heat source for the evolution of the temperature field.

[0057] In this embodiment, specifically, step S3 includes: Step S31: In the initial stage of heating, select a continuous time step and use the finite element method or finite difference method to solve the temperature field in a fully coupled manner:

[0058] in: Indicates at time step Below, located at spatial coordinates The actual temperature value was obtained through fully coupled numerical solution. Step S32: Save sample data on the evolution of the temperature field over time; Step S33: Establish the mapping relationship between temperature field changes, electromagnetic power distribution, and historical temperature, and construct a sample dataset for training the temperature field prediction model (such as Adaptive Neuro-Fuzzy Inference System, ANFIS).

[0059] In this embodiment, specifically, step S4 includes: Step S41: Construct a temperature field prediction model; its input: the current time step The temperature below Previous time step The temperature below Rate of change of relative permittivity Its output: the next time step Predicted temperature :

[0060] in: Indicates the first Index of rules; Indicates the first Rule #1 Fuzzy linguistic variables of input variables (where ); Indicates the first The rule outputs the predicted temperature for the next time step; Indicates the first Rule consequent weight; Step S42: Train the temperature field prediction model using historical full numerical solution data.

[0061] In this embodiment, specifically, step S5 includes: During non-critical time intervals, the temperature field is calculated using a temperature field prediction model:

[0062] in: For the first Rule activation level; In this embodiment, it should be noted that physical constraints (such as the conservation of the heat conduction equation) can also be introduced to improve the rationality of the prediction.

[0063] In this embodiment, specifically, step S6 aims to prevent the accumulation of prediction errors, ensure simulation accuracy, and simultaneously perform online correction of the prediction model, specifically including: Step S61: Determine whether the current time step meets the key time node determination conditions, wherein the determination conditions are any one or more of the following: (1) Prediction error exceeds the threshold:

[0064] in, Represents the prediction error norm. This indicates the preset temperature error threshold; (2) The rate of change of the material's dielectric constant exceeds the threshold:

[0065] in: This represents the preset threshold for the rate of change of dielectric constant; (3) The temperature gradient or heating rate exceeds the threshold:

[0066] in, Indicates the current time step Next position Temperature gradient at that location, This indicates the preset temperature gradient threshold; (4) Reaching the maximum number of prediction steps: The number of consecutive steps using the temperature field prediction model reaches the preset maximum number of prediction steps. ; Step S62: If it is determined to be a critical time node, use the actual temperature value obtained by fully coupled numerical solution. Compared with predicted temperature value Error field between weights of the consequent Perform online corrections; The fully coupled numerical solution includes: Update the electromagnetic field and calculate the power distribution:

[0067] Numerical solution of the temperature field:

[0068] Calculate the error field:

[0069] Weights of the successor Perform online corrections:

[0070] in: This represents the set of consequent parameters to be corrected, including ; This indicates the corrected parameter value; This indicates the parameter value before correction; This represents the learning rate, used to control the step size for parameter adjustment; This represents the partial derivative of the squared prediction error with respect to the consequent parameter; Step S63: After completing the online parameter correction, adaptively adjust the next prediction step size. : According to the prediction error Size dynamic adjustment When the prediction error If the value is less than the preset stability threshold, increase the next prediction step size. When the prediction error If the value is greater than the stability threshold, decrease the next prediction step size. .

[0071] In the cyclic regression prediction phase, calculations continue for non-critical time periods; this involves adaptive evaluation and cyclic execution of the simulation segments. The aim is to achieve adaptive adjustment of the simulation segments, reduce redundant numerical solutions, and ensure overall accuracy. ① The entire heating process is automatically divided into multiple simulation stages, each stage including: initial full numerical solution stage, prediction calculation stage, and verification and correction stage; ② The length of each prediction stage is dynamically adjusted based on the prediction error and changes in physical properties: for small errors, the prediction step size is increased; for large errors, the prediction step size is shortened and fully coupled nodes are added. ③ The process is repeated until the entire heating process is finished; Technical features: adaptive segmented control mechanism and error evaluation mechanism.

[0072] Based on the same technical concept, embodiments of the present invention also provide an electronic device that can implement the microwave heating accelerated calculation method provided in the above embodiments of the present invention. In one embodiment, the electronic device can be a server, a terminal device, or other electronic equipment. Figure 6 As shown, the electronic device may include: At least one processor and a memory connected to the at least one processor. In this embodiment of the invention, the specific connection medium between the processor and the memory is not limited. Figure 6 The example used is the connection between the processor and memory via a bus. The bus... Figure 6 The connections between other components are indicated by thick lines and are for illustrative purposes only, not as limiting information. Buses can be divided into address buses, data buses, control buses, etc., but for ease of representation, [the specific bus type is not shown here]. Figure 6 The processor is represented by a single thick line, but this does not imply that there is only one bus or one type of bus. Alternatively, a processor can also be called a controller; there are no restrictions on the name.

[0073] In this embodiment of the invention, the memory stores instructions executable by at least one processor. By executing the instructions stored in the memory, the at least one processor can execute the microwave heating acceleration calculation method described above. The processor can implement... Figure 6 The functions of each module in the device shown.

[0074] The processor is the control center of the device. It can connect to various parts of the control device through various interfaces and lines. By running or executing instructions stored in memory and calling data stored in memory, it can monitor the device's various functions and process data, thereby enabling overall monitoring of the device.

[0075] In an alternative design, the processor may include one or more processing units. The processor may integrate an application processor and a modem processor, wherein the application processor primarily handles the operating system, user interface, and applications, while the modem processor primarily handles wireless communication. It is understood that the modem processor may also not be integrated into the processor. In some embodiments, the processor and memory may be implemented on the same chip; in some embodiments, they may also be implemented separately on separate chips.

[0076] The processor can be a general-purpose processor, such as a CPU, digital signal processor, application-specific integrated circuit, field-programmable gate array or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the microwave heating acceleration calculation method disclosed in the embodiments of this invention can be directly manifested as being executed by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0077] Memory, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. Memory can include at least one type of storage medium, such as flash memory, hard disk, multimedia cards, card-type memory, random access memory (RAM), static random access memory (SRAM), programmable read-only memory (PROM), read-only memory (ROM), and electrically erasable programmable read-only memory (EPROM). Only memory (EEPROM), magnetic storage, magnetic disks, optical disks, etc. A memory is any other medium capable of carrying or storing desired program code in the form of instructions or data structures, and accessible by a computer, but is not limited thereto. The memory in embodiments of this invention can also be a circuit or any other device capable of performing storage functions for storing program instructions and / or data.

[0078] By designing and programming the processor, the code corresponding to the microwave heating accelerated computing method described in the foregoing embodiments can be embedded into the chip, enabling the chip to execute the steps of the method described in the foregoing embodiments during operation. How to design and program the processor is a technique well-known to those skilled in the art and will not be elaborated upon here.

[0079] Based on the same inventive concept, embodiments of the present invention also provide a storage medium storing computer instructions that, when executed on a computer, cause the computer to perform a microwave heating accelerated calculation method described above.

[0080] In some alternative embodiments, the present invention also provides that various aspects of the microwave heating accelerated computing method can also be implemented in the form of a program product, which includes program code that, when the program product is run on a device, causes the control device to perform the steps in the microwave heating accelerated computing method according to various exemplary embodiments of the present invention described above.

[0081] It should be noted that although several units or sub-units of the apparatus have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of the invention, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units. Furthermore, although the operation of the method of the invention is described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0082] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can be implemented in one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs) containing computer-usable program code. The form of a computer program product implemented on ROM, optical memory, etc.

[0083] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a server, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0084] Program code for performing the operations of this invention can be written using any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0085] In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0086] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0087] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0088] Example 2 Example 2 is a further illustration of the present invention. Identical components will not be described again here. To achieve the objectives of the present invention and the expected technical effects, the following technical features are indispensable and necessary: (1) Characteristics of numerical solution technology for electromagnetic fields in the whole time domain Throughout the simulation, the electromagnetic field was solved using a numerical method based on Maxwell's equations. The electromagnetic power loss density was used as the input for the heat source term of the temperature field evolution, and this feature was used to ensure the physical realism of the microwave energy deposition distribution.

[0089] (2) Characteristics of temperature-dependent dielectric parameter modeling techniques The relative permittivity and dielectric loss factor of the heated material are set as functions of temperature and updated in real time with the temperature field during the simulation. This feature ensures the accuracy of the electromagnetic-thermal coupling relationship.

[0090] (3) Characteristics of full numerical solution technology in the initial stage of temperature field In the early stages of simulation, traditional numerical methods are used to solve the temperature field in a fully coupled manner over continuous time steps to build training samples for the prediction model. This feature is the basis for the reliability of subsequent prediction calculations.

[0091] (4) Characteristics of temperature field prediction and calculation techniques In non-critical time intervals, the core technical feature for accelerating simulation is to directly generate temperature distribution results using a temperature field prediction model based on intelligent algorithms, thereby replacing numerical solutions for the temperature field.

[0092] (5) Key time node determination and full coupling correction technology features By evaluating prediction errors, dielectric parameter change rate, and temperature evolution characteristics, key time nodes are automatically identified, and fully coupled numerical solutions are re-executed at these nodes to correct prediction results and prevent error accumulation.

[0093] (6) Characteristics of Simulation-based Piecewise Adaptive Control Technology The prediction interval length is dynamically adjusted based on the prediction accuracy and changes in physical properties to achieve adaptive adjustment of the simulation segments, thereby striking a balance between accuracy and efficiency.

[0094] The following is a description of non-essential technical features: In the implementation of this embodiment, some technical features are used to improve simulation accuracy, stability or applicability, but are not necessary to achieve the technical purpose of this invention.

[0095] For example, the specific implementation of the temperature field prediction model does not constitute a limitation of this invention. It can be a deep learning model based on neural networks, or an adaptive neural fuzzy inference system (ANFIS) or other prediction algorithms with nonlinear mapping capabilities. The number of network layers, membership function type, training strategy, and hyperparameter settings used in the prediction model can all be adjusted according to the specific simulation object and computing resource conditions; The evaluation method for prediction error is not limited to a certain mathematical form; it can also use indicators such as maximum absolute error, mean square error, or temperature gradient deviation. Furthermore, whether parallel computing, GPU acceleration, or multi-core computing architecture is used, as well as the specific setting of thermal boundary conditions and convection heat transfer models, do not affect the core idea of ​​this invention to accelerate simulation by "predicting the temperature field to replace numerical solution for non-critical time periods." Therefore, the above-mentioned technical features are all non-essential technical features.

[0096] Description of technical features identical to those in the prior art In terms of technical implementation, this invention inherits and adopts several mature techniques in existing microwave heating multiphysics simulations. For example, the electromagnetic field part is still based on Maxwell's equations to establish a physical model, and numerical solutions are obtained using the finite element method or finite difference method. The temperature field still uses the heat conduction equation as the basic governing equation, and the electromagnetic power loss density is used as the heat source term for temperature field calculation. In the multiphysics coupling process, the dielectric constant, dielectric loss factor, and other electromagnetic parameters of the material are still set as functions of temperature and updated with temperature changes during the simulation. At the same time, the overall simulation process still uses a time-stepping method to perform transient analysis of the microwave heating process. The above technical features are all conventional methods known to those skilled in the art. This invention does not improve upon these features, but rather uses them as a basis to introduce new simulation processes and computational mechanisms.

[0097] Technical features that distinguish it from existing technologies Compared with existing microwave heating multiphysics simulation methods, the fundamental difference of this invention lies in the reconstruction of the multiphysics simulation process itself. Existing technologies generally perform fully coupled numerical solutions for the electromagnetic field and temperature field simultaneously at each time step, resulting in the need to repeatedly perform computationally intensive temperature field solutions during periods of slow temperature change. In contrast, this invention, based on the physical fact that the electromagnetic field and temperature field have significantly different time scales, separates the temperature field from the unified time-step solution process. In non-critical time intervals, an adaptive prediction model is used to directly generate the temperature field distribution. Only when the prediction error exceeds the limit, the dielectric parameters of the material change drastically, or the temperature evolution characteristics change significantly, is a new electromagnetic-thermal fully coupled numerical solution node triggered, and the high-precision numerical results of this node are used to correct the prediction model. At the same time, by dynamically evaluating the length of the prediction interval, adaptive adjustment of the simulation segments is achieved. Thus, while ensuring the physical accuracy of the electromagnetic power distribution and the reliability of the temperature field calculation, the number of temperature field numerical solutions is significantly reduced, and the overall simulation calculation time and computational resource consumption are significantly reduced. This technical concept and implementation method have not yet been seen in existing microwave heating multiphysics simulation technologies.

[0098] Example 3 Example 3 is a specific application of a microwave heating acceleration calculation method proposed in this invention.

[0099] First, the heating model is modeled, such as... Figure 2 As shown, 1 is the microwave feed port, which receives microwave energy; 2 is the load being heated, with dimensions of 40*40*10mm, and its dielectric constant changes with temperature, denoted as: ; 3 is a metal cavity with dimensions of 200mm*200mm*200mm.

[0100] The heating model is partitioned into external and internal regions based on the rate of change of relative permittivity, which improves algorithm accuracy. A finite element mesh is generated for the heated object in COMSOL, recording the coordinates (x, y, z) of each node, along with its temperature and corresponding relative permittivity. The rate of change η of the relative permittivity and temperature at each point per unit time is then calculated. Finally, the rates of change of relative permittivity for all points are statistically analyzed, with mesh points exhibiting high rates of change designated as the internal region and those exhibiting low rates as the surface region. This process ultimately yields the mesh points representing the interior and surface of the heated object. Figure 3 As shown.

[0101] Simulation results: When the input power is set to 200W, the predicted temperature is compared with the actual simulated temperature. Figure 4 As shown.

[0102] The data comparison is shown in the table below:

[0103] When the power is 500W, the temperature distribution is as follows: Figure 5 As shown.

[0104] The data above shows that the maximum absolute error of the external simulation obtained by this method is 2℃, while the maximum absolute error of the internal simulation is 1.6℃, indicating high prediction accuracy. The average response time of this model is 0.87 seconds / simulation, which is 69 times faster than traditional FEM simulation, significantly reducing computation time.

[0105] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

[0106] This background section is provided to generally present the context of the invention. The work of the currently named inventors, the work to the extent described in this background section, and aspects of this section that did not constitute prior art at the time of application are neither expressly nor impliedly acknowledged as prior art to the invention.

Claims

1. A method for accelerating computation of microwave heating, characterized by, The method comprises the following steps: Step S1: Constructing a microwave heating multi-physical field simulation model, the microwave heating multi-physical field simulation model comprising a microwave cavity, a microwave excitation source and a heated material, wherein the dielectric parameters of the heated material are set as functions of temperature; Step S2: In the simulation process, the electromagnetic field is solved by a numerical solution method to obtain electromagnetic power loss density, and the electromagnetic power loss density is taken as a heat source term of temperature field evolution; Step S3: In the initial stage of simulation, the temperature field is solved by a numerical solution method to obtain temperature field evolution data; Step S4: Establishing a temperature field prediction model, and training the temperature field prediction model by using the temperature field evolution data; Step S5: In a non-critical time interval, the trained temperature field prediction model is used to replace the numerical solution method to calculate a predicted temperature field distribution; Step S6: Monitoring the simulation process to determine a critical time node, and when the critical time node is determined, performing full-coupling numerical solution of the electromagnetic field and the temperature field, and modifying the temperature field prediction model by using the result of the full-coupling numerical solution.

2. The method of claim 1, wherein, The step S1 comprises: Step S11: Establishing a microwave heating multi-physical field simulation model, including cavity size, excitation source position, boundary conditions, heated material shape and position; Step S12: Defining an electromagnetic field control equation, the electromagnetic field control equation being used to describe the propagation and distribution of microwaves in the microwave cavity and the heated material, wherein the relative permittivity and the dielectric loss factor of the heated material are set as functions of temperature; Step S13: Defining a heat conduction equation, the heat conduction equation being used to describe the temperature evolution inside the heated material, and the electromagnetic power loss density being taken as a heat source term; Step S14: Setting thermal physical property parameters of the heated material, the thermal physical property parameters including density, specific heat capacity and thermal conductivity.

3. The method of claim 2, wherein, The specific expression of the electromagnetic field control equation is: Wherein: relative dielectric constant of the heated material as a function of temperature relative dielectric constant of the heated material as a function of temperature denotes the curl operator; denotes the vacuum permeability; represents the angular frequency; represents the vacuum permittivity; E represents the electric field vector; The specific expression of the heat conduction equation is: Wherein: ρ represents the density of the heated material; Cp represents the specific heat capacity of the heated material; representing time; k represents the thermal conductivity of the heated material; denotes the electromagnetic power loss density inside the heated material, , denotes the dielectric loss factor of the heated material as a function of temperature .

4. The method of claim 3, wherein, The step S2 comprises: Step S21: At each temperature update node, the relative permittivity of the heated material as a function of temperature is determined solving for steady state or quasi-steady state electromagnetic fields ; wherein, E(t) represents the temperature of the heated material at the current time step ; and E(t) represents the electric field vector magnitude solved for at the current time step ; and Step S22: At each temperature update node, the electromagnetic power loss density inside the heated material is calculated: Wherein: represents a spatial position coordinate; representing the dielectric loss factor of the heated material as a function of temperature representing the dielectric loss factor of the heated material as a function of temperature Step S23: internal electromagnetic power loss density of the heated material thermal source for the temperature field evolution.

5. The method of claim 4, wherein, The step S3 comprises: Step S31: In the initial stage of heating, a continuous time step is selected, and the temperature field is fully coupled and solved by using a finite element or finite difference method: Wherein: representing the real temperature values obtained by the fully coupled numerical solution at the spatial position coordinates at the time step ; Step S32: Sample data of the temperature field evolution over time is saved; Step S33: A mapping relationship between the temperature field change and the electromagnetic power distribution and the historical temperature is established, and a sample data set for training the temperature field prediction model is constructed.

6. The method of claim 5, wherein, The step S4 comprises: Step S41: Construct a temperature field prediction model; its input: the current time step The temperature below Previous time step The temperature below Rate of change of relative permittivity Its output: the next time step Predicted temperature : Wherein: an index representing the rule; represents the first rule for the first input variable (where ); represents the next time step prediction temperature output by the rule represents the first rule consequent weight; Step S42: The temperature field prediction model is trained by using historical full-numerical solution data.

7. The method of claim 6, wherein, The step S5 comprises: In a non-critical time interval, the temperature field is calculated by using the temperature field prediction model: Wherein: For the first Rule activation degree.

8. The method of claim 7, wherein, The step S6 comprises: Step S61: Determining whether the current time step meets the critical time node determination condition, the determination condition being any one or a combination of multiple conditions: (1) The prediction error exceeds a threshold value: wherein, denotes a prediction error norm, denotes a preset temperature error threshold; (2) The material dielectric constant change rate exceeds a threshold value: Wherein: represents a preset dielectric constant variation rate threshold value; (3) The temperature gradient or the temperature rise rate exceeds a threshold value: wherein, denotes the temperature gradient at the current time step at the current position denotes the temperature gradient at the current time step denotes a preset temperature gradient threshold value; (4) a maximum predicted number of steps is reached: The step number of continuously using the temperature field prediction model reaches a preset maximum prediction step number ; Step S62: If the critical time node is determined, the real temperature value obtained by full coupling numerical solution is used and the error field between the predicted temperature value is corrected online , the consequent weight ​ wherein: represents the set of antecedent parameters to be corrected, including ; represents the corrected parameter value; denotes the parameter value before correction; denotes a learning rate for controlling the step size of the parameter adjustment; denotes the partial derivative of the prediction error square with respect to the consequent parameters; Step S63: After completing the online correction of parameters, the next prediction step is adaptively adjusted : According to the prediction error , dynamically adjust the size of , when the prediction error is less than a preset stable threshold, increase the next prediction step ; when the prediction error is greater than the stable threshold, decrease the next prediction step .

9. An electronic device, comprising: includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the at least one processor, by executing the instructions stored in the memory, causes the at least one processor to perform the method of any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium is for storing instructions that, when executed, cause the method of any one of claims 1-8 to be implemented.

Images