Sound pressure calculation method for high-frequency transformer

Abstract

The invention discloses a high-frequency transformer sound pressure calculation method, and relates to the technical field of circuits, and the method comprises the steps: obtaining a microstructure parameter of a gap beryllium copper reed to generate a microstructure feature vector; on the basis of the microstructure feature vector, calculating magnetostriction strain of microstructure regulation and control to determine coupling field distribution description of microstructure regulation and control magnetostriction; on the basis of the coupling field distribution description, performing finite element grid division on an air domain around the high-frequency transformer to generate an air domain finite element grid; enabling a gap beryllium copper reed to be equivalent to an acoustic impedance boundary condition, and defining a pseudo-density variable in an air domain finite element grid to determine the optimal sound wave propagation path tracking in the air domain; and calculating the sound pressure of the high-frequency transformer based on the tracked sound wave propagation path. According to the method, the feature vectors are generated by obtaining the structure parameters, the magnetostrictive strain is accurately calculated, grid division is optimized, the calculation precision is improved based on boundary condition processing and multi-link cooperation, and a reliable basis is provided for product design.

Description

Technical Field

[0001] The present application relates to the field of circuit technology, and in particular to a method for calculating the sound pressure of a high-frequency transformer. Background Art

[0002] High-frequency transformers are critical equipment, and the noise they generate during operation has drawn considerable attention. Accurately calculating the sound pressure of high-frequency transformers is crucial for assessing their noise levels, optimizing their designs, and meeting relevant noise standards.

[0003] Currently, existing methods for calculating the sound pressure of high-frequency transformers primarily rely on a combination of macroscopic electromagnetic parameters and empirical acoustic models. This approach first obtains certain macroscopic electromagnetic parameters of the high-frequency transformer, such as the core's magnetic permeability and the winding's resistance and inductance, through experimental measurements or theoretical calculations. These macroscopic parameters reflect the transformer's overall electromagnetic properties but overlook the complex influences within the transformer on the electromagnetic-to-acoustic conversion process. Specifically, in actual high-frequency transformers, structural features (such as gapped beryllium copper reeds) significantly influence the magnetostrictive effect, which in turn affects the distribution of electromagnetic forces and the excitation and propagation of sound waves. Existing methods, which fail to account for these factors, result in significant errors in the sound pressure calculation, failing to accurately reflect the transformer's actual noise conditions and failing to meet the increasingly demanding requirements for transformer noise control precision. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a method for calculating the sound pressure of a high-frequency transformer to at least alleviate the above technical problems.

[0005] The technical solutions provided in the embodiments of this application are as follows: A method for calculating the sound pressure of a high-frequency transformer, comprising: Step 1, obtaining microstructure parameters of the gap beryllium copper reed to generate a microstructure feature vector; Step 2: Based on the microstructure characteristic vector, the magnetostrictive strain regulated by the microstructure is calculated to determine the coupled field distribution description of the magnetostriction regulated by the microstructure; Step 3: Based on the description of the coupled field distribution, the air domain around the high-frequency transformer is divided into finite element meshes to generate an air domain finite element mesh; Step 4: Equivalent the gap beryllium copper reed to an acoustic impedance boundary condition, and define a pseudo-density variable in the air domain finite element mesh to determine the optimal acoustic wave propagation path tracking in the air domain; Step 5: Calculate the sound pressure of the high-frequency transformer based on the traced sound wave propagation path.

[0006] The proposed method for calculating the sound pressure of high-frequency transformers breaks through traditional limitations. By generating eigenvectors from the microstructural parameters of gapped beryllium copper reeds, this method accurately captures information such as grain structure and defect distribution. Based on this, the method deeply analyzes how microstructure regulates the magnetostrictive effect. Compared to traditional macroscopic estimates, it can more accurately calculate magnetostrictive strain, providing reliable data for electromagnetic force distribution calculations and acoustic excitation. In the air domain meshing process, the magnetostrictive effect is incorporated to optimize the mesh layout and enhance the accuracy of acoustic wave propagation simulations. Furthermore, the equivalent acoustic impedance boundary condition and pseudo-density variables are used to meticulously characterize the complex effects of gap components on acoustic waves, accurately tracking propagation paths and reducing errors in sound pressure calculations. Through systematic optimization of multiple steps, this method fully considers the impact of structure on the entire electromagnetic-to-acoustic conversion process, making the sound pressure calculation results of high-frequency transformers more realistic. This lays a solid technical foundation for noise assessment and product optimization design, and effectively promotes the performance improvement of high-frequency transformers to meet stringent noise standards.

[0007] The method for calculating the sound pressure of a high-frequency transformer provided in this application has the following technical advantages: (1) By obtaining the microstructural parameters of the gapped beryllium copper reed to generate microstructural feature vectors, the structural information of key components such as the gapped beryllium copper reed is fully considered. These structural parameters can accurately describe the component's grain structure, defect distribution, and other characteristics, which are completely ignored in existing solutions. By incorporating this structural information into the calculation system, the physical state inside the transformer can be more realistically reflected, laying the foundation for the subsequent accurate calculation of sound pressure.

[0008] (2) By considering the structural influence on the magnetostrictive effect, the magnitude and distribution of magnetostrictive strain under different structures can be accurately calculated. This greatly improves the accuracy of the calculation compared to existing schemes that simply estimate the magnetostrictive effect based on macroscopic parameters. The magnetostrictive effect is a key link in the electromagnetic-to-acoustic conversion process. Accurately calculating the magnetostrictive strain helps accurately determine the distribution of the electromagnetic force, thereby providing more reliable data support for the excitation of acoustic waves.

[0009] (3) Because the description of the coupled field distribution fully considers the magnetostrictive effect of structural regulation, the mesh layout can be more rationally arranged when meshing the air domain. Compared with the coarse meshing method in existing schemes, the finite element mesh generated by this scheme can more accurately simulate the propagation process of sound waves in the air domain, improving the calculation accuracy.

[0010] (4) Existing solutions often oversimplify the transformer boundary conditions and fail to fully consider the complex effects of gap components on sound wave propagation. However, this solution, through the equivalent acoustic impedance boundary condition and the definition of pseudo-density variables, can more accurately simulate the reflection, refraction, and absorption of sound waves by gap beryllium copper reeds and other components, thereby accurately tracking the propagation path of sound waves in the air domain and avoiding errors in sound pressure calculation caused by inaccurate path calculation.

[0011] (5) The influence of the structure on the electromagnetic-acoustic conversion process is fully considered, the magnetostrictive effect is accurately calculated, the air domain grid is reasonably divided, and the sound wave propagation path is accurately tracked. Therefore, the final calculated sound pressure result of the high-frequency transformer is closer to the actual measurement value. Compared with the existing scheme, this scheme can significantly improve the accuracy of the sound pressure calculation, provide a more reliable technical basis for the noise evaluation and optimization design of high-frequency transformers, help meet the increasingly stringent noise standards, and improve the performance and quality of high-frequency transformers. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Schematic diagram of the flow of the method for calculating the sound pressure of a high-frequency transformer according to an embodiment of the present application. DETAILED DESCRIPTION

[0013] like Figure 1 As shown, an embodiment of the present application provides a method for calculating the sound pressure of a high-frequency transformer, which includes: Step 1, obtaining microstructure parameters of the gap beryllium copper reed to generate a microstructure feature vector; Step 2: Based on the microstructure characteristic vector, the magnetostrictive strain regulated by the microstructure is calculated to determine the coupled field distribution description of the magnetostriction regulated by the microstructure; Step 3: Based on the description of the coupled field distribution, the air domain around the high-frequency transformer is divided into finite element meshes to generate an air domain finite element mesh; Step 4: Equivalent the gap beryllium copper reed to an acoustic impedance boundary condition, and define a pseudo-density variable in the air domain finite element mesh to determine the optimal acoustic wave propagation path tracking in the air domain; Step 5: Calculate the sound pressure of the high-frequency transformer based on the traced sound wave propagation path.

[0014] Optionally, step 1, obtaining microstructure parameters of the gap beryllium copper reed to generate a microstructure feature vector, specifically includes the following steps: Step 11: reconstructing the micron-level three-dimensional microstructure of the interstitial beryllium copper reed to generate a scanning electron microscope image and calculate the grain size distribution characteristics based on the image; Step 12: Perform in-situ stress testing on the interstitial beryllium copper spring to obtain grain boundary stress distribution and generate synchrotron X-ray diffraction data to extract grain boundary stress characteristics; Step 13: collecting element distribution and crystal orientation data of the interstitial beryllium copper spring to generate electron backscatter diffraction data and generate orientation difference distribution characteristics based on the data; Step 14: Establish a probability correlation matrix among grain size distribution characteristics, grain boundary stress characteristics, and orientation difference distribution characteristics; Step 15: Perform eigenvalue decomposition on the probability association matrix to extract the principal components whose cumulative contribution rate is greater than the contribution rate threshold and generate a microstructure feature vector based on the principal components.

[0015] Preferably, for step 11, imaging technology is used to obtain three-dimensional structural information of the grains within the interstitial beryllium copper reed. First, the reed is sliced ​​layer by layer using a focused ion beam (FIB), with the cutting thickness controlled to the micron level. Simultaneously, a scanning electron microscope (SEM) is used to scan and image each cross-section, producing a high-resolution two-dimensional image sequence. These two-dimensional images are then stitched together using a three-dimensional reconstruction algorithm to construct a three-dimensional model of the grains within the reed. In this process, to accurately calculate the grain size distribution characteristics, each grain in the three-dimensional model must be identified and segmented. Methods based on image gradients and edge detection are typically used to determine grain boundaries. Next, the equivalent diameter (such as the area-equivalent circular diameter or volume-equivalent spherical diameter) of each grain is calculated, and the grain size distribution is statistically analyzed to obtain grain size distribution characteristics, such as the average grain size and the standard deviation of the grain size.

[0016] Preferably, for step 12, the in-situ stress test is performed to determine the stress distribution within the grain boundaries of the reed under actual operating conditions. A gapped beryllium copper reed is mounted on an in-situ tension / compression testing machine. While applying a certain stress, the structural changes in the grain boundary region are monitored in real time using a scanning electron microscope (SEM). Digital image correlation (DIC) is used to analyze the SEM images and calculate the strain at the grain boundary. Based on the elastic modulus of the material, the grain boundary stress is then inferred using Hooke's law (σ = E·ε). To more accurately characterize the grain boundary stress, synchrotron X-ray diffraction is employed. A high-intensity, highly collimated X-ray beam generated by a synchrotron radiation source penetrates the reed, irradiating the grain boundary region and collecting a diffraction pattern. Based on the Bragg equation (2d·sinθ = nλ), interplanar spacing information is extracted from the diffraction pattern, and lattice strain is then calculated. Combined with the stress-strain relationship, grain boundary stress characteristics, such as the distribution of stress concentration areas and stress gradients, are then determined.

[0017] In the formula σ=E·ε: σ represents stress, which is the force per unit area within an object. Its unit is Pascal (Pa). Here, it specifically refers to the stress at the grain boundaries of the reed. E represents the material's elastic modulus, also known as Young's modulus. It is a physical quantity that measures a material's resistance to elastic deformation. Its unit is Pascal (Pa). Its value is determined by the material's properties (such as the inherent properties of interstitial beryllium copper). ε represents strain, which is the relative deformation of an object under the action of an external force. It is a dimensionless quantity (usually expressed as a percentage or decimal). Here, it refers to the strain at the grain boundaries calculated using digital image correlation (DIC).

[0018] In the equation 2d·sinθ=nλ, d represents the interplanar spacing (IS), the perpendicular distance between two adjacent parallel crystal planes in a crystal, measured in meters (m) or nanometers (nm). It is a key parameter of the crystal structure. θ represents the Bragg angle, also known as the grazing angle, which is the angle between the incident X-ray beam and the crystal plane, measured in degrees (°). n represents the diffraction order, a positive integer (n=1,2,3,...) representing the number of times the X-rays are reflected between crystal planes. λ represents the wavelength of the incident X-rays, measured in meters (m) or nanometers (nm). It is determined by the characteristics of the synchrotron radiation source and, in this case, is the wavelength of the highly collimated X-ray beam.

[0019] Preferably, for step 13, when collecting element distribution and crystal orientation data, a spatial correlation and interactive analysis between the two should be established. Using an SEM equipped with an EDS and EBSD system, the same area of ​​the interstitial beryllium copper reed is simultaneously scanned: element distribution maps are obtained through EDS to clarify the compositional differences between different microregions (such as alloy element enrichment areas or impurity phase distribution). Simultaneously, the reed is tilted to 70°, and EBSD is used to collect Kikuchi diffraction patterns in this area to preliminarily determine the grain orientation. Using data overlay technology, the element distribution characteristics are spatially matched with the diffraction pattern—for example, identifying areas with sudden changes in element composition (possibly corresponding to grain boundaries or secondary phases), assisting the EBSD software in more accurately delineating grain boundaries. For the interior of grains with uniform composition, the diffraction pattern indexing algorithm is optimized by combining the influence of element type on the crystal structure (such as the fine-tuning of lattice parameters by solute atoms), thereby improving the accuracy of crystal orientation determination. Based on this interactively corrected EBSD data, when calculating the orientation difference between adjacent grains, we can focus on the orientation difference characteristics of the element segregation area (such as whether the element-enriched grain boundaries are more likely to form large-angle orientation differences). Finally, we can obtain the orientation difference distribution characteristics containing composition-structure correlation information, such as the difference in the proportion of small-angle grain boundaries and large-angle grain boundaries in a specific element distribution area.

[0020] Preferably, for step 15, the numpy.linalg.eig function is used to perform eigenvalue decomposition on the probability association matrix. The eigenvalues ​​and eigenvectors are solved using the QR iterative algorithm to obtain an orthogonal matrix V composed of eigenvectors and a diagonal matrix Λ containing the eigenvalues, ensuring that the eigenvalues ​​are arranged in descending order. An eigenvalue accumulator is constructed, and the proportion of each eigenvalue to the total sum of all eigenvalues ​​is calculated in sequence. The cumulative sum is then accumulated to generate a cumulative contribution rate sequence. The rate of change of adjacent contribution rates is simultaneously calculated to identify contribution rate mutation points. A dynamic contribution rate threshold interval is set between [85%, 95%]. The optimal threshold is automatically determined based on the rate of change of the cumulative contribution rate sequence. The first k eigenvalues ​​whose cumulative contribution rates exceed the dynamic contribution rate threshold interval are selected, and their corresponding eigenvectors are extracted as principal components to form a principal component matrix. An exponential decay weight function is constructed based on the eigenvalue size, and each vector in the principal component matrix is ​​weighted. A weighted principal component vector is generated through a vector dot product operation, and is output as a microstructure feature vector after L2 norm normalization.

[0021] Optionally, step 14, establishing a probability correlation matrix among grain size distribution characteristics, grain boundary stress characteristics, and orientation difference distribution characteristics, specifically includes the following steps: Step 141: Using grain size distribution characteristics, grain boundary stress characteristics, and orientation difference distribution characteristics as input variables, construct a joint probability distribution function of the three and generate a joint distribution model that describes the dependency relationship between the characteristics based on the function; Step 142: Calculate the Kendall rank correlation coefficient between each feature according to the joint distribution model to determine the conditional probability distribution between the features; Step 143: Map the probability association indicators between the features in the conditional probability distribution into matrix elements to generate a probability association matrix.

[0022] Preferably, for step 141, the three variables of grain size distribution, grain boundary stress, and misorientation distribution are used as input to construct a joint probability distribution function between them. By collecting a large amount of sample data and using methods such as maximum likelihood estimation, the parameters of the joint distribution function, such as the mean vector and covariance matrix, are estimated. In this way, a joint distribution model describing the dependency relationship between the three features is obtained. This model can reflect the change trend and probability distribution of the other two features when one of the features changes, providing a basis for subsequent analysis of the correlation between the features.

[0023] Preferably, for step 142, after obtaining the joint distribution model, the Kendall rank correlation coefficient is used to calculate the correlation between the grain size distribution characteristics, the grain boundary stress characteristics and the orientation difference distribution characteristics. The Kendall rank correlation coefficient is a non-parametric statistic that measures the correlation between variables based on the rank of the data. For each pair of features, such as grain size and grain boundary stress, the sample data are sorted, their ranks are calculated, and then the number of consistent pairs and inconsistent pairs is counted according to the calculation formula of the Kendall rank correlation coefficient to obtain the correlation coefficient between them. The value range of this correlation coefficient is between -1 and 1. The larger the absolute value, the stronger the correlation between the two features. By calculating the Kendall rank correlation coefficient between each pair of features, the conditional probability distribution between them can be determined, that is, the probability distribution of another feature when one feature takes a specific value.

[0024] Preferably, for step 143, the grain size distribution characteristics, grain boundary stress characteristics, and orientation difference distribution characteristics are preprocessed, the isolation forest algorithm is used to identify and eliminate outliers, the random forest regression model is used to fill missing values, and the Z-score normalization method is used to convert each feature data into a standard normal distribution with a mean of 0 and a standard deviation of 1 to obtain a standardized data set. Based on the standardized data set, the kendalltau function in the scipy.stats library is used to build a parallel computing framework to calculate the Kendall rank correlation coefficients between the grain size distribution characteristics and the grain boundary stress characteristics, the grain size distribution characteristics and the orientation difference distribution characteristics, and the grain boundary stress characteristics and the orientation difference distribution characteristics, respectively, to generate a set of correlation coefficient triplets. According to the symmetry principle of the Kendall rank correlation coefficient, a 3×3 symmetric matrix structure is constructed, and the elements in the correlation coefficient triplets are mapped to the corresponding non-diagonal positions of the matrix. At the same time, the matrix diagonal elements are uniformly assigned a value of 1, and finally a probability association matrix is ​​generated.

[0025] Optionally, step 2, calculating the magnetostrictive strain regulated by the microstructure based on the microstructure characteristic vector to determine a coupled field distribution description of the magnetostriction regulated by the microstructure, specifically comprises the following steps: Step 21: using the microstructure characteristic vector as an input parameter and inputting it into the constructed microstructure-magnetostriction nonlinear constitutive model to calculate the magnetostrictive strain of the synergistic effect of grain size, grain boundary stress and misorientation; Step 22: Generate a multi-physics field coupling interaction matrix based on the magnetostrictive strain; Step 23: Perform spatiotemporal fusion processing on the multi-physics field coupling interaction matrix to determine the coupling field distribution description of the microstructure-controlled magnetostriction.

[0026] Optionally, in step 21, the microstructure characteristic vector is used as an input parameter and input into the constructed microstructure-magnetostriction nonlinear constitutive model to calculate the magnetostrictive strain of the synergistic effect of grain size, grain boundary stress and orientation difference, which specifically includes the following steps: Step 211 , mapping the microstructure characteristic vector to the crystal elasticity theory framework to generate a microstructure corrected elastic constant matrix; Step 212: Input the microstructure-corrected elastic constant matrix into the microstructure-magnetostriction nonlinear constitutive model to perform thermodynamic potential function expansion to obtain a nonlinear magnetostrictive strain expression; Step 213: Generate a magnetostrictive strain tensor including microstructure parameters based on the nonlinear magnetostrictive strain expression; Step 214 : Perform multi-scale consistency correction and physical constraint optimization processing on the magnetostrictive strain tensor to calculate the magnetostrictive strain of the synergistic effect of grain size, grain boundary stress and orientation difference.

[0027] Preferably, for step 211, the microstructure feature vector is first decoupled, the grain size distribution characteristics therein are mapped into elastic modulus correction coefficients, the grain boundary stress characteristics are converted into shear modulus adjustment parameters, and the orientation difference distribution characteristics are generated into an anisotropy tensor. Based on the crystal periodic structure, the crystal elasticity theory framework that converts interatomic forces into macroscopic elastic parameters and reflects mechanical properties determined by structure such as anisotropy is constructed. A microstructure-sensitive elastic constant matrix generation model is constructed, and the anisotropy tensor is subjected to multi-scale weighted fusion and physical constraint correction processing: the principal components of the anisotropy tensor are scaled to obtain the corresponding shear components by combining the elastic modulus correction coefficient corresponding to the grain size; the shear modulus adjustment parameter associated with the grain boundary stress is introduced to dynamically correct the shear components to obtain the corrected shear components; and crystal symmetry constraints are embedded, and the corrected tensor is subjected to physical symmetry projection and thermodynamic stability constraint processing to integrate the multi-dimensional microstructure characteristics into the elastic constant matrix, generating an elastic constant matrix containing microstructure information such as grain size, grain boundary stress, and orientation difference.

[0028] Preferably, in step 212, a microstructure-magnetostriction nonlinear constitutive model is constructed based on the Landau free energy theory, and the microstructure-corrected elastic constant matrix is ​​used as the input parameter of the model. Energy terms such as magnetocrystalline anisotropy energy, elastic energy, and magnetoelastic coupling energy are introduced into the model to construct a complete thermodynamic potential function. With respect to this thermodynamic potential function, the influence of microstructure on magnetic domain wall movement and magnetization rotation is considered, and microstructure-sensitive energy correction terms are introduced to optimize it. The optimized thermodynamic potential function is subjected to a variational expansion, and combined with the Maxwell relation and constitutive equations, a nonlinear magnetostrictive strain expression including microstructure parameters is derived. This nonlinear magnetostrictive strain expression can reflect the complex nonlinear relationship between magnetic field, stress, and microstructure parameters.

[0029] Preferably, in step 213, the magnetic field intensity, direction, and external stress conditions in the actual working conditions are substituted into the nonlinear magnetostrictive strain expression to provide input parameters for subsequent calculations. Based on these input parameters, the computational domain is divided into a regular grid using a finite element discretization method to obtain a discrete computational model consisting of multiple grid cells. For each grid cell in the discrete model, the magnetostrictive strain component of each grid cell is calculated based on its position and orientation, combined with the local microstructure parameters in the microstructure eigenvector. Based on these strain components, considering the anisotropic properties of the material, a three-dimensional strain tensor structure is constructed, and the strain components in each direction are filled into the corresponding positions of the tensor. The strain values ​​at the grid boundaries in the three-dimensional strain tensor are processed using an interpolation algorithm to obtain a strain tensor with continuity and physical rationality. The processed strain tensor ultimately generates a magnetostrictive strain tensor containing spatial distribution information, which fully reflects the regulatory effect of the microstructure parameters on the magnetostrictive strain.

[0030] Preferably, in step 214, a cross-scale correlation model is established to couple the grain structure with the strain response at the macroscale to obtain a physically consistent strain tensor at different scales. With respect to the strain tensor, grain boundary constraints are introduced to reflect the hindering effect of grain boundaries on grain deformation, and the strain components near the grain boundaries are corrected. Based on the corrected strain components, a physical constraint optimization objective function is constructed based on the energy minimization principle so that the function can reflect the constraints of the basic laws of elastic mechanics and the physical properties of the material on the tensor. The Lagrange multiplier method is used to process the constraints in this objective function, and the optimal strain distribution is obtained by solving the problem through an iterative optimization algorithm. The optimal strain distribution is further processed to finally obtain the magnetostrictive strain after multi-scale consistency correction and physical constraint optimization, so as to achieve the accurate calculation of the synergistic effect of grain size, grain boundary stress and orientation difference.

[0031] Optionally, step 22, generating a multi-physics field coupling interaction matrix based on magnetostrictive strain, specifically includes the following steps: Step 221: derive the magnetic permeability change tensor caused by the magnetostrictive strain based on the influence of the magnetostrictive strain on the magnetic field distribution; Step 222: Calculate the Joule heat and mechanical energy dissipation changes generated by the stretching strain according to the coupling coefficient between the stress component and the temperature gradient, and generate a stress-temperature field change tensor accordingly. Step 223: Perform tensor fusion and nonlinear correction on the magnetic permeability change tensor and the stress-temperature field change tensor to generate a multi-physics field coupling interaction matrix.

[0032] Preferably, for step 221, the coupling relationship between magnetostrictive strain and magnetic field distribution is analyzed to clarify the mechanism by which the rearrangement of the internal magnetic domain structure affects the magnetization characteristics when the material generates magnetostrictive strain. Based on this influence mechanism and magnetoelastic coupling theory, a nonlinear mapping model between magnetostrictive strain and magnetic permeability is established. Based on this nonlinear mapping model, the finite element method is used to simulate the changes in magnetic domain distribution under different magnetostrictive strains, and the spatial distribution characteristics of magnetic domain orientation and magnetization intensity are extracted. Based on these spatial distribution characteristics, the magnetostrictive strain is decomposed into components in different directions, and the influence of each component on the anisotropy of magnetic permeability is analyzed. Based on micromagnetic theory and the influence of the above-mentioned components on the anisotropy of magnetic permeability, a magnetic permeability tensor that includes the magnetostrictive strain effect is derived. This magnetic permeability tensor can reflect the changes in magnetic permeability anisotropy caused by strain, thereby achieving a quantitative description of the influence of magnetostrictive strain on the magnetic field distribution.

[0033] Preferably, for step 222, the coupling effect between the stress field and the temperature field caused by magnetostrictive strain is considered. The coupling coefficient between the stress component and the temperature gradient is first determined. This coefficient is derived from a combination of experimental data and thermodynamic theory. Based on this coupling coefficient, the physical phenomena occurring when the material generates magnetostrictive strain are analyzed: stress concentration is generated internally, and hysteresis loss and eddy current loss are converted into Joule heating. In response to these phenomena, an energy conservation equation is established to convert the mechanical work and magnetic energy dissipation generated by the magnetostrictive strain into thermal energy, and the temperature field distribution within the material is calculated. Based on the temperature field distribution, a finite element thermal conduction analysis is performed to simulate the conduction process of the temperature gradient within the material, while also incorporating the effect of the stress field on the thermal conductivity coefficient. The changes in the stress components and temperature gradients obtained from the conduction analysis are used as tensor elements to construct a stress-temperature field change tensor. This stress-temperature field change tensor can reflect the coupled change characteristics of the stress field and temperature field caused by the magnetostrictive strain.

[0034] Preferably, for step 223, a multi-physics field tensor fusion model is established using the magnetic permeability change tensor and the stress-temperature field change tensor as the objects to account for the nonlinear coupling relationship between the two. The two tensors are dimensionalized and coordinate transformed to obtain comparable tensors in the same physical space. The comparable tensors are fused using a tensor product operation and a weighted summation method (the weight coefficients are calibrated based on the coupling strength of each physical field using experimental data) to obtain a preliminary fused tensor. A nonlinear correction term is introduced into the preliminary fused tensor (to account for nonlinear changes in the physical field caused by the magnetostrictive effect, such as the saturation characteristics of the magnetic permeability and the influence of temperature on the material's elastic modulus) to obtain a fused tensor that incorporates nonlinear factors. A nonlinear correction function is constructed to optimize the fused tensor that incorporates nonlinear factors, obtaining an optimized tensor that accurately reflects the coupling interactions between the multiple physical fields. The optimized tensor is generated as a multi-physics field coupling interaction matrix, which comprehensively describes the interactions between the magnetic field, stress field, and temperature field caused by magnetostrictive strain.

[0035] Optionally, step 23, performing spatiotemporal fusion processing on the multi-physics field coupling interaction matrix to determine a coupling field distribution description of microstructure-controlled magnetostriction, specifically includes the following steps: Step 231: performing time-scale decomposition on the multi-physics field coupling interaction matrix, and using wavelet packet transform to decompose the matrix elements into dynamic components and steady-state components in the time dimension to obtain a time-scale decomposition tensor; Step 232: performing spatial fractal interpolation processing on the time scale decomposition tensor to implement non-uniform grid division on the interface region between the iron core and the interstitial beryllium copper reed to generate a spatial scale characteristic matrix; Step 233: Based on the spatial scale characteristic matrix, a three-dimensional coupling operator including time, space and physical field is constructed to determine the coupling field distribution description of microstructure-controlled magnetostriction.

[0036] Preferably, for step 231, the time-varying characteristics of each element in the multi-physics field coupling interaction matrix are analyzed to clarify the dynamic characteristic differences of the magnetostrictive effect at different time scales. Based on this dynamic characteristic difference, a time scale decomposition model is established to provide a framework for subsequent time-frequency analysis. The multi-physics field coupling interaction matrix is ​​incorporated into the model, and the wavelet packet transform technology is used to perform time-frequency analysis on it. The time series signal of the matrix is ​​decomposed into sub-signals of different frequency bands to achieve multi-resolution analysis of the time characteristics of the matrix elements. For the time dimension of the matrix, an appropriate wavelet basis function (such as Daubechies wavelet) is selected for layer-by-layer decomposition to separate high-frequency dynamic components (such as transient coupling caused by rapid magnetic domain reversal) and low-frequency steady-state components (such as slow coupling caused by heat accumulation). Based on the decomposed frequency sub-bands, a time scale decomposition tensor is constructed, which retains the dynamic characteristics of physical field coupling at different time scales.

[0037] Preferably, for step 232, the physical field gradient characteristics (e.g., stress concentration and magnetic field distortion) at the interface between the core and the interstitial beryllium copper reed are analyzed to clarify the need for a fine spatial scale description of this region. The time-scale decomposition tensor is analyzed, and its spatial distribution characteristics are analyzed using fractal geometry theory. The degree of spatial irregularity of the physical field is characterized by calculating the fractal dimension. Adaptive interpolation is performed on the physical field parameters in the interface region based on a fractal interpolation algorithm. The interpolation density is increased in regions with large gradients (e.g., grain boundaries and contact surfaces), while the density is reduced in non-critical regions, achieving a non-uniform description of the spatial scale. The interpolated spatial distribution characteristics are combined with the principles of finite element meshing, and a non-uniform meshing strategy is adopted. A submicron-level fine mesh is used in the interface region, while a conventional mesh is used in non-interface regions. This constructs a spatial scale characteristic matrix. This spatial scale characteristic matrix achieves multi-resolution representation of the spatial scale by establishing a correlation between mesh density and physical field gradients, accurately capturing the strong coupling effects of the interface region.

[0038] Preferably, for step 233, a three-dimensional coupling model of time, space, and physical fields is established based on the spatial scale characteristic matrix and the time scale decomposition tensor, with the former serving as the spatial dimension input parameter and the latter providing time dimension information, thereby constructing a multiscale coupling operator. For this multiscale coupling operator, an interpolation kernel function is defined in three-dimensional space. Based on the grid density of the spatial scale characteristic matrix, the physical field coupling parameters of different regions are weighted, so that fine-grid regions obtain higher coupling weights. A time evolution operator is introduced based on the weighted physical field coupling parameters. Through convolution operations, the dynamic and steady-state components in the time scale decomposition tensor are associated with the spatial field distribution, achieving dynamic coupling in the time dimension. For the dynamically coupled model, interaction rules between the magnetic field, stress field, and temperature field are designed (such as stress-magnetic field coupling caused by magnetostrictive strain and temperature-stress coupling caused by Joule heating). These rules are embedded in the three-dimensional coupling operator to improve the operator's ability to describe multi-physical field interactions. By using the improved three-dimensional coupling operator, the multi-physical field coupling interaction matrix is ​​transformed in time and space to generate a coupling field distribution description that includes the microstructure regulation effect, realizing full-scale coupling characterization from time dynamics to spatial distribution.

[0039] Optionally, step 3, based on the coupled field distribution description, performs finite element meshing on the air domain around the high-frequency transformer to generate an air domain finite element mesh, specifically comprising the following steps: Step 31: Using the constructed air domain adaptive meshing criterion and based on the description of the coupled field distribution of microstructure-controlled magnetostriction, the core vibration displacement gradient and the area with intense acoustic pressure fluctuation are extracted as key location features; Step 32: Based on the key part characteristics, the air domain around the high-frequency transformer is divided into finite element meshes to generate an air domain finite element mesh.

[0040] Optionally, step 31, using the constructed air domain adaptive meshing criterion and based on the description of the coupled field distribution of microstructure-controlled magnetostriction, extracts the core vibration displacement gradient and the area with intense acoustic pressure fluctuation as key location features, specifically comprising the following steps: Step 311: Based on the physical field gradient energy density, a grid density allocation function is constructed as an air domain adaptive grid division criterion; Step 312: Calculate the magnetic field gradient energy density, stress gradient energy density, and acoustic pressure gradient energy density according to the coupled field distribution description to form a comprehensive gradient energy density field; Step 313: extract fractal features of the core surface and the air domain interface based on the grid density distribution function, and calculate the fractal dimension; Step 314: Calculate the core vibration displacement gradient and the area with intense sound pressure fluctuation as key location features based on the fractal dimension and the comprehensive gradient energy density field.

[0041] Preferably, for step 311, the gradient change rates of the magnetic field, stress field and acoustic pressure field are used as input parameters to establish a mapping relationship between the physical field gradient energy density and the grid density. Based on the mapping relationship, the influence weights of the gradient energy density of each physical field on the grid density are determined through experimental data calibration and numerical simulation verification. Using these influence weights, a nonlinear mapping function is designed to allocate a higher grid density in areas where the physical field gradient changes drastically, and a lower grid density in areas where the change is gentle. For the nonlinear mapping function, a regularization term is introduced to ensure its smoothness and avoid sudden changes in the grid density. The processed nonlinear mapping function forms a grid density allocation function, which can automatically adjust the grid distribution according to the physical field characteristics to achieve a balance between calculation accuracy and efficiency.

[0042] Preferably, for step 312, the coupled field distribution description of the microstructure-controlled magnetostriction obtained in step 2 is used as a basis to extract the spatial distribution data of the magnetic field, stress field, and acoustic pressure field respectively. Spatial differential operations are performed on these spatial distribution data to calculate the gradient change rate of each physical field, and the square of the gradient amplitude is used as the energy density index. Based on the characteristics of different physical fields, a normalization processing method is designed to process the energy density index to eliminate dimensional differences. The normalized gradient energy densities of the three physical fields are fused using a weighted summation method (the weight coefficient is determined based on the contribution of each physical field to acoustic radiation) to form a comprehensive gradient energy density field.

[0043] Preferably, in step 313, a spatial scan of the core surface and the air interface is performed based on a grid density distribution function to extract detailed geometric features of the interface. These detailed features are then subjected to fractal analysis using a box counting method, where the complexity of the interface is statistically analyzed using covering boxes of different scales. The relationship between box count and scale at different scales is calculated, and the fractal dimension is obtained through linear regression fitting.

[0044] Optionally, step 314, calculating the core vibration displacement gradient and the area with intense sound pressure fluctuation as key location features based on the fractal dimension and the comprehensive gradient energy density field, specifically includes the following steps: Step 3141: Substitute the fractal dimension into the grid density distribution function to obtain a grid density distribution based on geometric complexity; Step 3142: setting a physical field change threshold in combination with the comprehensive gradient energy density field, establishing a multi-dimensional correlation model, and performing tensor fusion of the grid density distribution and the physical field change threshold to obtain a physical-geometric coupling tensor field; Step 3143: Adaptive threshold segmentation is performed on the physical-geometric coupling tensor field to extract the areas where the core vibration displacement gradient exceeds the set gradient threshold and the areas where the sound pressure fluctuation energy density is higher than the fluctuation energy threshold, so as to determine the key part features.

[0045] Preferably, for step 3141, the fractal dimension of the core surface and the air domain interface calculated in step 313 is used as a geometric complexity indicator and input into the mesh density allocation function constructed in step 311. This function outputs a corresponding mesh density value based on the numerical value of the fractal dimension (a higher fractal dimension indicates a more complex regional geometry and potentially more drastic physical field changes) to generate a mesh density distribution. A higher mesh density is allocated to geometrically complex regions to accurately capture physical field changes, while a relatively sparse mesh is allocated to regions with lower fractal dimensions and simpler geometries.

[0046] Preferably, for step 3142, a statistical analysis is performed on the comprehensive gradient energy density field obtained in step 312, and a physical field change threshold is set based on its data distribution characteristics (such as mean and standard deviation) to distinguish between areas with drastic physical field changes and areas with gentle changes. With the physical field change threshold as a reference, a multidimensional correlation model is established to associate the grid density distribution based on geometric complexity with the threshold. The grid density distribution and the physical field change threshold tensor are fused through tensor operations to combine the grid density of each spatial position with the physical field change characteristics of that position. During the fusion process, a higher weight is given to areas with drastic physical field changes to highlight their importance in mesh division, forming a physical-geometric coupling tensor field. This physical-geometric coupling tensor field integrates the dual information of geometric morphology and physical field changes, and can more comprehensively reflect the key areas in the air domain that require fine mesh division.

[0047] Preferably, for step 3143, an adaptive threshold segmentation algorithm (such as the Otsu algorithm) is used to process the physical-geometric coupling tensor field. The algorithm automatically calculates the optimal segmentation threshold based on the data distribution of the physical-geometric coupling tensor field and divides the physical-geometric coupling tensor field into different regions. For the core vibration displacement gradient in these regions, the gradient value at each spatial position is compared with a pre-set gradient threshold, and regions where the vibration displacement gradient exceeds the threshold are screened out (such regions produce large vibrations due to the magnetostrictive effect, which significantly affects sound radiation). For the sound pressure fluctuation energy density in the region, the energy density value at each position is compared with the fluctuation energy threshold, and regions with violent fluctuations where the energy density exceeds the threshold are extracted. Through the above processing, regions with large core vibration displacement gradients and violent sound pressure fluctuations are accurately identified and determined as key location features in the air domain meshing.

[0048] Optionally, step 32, based on the key part characteristics, performs finite element meshing on the air domain around the high-frequency transformer to generate an air domain finite element mesh, specifically comprising the following steps: Step 321: construct a grid density mapping model driven by key part features to establish a nonlinear mapping relationship between physical field gradient energy density and grid size; Step 322: Generate a boundary layer mesh at the interface between the core and the air domain based on the nonlinear mapping relationship to capture the energy transfer effect characteristics of the structural vibration to the acoustic field; Step 323: Based on the energy transfer effect characteristics, perform finite element meshing on the air domain around the high-frequency transformer to generate an air domain finite element mesh.

[0049] Preferably, for step 321, the intrinsic connection between the characteristics of key parts and the changes in the physical field is analyzed, and the requirements for meshing accuracy in areas where the core vibration displacement gradient is large and the sound pressure fluctuation is severe are clarified. Simulation and experimental data under different physical field gradient energy densities are collected, and a nonlinear mapping model is constructed using the support vector regression (SVR) algorithm or artificial neural network (ANN) in machine learning. The physical field gradient energy density is used as the input of the model, and the target grid size is used as the output. The model is trained with a large amount of sample data to enable it to learn the complex mapping relationship between the degree of physical field change and the reasonable grid size. Regularization terms are introduced into the model to avoid overfitting, so that it can accurately output the optimal grid density corresponding to the physical field gradient under different working conditions, providing a quantitative basis for subsequent grid division.

[0050] Preferably, for step 322, based on the nonlinear mapping relationship established in step 321, a progressive mesh encryption strategy is adopted for the key energy transfer area, which is the interface between the iron core and the air domain. Based on the gradient energy density of the physical field near the interface, the initial mesh size of the boundary layer is determined by the mapping relationship. Based on the initial mesh size, the boundary layer mesh is generated by expanding layer by layer into the air domain according to a preset growth rate (such as geometric series growth), so that the mesh close to the surface of the iron core is finer, so as to accurately capture the high-frequency response generated by the structural vibration and the transient changes when transferring energy to the sound field. The generated boundary layer mesh is evaluated and optimized using a mesh quality inspection algorithm (such as Jacobian determinant inspection) to ensure the shape regularity of the mesh unit, avoid calculation errors caused by mesh distortion, and effectively improve the simulation accuracy of the energy transfer effect.

[0051] Preferably, for step 323, the boundary layer grid is used as the starting point, and the frontier advancing method or Delaunay triangulation algorithm is used to perform overall grid division in combination with the distribution of physical field gradient energy density in different regions within the air domain. For areas far away from the core and where the physical field changes slowly, a sparser grid is allocated according to the mapping relationship to reduce the amount of calculation while ensuring the accuracy of the calculation; for areas where the physical field changes drastically as identified by key part features, a fine grid is generated according to the mapping relationship. During the division process, the matching degree between the grid density and the physical field gradient is monitored in real time, and the areas that do not meet the accuracy requirements are corrected by the local grid re-division technology. The air domain finite element grid finally generated can not only accurately capture the sound propagation characteristics and energy transfer process caused by the vibration of the core, but also reasonably control the number of grids in non-critical areas, thereby achieving a balance between computational efficiency and simulation accuracy.

[0052] Optionally, step 4, equating the gap beryllium copper reed to an acoustic impedance boundary condition, and defining a pseudo-density variable in the air domain finite element mesh to determine the optimal acoustic wave propagation path tracking in the air domain, specifically includes the following steps: Step 41: Based on the influence of grain size distribution characteristics and grain boundary stress characteristics on the acoustic impedance of the material, the dynamic acoustic impedance of the gapped beryllium copper reed is derived, so as to treat the gapped beryllium copper reed as equivalent to the acoustic impedance boundary condition; Step 42: According to the distance relationship between the air domain finite element mesh and the boundary of the interstitial beryllium copper reed, a pseudo-density variable is defined in the air domain finite element mesh based on the dynamic acoustic impedance to obtain a pseudo-density distribution characteristic; Step 43: Consider each cell of the air domain finite element grid as a state space and the pseudo-density distribution characteristics as environmental parameters to define a reward function with the goal of minimizing the sound wave propagation loss; Step 44: Determine the optimal sound wave propagation path tracking in the air domain according to the reward function.

[0053] Preferably, for step 41, beryllium copper reed is used as the research object, and the sound velocity and density data under different grain sizes and grain boundary stress states are experimentally measured to establish a database of microstructure parameters - acoustic impedance. Based on this database, combined with mechanical theory, the influence of grain size on the sound wave scattering effect and the change of grain boundary stress on the elastic modulus and damping characteristics of the material are analyzed to clarify the mechanism of action of these factors on acoustic impedance. The random forest regression algorithm in machine learning is used, and the grain size distribution characteristics and grain boundary stress characteristics are used as input variables to fit a nonlinear relationship model between acoustic impedance and microstructure parameters. For this model, a frequency domain correction factor is introduced to consider the dynamic influence of sound wave frequency on acoustic impedance, and a dynamic acoustic impedance expression that can reflect the characteristics under different working conditions is derived. This dynamic acoustic impedance expression equates the gap beryllium copper reed to an acoustic impedance boundary condition with frequency and microstructure sensitivity.

[0054] Preferably, for step 42, a distance field is established to calculate the shortest distance from the center of each cell in the air domain finite element mesh to the boundary of the interstitial beryllium copper reed. Based on this distance field and combined with the dynamic acoustic impedance obtained in step 41, an attenuation function is designed so that the pseudo-density variable of the mesh cell closer to the boundary approaches the dynamic acoustic impedance of the reed. As the distance increases, the value of the pseudo-density variable gradually decreases according to the attenuation function until it approaches the acoustic impedance of air. This attenuation function is used to assign a corresponding pseudo-density value to each cell in the entire air domain finite element mesh, forming a pseudo-density distribution characteristic that reflects the acoustic influence of the boundary and provides environmental parameters for subsequent acoustic wave propagation path tracking.

[0055] Preferably, for step 44, a path search algorithm from reinforcement learning (such as Q-learning or a deep Q-network (DQN)) is employed, using the cells of the air domain finite element mesh as state nodes and treating the propagation of sound waves between mesh cells as a state transition process. During the algorithm initialization phase, a starting cell is randomly selected as the starting point for the sound wave. For each state transition, the reward value for all neighboring cells is calculated based on the pseudo-density distribution characteristics of the current cell and a reward function. The reward value reflects the magnitude of the sound wave loss when propagating in that direction. The algorithm preferentially selects the neighboring cell with the highest reward value (i.e., the lowest sound wave propagation loss) as the next state, thereby gradually extending the path. Through multiple iterative training, the algorithm continuously optimizes the path selection strategy and gradually converges to the optimal propagation path from the starting point to the target point. During the training process, an experience replay mechanism and a target network update strategy are introduced to prevent the algorithm from falling into local optimal solutions. Ultimately, the optimal path that truly reflects the sound wave propagation in the air domain, considering the influence of the gap between the beryllium copper reeds, is obtained, i.e., the optimal sound wave propagation path.

[0056] Optionally, step 43 considers each cell of the air domain finite element mesh as a state space and uses the pseudo-density distribution characteristics as environmental parameters to define a reward function with the goal of minimizing the sound wave propagation loss, specifically including the following steps: Step 431: Consider each cell of the air domain finite element grid as a state space and the pseudo-density distribution characteristics as environmental parameters to calculate the acoustic wave energy attenuation per unit distance within each cell of the air domain finite element grid; Step 432: Determine the acoustic wave propagation direction, the acoustic impedance difference between cells, and the pseudo-density gradient based on the acoustic wave energy attenuation per unit distance within each cell of the air domain finite element grid; Step 433: Define a reward function that aims to minimize the sound wave propagation loss based on the sound wave propagation direction, the acoustic impedance difference between units, and the pseudo-density gradient.

[0057] Preferably, for step 431, each cell of the air domain finite element mesh is used as an independent analysis unit, and its pseudo-density value is extracted as an environmental parameter. This parameter can reflect the acoustic impact of the interstitial beryllium copper reed boundary on the air domain (the closer the cell is to the boundary, the higher the pseudo-density value, and the corresponding sound wave propagation loss is greater). Based on the theory of acoustic energy attenuation, the sound wave attenuation rate at different pseudo-density values ​​is experimentally calibrated to form a lookup table or interpolation model, and a mapping relationship between pseudo-density and sound wave attenuation coefficient is established. The pseudo-density value of each cell is substituted into the mapping model to calculate the sound wave energy attenuation coefficient per unit distance within the cell. This coefficient represents the proportion of energy lost per meter of sound wave propagation when passing through the cell.

[0058] Preferably, for step 432, the spatial positional relationship between the current cell and adjacent cells is analyzed, with the current cell as the center, to determine possible sound wave propagation directions (e.g., six adjacent directions in three-dimensional space). For each propagation direction, the pseudo-density difference between the current cell and adjacent cells is calculated. This difference reflects the "acoustic impedance difference" between cells (the more dramatic the pseudo-density change, the greater the sound wave reflection and scattering loss). Using the finite difference method, the pseudo-density gradient of the current cell in each propagation direction is calculated. The larger the absolute value of the gradient, the more significant the change in the acoustic environment in that direction. The three parameters—propagation direction, inter-cell pseudo-density difference (i.e., equivalent acoustic impedance difference), and pseudo-density gradient—are normalized to obtain unified-dimensional parameters. Based on these unified-dimensional parameters, a multidimensional feature vector is formed to quantitatively describe the acoustic loss characteristics of each propagation direction.

[0059] Preferably, for step 433, a nonlinear reward function is designed based on the propagation direction eigenvector obtained in step 432. The inter-unit acoustic impedance difference and pseudo-density gradient are used as negative reward factors (greater differences result in lower rewards), and the energy attenuation coefficient in the propagation direction is used as the core penalty term. A weighted summation approach is used to construct the function: reward = -α × (acoustic impedance difference) - β × (pseudo-density gradient) - γ × (energy attenuation per unit distance), where α, β, and γ are weight coefficients calibrated using experimental data, reflecting the priority of each factor's impact on acoustic wave loss. The function value is normalized to ensure that the reward value falls within a reasonable range (e.g., [-1, 0]). Smaller negative values ​​indicate lower acoustic wave loss in that propagation direction, corresponding to higher rewards. The resulting reward function guides the algorithm to prioritize propagation paths with smooth acoustic impedance variations, small pseudo-density gradients, and low energy attenuation.

[0060] Optionally, step 5, calculating the sound pressure of the high-frequency transformer based on the tracked sound wave propagation path, specifically includes the following steps: Step 51: Discretize the sound wave propagation path into a series of acoustic propagation units to calculate the sound energy transmission coefficient between adjacent acoustic propagation units and generate a path energy attenuation coefficient matrix based on the coefficient; Step 52: Determine the sound pressure distribution taking into account the multipath interference effect based on the path energy attenuation coefficient matrix; Step 53: Based on the description of the coupling field distribution of the microstructure-controlled magnetostriction, the vibration velocity distribution of the surface of the gap beryllium copper reed is extracted; Step 54 : Perform acoustic-solid coupling boundary integration processing on the sound pressure distribution of the multipath interference effect and the vibration velocity distribution on the surface of the interstitial beryllium copper reed to calculate the sound pressure of the high-frequency transformer.

[0061] Preferably, for step 51, the tracked continuous sound wave propagation path is discretized according to the spatial position, and the path is divided into a series of interconnected acoustic propagation units based on the size and shape of the air domain finite element grid. Each unit is regarded as an independent sound wave propagation medium, and its characteristics are determined by physical parameters such as the air density and sound speed at its location. For adjacent acoustic propagation units, the energy transfer characteristics of the sound wave at the unit interface are analyzed based on the acoustic interface transmission theory. Since there may be differences in acoustic impedance between different units, reflection and transmission phenomena will occur when the sound wave propagates to the interface. By calculating the acoustic impedance ratio of adjacent units, combining Snell's law and acoustic boundary conditions, the sound energy transmission coefficient of the sound wave at the interface is determined. This coefficient represents the proportion of energy that can continue to propagate forward when the sound wave crosses the interface. The higher the transmission coefficient, the smaller the energy loss. The acoustic energy transmission coefficients between all adjacent units are filled into the corresponding positions of the matrix in the order of the propagation path to construct a path energy attenuation coefficient matrix. The rows and columns of the matrix correspond to different acoustic propagation units, and the elements on the diagonal are set to 1 by default (indicating no energy attenuation inside the unit), while the non-diagonal elements record the acoustic energy transmission coefficients between adjacent units.

[0062] Optionally, step 52, determining the sound pressure distribution taking into account the multipath interference effect based on the path energy attenuation coefficient matrix, specifically includes the following steps: Step 521: Calculate the propagation time of the sound wave on the sound wave propagation path based on the physical parameters of each sound wave propagation unit in the path energy attenuation coefficient matrix to determine the phase delay and generate a phase delay matrix accordingly; Step 522: Calculate the sound pressure attenuation factor based on the path energy attenuation coefficient matrix and perform a tensor product operation with the phase delay matrix to obtain sound pressure description data; Step 523: Use each grid node in the air domain finite element grid as a sound wave receiving point to generate a sound pressure distribution matrix that takes into account the multipath interference effect based on the sound pressure description data.

[0063] Preferably, for step 521, based on the path energy attenuation coefficient matrix, the physical parameters corresponding to each acoustic propagation unit in the matrix (including information such as the air density and sound speed within the unit) are extracted. These parameters directly affect the propagation speed of the sound wave within the unit (the faster the sound speed, the shorter the time it takes for the sound wave to traverse the unit). The length of each unit is divided by the corresponding sound speed to calculate the propagation time of the sound wave within each unit. The propagation time of each unit along the path is accumulated to obtain the total propagation time of the sound wave from the starting point to the end point. Based on the relationship between time and phase in the wave equation (the phase change of the sound wave is directly related to its propagation time), the total propagation time is converted into a phase delay. This phase delay reflects the phase change caused by the accumulation of time during the propagation of the sound wave and is a key parameter for analyzing multipath interference effects. The phase delay corresponding to each propagation unit is filled into the corresponding position in the matrix to generate a phase delay matrix.

[0064] Preferably, for step 522, the path energy attenuation coefficient matrix is ​​used as the analysis object. The physical relationship between energy and sound pressure (the greater the energy attenuation, the more significant the drop in sound pressure amplitude) is utilized to convert the acoustic energy transmission coefficients between each unit in the matrix into corresponding sound pressure attenuation factors. This factor represents the degree of attenuation of the sound pressure amplitude relative to the initial value after the sound wave passes through a specific unit or interface. A tensor product operation is performed on these sound pressure attenuation factors with the phase delay matrix generated in step 521. This tensor product combines the sound pressure attenuation information with the phase delay information, simultaneously considering the effects of amplitude attenuation and phase change on sound pressure. This operation combines the sound pressure attenuation factor on each propagation path with the corresponding phase delay to generate sound pressure description data containing amplitude and phase information (the data is presented in complex form, with the real part corresponding to the sound pressure amplitude and the imaginary part corresponding to the phase information).

[0065] Optionally, step 523, using each grid node in the air domain finite element grid as a sound wave receiving point to generate a sound pressure distribution matrix taking into account the multipath interference effect based on the sound pressure description data, specifically includes the following steps: Step 5231: traverse all mesh nodes in the air domain finite element mesh and use each mesh node as an acoustic wave receiving point; Step 5232: Calculate the path length difference based on the spatial position relationship between the sound wave receiving point and the sound wave propagation path, and then obtain the additional phase difference between the paths; Step 5233: perform coherent superposition on the additional phase difference and sound pressure description data to generate a sound pressure distribution matrix that takes into account the multipath interference effect.

[0066] Optionally, step 53, based on the description of the coupling field distribution of microstructure-controlled magnetostriction, extracting the vibration velocity distribution of the surface of the gap beryllium copper reed, specifically includes the following steps: Step 531: Based on the description of the coupled field distribution of microstructure-controlled magnetostriction, determine the grain orientation difference distribution and grain boundary stress concentration coefficient for the gap beryllium copper spring; Step 532: Based on the grain orientation difference distribution and the grain boundary stress concentration factor, a representative volume element model is used to determine the vibration response of the gapped beryllium copper reed. Step 533: Based on the vibration response, calculate the dynamic strain energy density of each point on the reed surface to extract the vibration velocity distribution on the surface of the gapped beryllium copper reed.

[0067] Preferably, in step 531, based on the description of the coupled field distribution of microstructurally regulated magnetostriction, a gapped beryllium copper reed is scanned using materials characterization techniques such as electron backscatter diffraction (EBSD) to obtain crystallographic orientation data for each grain. By comparing the crystallographic orientation data of adjacent grains, the grain orientation difference is calculated, and a grain orientation difference distribution map for the entire reed is generated. This map can intuitively reflect the degree of irregularity in grain arrangement. To calculate the grain boundary stress concentration coefficient, a finite element analysis model is constructed, incorporating microstructural features (such as grain size and grain boundary shape) and the magnetostrictive effect. Based on this model, the stress distribution within the reed under the action of a magnetic field is simulated, focusing on the grain boundaries (where magnetostrictive stress tends to accumulate due to the irregular atomic arrangement). Stress analysis of the grain boundaries is used to calculate the degree of stress concentration and determine the grain boundary stress concentration coefficient, which quantifies the stress amplification effect of the grain boundaries.

[0068] Preferably, for step 532, a representative volume element (RVE) model is selected that contains a sufficient number of grains and grain boundaries to accurately reflect the structural characteristics of the reed. The grain misorientation distribution and grain boundary stress concentration factor obtained in step 531 are assigned as input parameters to the RVE model. A multi-physics coupling equation is established by combining the magnetostrictive effect and electromagnetic force. Numerical calculation methods are used to solve this multi-physics coupling equation, simulating the deformation and vibration caused by magnetostriction within the RVE model under high-frequency magnetic field excitation. The model's dynamic response is analyzed to obtain vibration parameters such as displacement, velocity, and acceleration of the RVE at different times. The vibration parameters of the RVE model are statistically averaged and extrapolated, and the results are mapped to the entire gapped beryllium copper reed to determine the overall vibration response of the reed.

[0069] Preferably, for step 533, the reed surface is divided into multiple micro-regions based on the reed vibration response obtained in step 532. For each micro-region, the strain of the region is calculated based on the displacement change during the vibration process using geometric relationships and elastic mechanics theory. Combined with parameters such as the elastic modulus of the material, the dynamic strain energy density of each region is calculated based on the strain results (this value reflects the energy stored per unit volume during the vibration process). Based on the principle of conservation of energy and kinetic equations, a conversion relationship between dynamic strain energy density and vibration velocity is established using the inherent connection between the two. Through this conversion relationship, the dynamic strain energy density of each region is converted into the corresponding vibration velocity. Using a spatial interpolation method, the vibration velocity of the discrete region is expanded to the entire reed surface, generating a continuous and accurate vibration velocity distribution.

[0070] Optionally, step 54, performing acoustic-solid coupling boundary integral processing on the sound pressure distribution of the multipath interference effect and the vibration velocity distribution on the surface of the interstitial beryllium copper reed to calculate the sound pressure of the high-frequency transformer, specifically includes the following steps: Step 541: Using the vibration velocity distribution as an acoustic boundary condition, introducing the three-dimensional free space Green's function to establish a frequency domain sound pressure integral expression; Step 542: Based on the frequency domain sound pressure integral expression, perform acoustic-solid coupling boundary integral processing on the sound pressure distribution of the multipath interference effect to generate an acoustic-solid coupling correction term; Step 543: Perform complex superposition processing on the acoustic-solid coupling correction term and the sound pressure distribution to obtain the sound pressure field of the high-frequency transformer.

[0071] Preferably, for step 541, the vibration velocity distribution is defined as an acoustic boundary condition, and the continuity relationship between the vibration velocity of the reed surface and the normal velocity of the air-domain particle is clarified. Based on this continuity relationship, a three-dimensional free-space Green's function is introduced for the air-domain sound field, and is used to characterize the sound pressure response characteristics generated by a unit point sound source at any position in space. The Green's function is combined with the acoustic boundary condition, and the sound pressure contributions of each vibration point on the reed surface are superimposed through an integral operation to obtain a quantitative relationship between the sound pressure at any field point in the air domain and the reed vibration velocity distribution. This quantitative relationship is presented in the form of an integral, forming a frequency-domain sound pressure integral expression that describes the sound field radiated by the entire vibrating surface.

[0072] Preferably, for step 542, after establishing the frequency domain sound pressure integral expression, the sound pressure distribution obtained in step 52 taking into account the multi-path interference effect is combined to perform acoustic-solid coupling boundary integral processing. The multi-path interference effect reflects the phenomenon of sound waves superimposing on each other after propagating along different paths in the air domain, while the acoustic-solid coupling takes into account the influence of the reed vibration on the sound field. In the specific operation, the sound pressure distribution of the multi-path interference effect is used as the initial sound field. On this basis, the frequency domain sound pressure integral expression in step 541 is used to calculate the additional sound pressure contribution caused by the reed vibration. Through the boundary integral method, the vibration velocity information of the reed surface is coupled with the initial sound field to quantify the correction effect of the vibration on the sound field. This process takes into account the interaction between vibration and sound wave propagation, and finally generates an acoustic-solid coupling correction term, which reflects the adjustment of the acoustic-solid coupling effect to the original sound pressure distribution.

[0073] Preferably, for step 543, since the sound pressure distribution and the acoustic-solid coupling correction term are both represented in complex form in the frequency domain (the real part represents the sound pressure amplitude, and the imaginary part represents the phase information), complex superposition processing is required to obtain the final sound pressure field. Complex superposition can simultaneously account for changes in amplitude and phase, accurately reflecting the sound pressure distribution under the combined effects of acoustic-solid coupling and multipath interference. In specific implementation, the acoustic-solid coupling correction term generated in step 542 is added point by point to the sound pressure distribution obtained in step 52, which takes into account the multipath interference effect. For each node in the air domain finite element mesh, the complex correction term at the corresponding position is added to the complex original sound pressure distribution to obtain a new complex sound pressure value. Through this operation, the dual effects of vibration radiation and acoustic wave interference are integrated, ultimately obtaining a high-frequency transformer sound pressure field that includes the acoustic-solid coupling effect, achieving a more accurate simulation of the actual sound field.

[0074] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for calculating the sound pressure of a high-frequency transformer, characterized in that: include: Step 1, obtaining microstructure parameters of the gap beryllium copper reed to generate a microstructure feature vector; Step 2: Based on the microstructure characteristic vector, the magnetostrictive strain regulated by the microstructure is calculated to determine the coupled field distribution description of the magnetostriction regulated by the microstructure; Step 3: Based on the description of the coupled field distribution, the air domain around the high-frequency transformer is divided into finite element meshes to generate an air domain finite element mesh; Step 4: Equivalent the gap beryllium copper reed to an acoustic impedance boundary condition, and define a pseudo-density variable in the air domain finite element mesh to determine the optimal acoustic wave propagation path tracking in the air domain; Step 5: Calculate the sound pressure of the high-frequency transformer based on the traced sound wave propagation path.

2. The method according to claim 1, characterized in that Step 1 specifically includes the following steps: Step 11: reconstructing the micron-level three-dimensional microstructure of the interstitial beryllium copper reed to generate a scanning electron microscope image and calculate the grain size distribution characteristics based on the image; Step 12: Perform in-situ stress testing on the interstitial beryllium copper spring to obtain grain boundary stress distribution and generate synchrotron X-ray diffraction data to extract grain boundary stress characteristics; Step 13: collecting element distribution and crystal orientation data of the interstitial beryllium copper spring to generate electron backscatter diffraction data and generate orientation difference distribution characteristics based on the data; Step 14: Establish a probability correlation matrix among grain size distribution characteristics, grain boundary stress characteristics, and orientation difference distribution characteristics; Step 15: Perform eigenvalue decomposition on the probability association matrix to extract the principal components whose cumulative contribution rate is greater than the contribution rate threshold and generate a microstructure feature vector based on the principal components.

3. The method according to claim 1, characterized in that Step 2 specifically includes the following steps: Step 21: using the microstructure characteristic vector as an input parameter and inputting it into the constructed microstructure-magnetostriction nonlinear constitutive model to calculate the magnetostrictive strain of the synergistic effect of grain size, grain boundary stress and misorientation; Step 22: Generate a multi-physics field coupling interaction matrix based on the magnetostrictive strain; Step 23: Perform spatiotemporal fusion processing on the multi-physics field coupling interaction matrix to determine the coupling field distribution description of the microstructure-controlled magnetostriction.

4. The method according to claim 3, characterized in that Step 21 specifically includes the following steps: Step 211 , mapping the microstructure characteristic vector to the crystal elasticity theory framework to generate a microstructure corrected elastic constant matrix; Step 212: Input the microstructure-corrected elastic constant matrix into the microstructure-magnetostriction nonlinear constitutive model to perform thermodynamic potential function expansion to obtain a nonlinear magnetostrictive strain expression; Step 213: Generate a magnetostrictive strain tensor including microstructure parameters based on the nonlinear magnetostrictive strain expression; Step 214 : Perform multi-scale consistency correction and physical constraint optimization processing on the magnetostrictive strain tensor to calculate the magnetostrictive strain of the synergistic effect of grain size, grain boundary stress and orientation difference.

5. The method according to claim 3, characterized in that Step 22 specifically includes the following steps: Step 221: derive the magnetic permeability change tensor caused by the magnetostrictive strain based on the influence of the magnetostrictive strain on the magnetic field distribution; Step 222: Calculate the Joule heat and mechanical energy dissipation changes generated by the stretching strain according to the coupling coefficient between the stress component and the temperature gradient, and generate a stress-temperature field change tensor accordingly. Step 223: Perform tensor fusion and nonlinear correction on the magnetic permeability change tensor and the stress-temperature field change tensor to generate a multi-physics field coupling interaction matrix.

6. The method according to claim 1, characterized in that Step 3 specifically includes the following steps: Step 31: Using the constructed air domain adaptive meshing criterion and based on the description of the coupled field distribution of microstructure-controlled magnetostriction, the core vibration displacement gradient and the area with intense acoustic pressure fluctuation are extracted as key location features; Step 32: Based on the key part characteristics, the air domain around the high-frequency transformer is divided into finite element meshes to generate an air domain finite element mesh.

7. The method according to claim 6, characterized in that Step 31 specifically includes the following steps: Step 311: Based on the physical field gradient energy density, a grid density allocation function is constructed as an air domain adaptive grid division criterion; Step 312: Calculate the magnetic field gradient energy density, stress gradient energy density, and acoustic pressure gradient energy density according to the coupled field distribution description to form a comprehensive gradient energy density field; Step 313: extract fractal features of the core surface and the air domain interface based on the grid density distribution function, and calculate the fractal dimension; Step 314: Calculate the core vibration displacement gradient and the area with intense sound pressure fluctuation as key location features based on the fractal dimension and the comprehensive gradient energy density field.

8. The method according to claim 6, characterized in that Step 32 specifically includes the following steps: Step 321: construct a grid density mapping model driven by key part features to establish a nonlinear mapping relationship between physical field gradient energy density and grid size; Step 322: Generate a boundary layer mesh at the interface between the core and the air domain based on the nonlinear mapping relationship to capture the energy transfer effect characteristics of the structural vibration to the acoustic field; Step 323: Based on the energy transfer effect characteristics, perform finite element meshing on the air domain around the high-frequency transformer to generate an air domain finite element mesh.

9. The method according to claim 2, characterized in that Step 4 specifically includes the following steps: Step 41: Based on the influence of grain size distribution characteristics and grain boundary stress characteristics on the acoustic impedance of the material, the dynamic acoustic impedance of the gapped beryllium copper reed is derived, so as to treat the gapped beryllium copper reed as equivalent to the acoustic impedance boundary condition; Step 42: According to the distance relationship between the air domain finite element mesh and the boundary of the interstitial beryllium copper reed, a pseudo-density variable is defined in the air domain finite element mesh based on the dynamic acoustic impedance to obtain a pseudo-density distribution characteristic; Step 43: Consider each cell of the air domain finite element grid as a state space and the pseudo-density distribution characteristics as environmental parameters to define a reward function with the goal of minimizing the sound wave propagation loss; Step 44: Determine the optimal sound wave propagation path tracking in the air domain according to the reward function.

10. The method according to claim 1, characterized in that Step 5 specifically includes the following steps: Step 51: Discretize the sound wave propagation path into a series of acoustic propagation units to calculate the sound energy transmission coefficient between adjacent acoustic propagation units and generate a path energy attenuation coefficient matrix based on the coefficient; Step 52: Determine the sound pressure distribution taking into account the multipath interference effect based on the path energy attenuation coefficient matrix; Step 53: Based on the description of the coupling field distribution of the microstructure-controlled magnetostriction, the vibration velocity distribution of the surface of the gap beryllium copper reed is extracted; Step 54 : Perform acoustic-solid coupling boundary integration processing on the sound pressure distribution of the multipath interference effect and the vibration velocity distribution on the surface of the interstitial beryllium copper reed to calculate the sound pressure of the high-frequency transformer.

Images