Near-field radiation source space pre-marking method and device, electronic equipment and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for spatial pre-calibration of near-field radiation sources. The method includes: screening strong radiation grid points in near-field scanning data where the electromagnetic field intensity is higher than the background noise threshold to form a set of strong radiation contribution points; dynamically adjusting the number of clusters to match the number of equivalent dipoles based on the radiation energy distribution characteristics of this point set and a preset number of equivalent dipoles; performing K-means clustering with the adjusted number of clusters to automatically divide the strong radiation region into spatial clusters equal to the number of equivalent dipoles; and outputting the geometric boundary of each spatial cluster. This method accurately solves the core technical problems of traditional near-field radiation source reconstruction, such as the lack of effective spatial pre-calibration, low efficiency of blind full-plane search, poor reconstruction accuracy, difficulty in adapting to low-cost, phase-free testing scenarios, and inability to conform to the radiation distribution characteristics of different devices under test. It achieves efficient, accurate, and highly adaptable spatial pre-calibration of near-field radiation sources.

Description

Technical Field

[0001] This application relates to the field of electromagnetic compatibility prediction, and in particular to a method, apparatus, electronic device and storage medium for spatial precalibration of near-field radiation sources. Background Technology

[0002] With the rapid development of electronic devices towards miniaturization, high frequency, and high integration, electromagnetic interference (EMI) problems are becoming increasingly prominent, directly affecting the normal operation of equipment and electromagnetic compatibility (EMC) compliance. Near-field radiation source reconstruction and far-field prediction technologies have become core methods for EMC design and optimization of electronic equipment due to their elimination of the need for large microwave anechoic chambers and low testing costs.

[0003] Existing schemes typically select magnetic dipoles or electric dipoles as equivalent sources and solve for the equivalent source parameters using inversion algorithms such as least squares method and genetic algorithm, thereby deriving far-field radiation characteristics. This scheme achieves equivalent source reconstruction based on near-field data.

[0004] However, the initial position of the equivalent dipole in the above methods is mostly randomly set in the full scanning plane. The optimal solution is found only by the algorithm's own global search capability. The initial position of the equivalent dipole has no spatial constraints and requires a large number of invalid iterations in low-radiation or even non-radiation regions, resulting in low reconstruction efficiency, serious waste of computational resources, and affecting the accuracy of subsequent far-field prediction. Summary of the Invention

[0005] This application provides a near-field radiation source spatial pre-calibration method, device, electronic device, and storage medium to provide accurate spatial constraints for equivalent source reconstruction, thereby improving reconstruction efficiency and accuracy while adapting to phase-free data scenarios.

[0006] In a first aspect, embodiments of this application provide a spatial pre-calibration method for near-field radiation sources, including:

[0007] Strong radiation grid points with electromagnetic field strength higher than the background noise threshold in near-field scanning data are selected to form a set of strong radiation contribution points.

[0008] Based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles, the number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles.

[0009] K-means clustering is performed with the adjusted number of clusters to automatically divide the strong radiation region into spatial clusters with the same number of equivalent dipoles, and the geometric boundary of each spatial cluster is output. The geometric boundary is used to constrain the position search range of each equivalent dipole in the subsequent equivalent source reconstruction process from the full scan plane to its corresponding spatial cluster.

[0010] In one possible implementation, dynamically adjusting the number of clusters to be the same as the number of equivalent dipoles based on the radiation energy distribution characteristics of the strong radiation contribution point set and a preset number of equivalent dipoles includes:

[0011] Analyze the radiation energy distribution characteristics to determine the initial number of clusters;

[0012] Perform K-means clustering on the set of strong radiation contribution points and calculate the radiation contribution of each cluster;

[0013] The initial number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles based on all radiation contributions.

[0014] In one possible implementation, the radiation energy distribution characteristics include at least one of the radiation energy entropy value of the set of strong radiation contribution points and the spatial density distribution, wherein the radiation energy entropy value is used to characterize the degree of concentration of radiation energy, and the spatial density distribution is used to characterize the aggregation pattern of the strong radiation region in the scanning plane.

[0015] In one possible implementation, analyzing the radiation energy distribution characteristics of the strong radiation contribution point set and determining the initial number of clusters includes:

[0016] The initial number of clusters is determined based on the radiation energy distribution characteristics and a preset mapping relationship, wherein the mapping relationship includes the correspondence between the radiation energy distribution characteristics and the initial number of clusters.

[0017] In one possible implementation, performing K-means clustering on the set of strong radiation contribution points to calculate the radiation contribution of each cluster includes:

[0018] For each cluster, the radiation contribution of the cluster is calculated by the ratio of the sum of the radiation intensities of all grid points within the cluster to the sum of the radiation intensities of all grid points with full intensity radiation.

[0019] In one possible implementation, dynamically adjusting the initial cluster number to be the same as the equivalent dipole number based on all radiation contributions includes:

[0020] If the initial number of clusters is greater than the number of equivalent dipoles, then adjacent clusters are merged in order of radiation contribution from low to high until the number of clusters is reduced to the number of equivalent dipoles.

[0021] If the initial number of clusters is less than the number of equivalent dipoles, the corresponding clusters are split in descending order of radiation contribution until the number of clusters increases to the number of equivalent dipoles.

[0022] If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters.

[0023] In one possible implementation, the method further includes:

[0024] Phase-free magnetic field amplitude data within a preset fixed distance range between the scanning plane and the surface of the device under test are collected as the near-field scanning data.

[0025] Secondly, embodiments of this application provide a near-field radiation source spatial pre-calibration device, comprising:

[0026] The filtering module is used to filter out strong radiation grid points in near-field scanning data whose electromagnetic field intensity is higher than the background noise threshold, forming a set of strong radiation contribution points.

[0027] The dynamic adjustment module is used to dynamically adjust the number of clusters to be the same as the number of equivalent dipoles based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles.

[0028] The division output module is used to perform K-means clustering with the adjusted number of clusters, automatically dividing the strong radiation region into spatial clusters with the same number of equivalent dipoles, and outputting the geometric boundary of each spatial cluster. The geometric boundary is used to constrain the position search range of each equivalent dipole in the subsequent equivalent source reconstruction process from the full scan plane to its corresponding spatial cluster.

[0029] In one possible implementation, the dynamic adjustment module includes:

[0030] Analyze the radiation energy distribution characteristics to determine the initial number of clusters;

[0031] Perform K-means clustering on the set of strong radiation contribution points and calculate the radiation contribution of each cluster;

[0032] The initial number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles based on all radiation contributions.

[0033] In one possible implementation, the radiation energy distribution characteristics include at least one of the radiation energy entropy value of the set of strong radiation contribution points and the spatial density distribution, wherein the radiation energy entropy value is used to characterize the degree of concentration of radiation energy, and the spatial density distribution is used to characterize the aggregation pattern of the strong radiation region in the scanning plane.

[0034] In one possible implementation, the dynamic adjustment module analyzes the radiation energy distribution characteristics of the strong radiation contribution point set to determine the initial number of clusters, specifically including:

[0035] The initial number of clusters is determined based on the radiation energy distribution characteristics and a preset mapping relationship, wherein the mapping relationship includes the correspondence between the radiation energy distribution characteristics and the initial number of clusters.

[0036] In one possible implementation, the dynamic adjustment module performs K-means clustering on the strong radiation grid points to calculate the radiation contribution of each cluster, specifically including:

[0037] For each cluster, the radiation contribution of the cluster is calculated by the ratio of the sum of the radiation intensities of all grid points within the cluster to the sum of the radiation intensities of all grid points with full intensity radiation.

[0038] In one possible implementation, the dynamic adjustment module dynamically adjusts the initial cluster number to be the same as the equivalent dipole number based on all radiation contributions, including:

[0039] If the initial number of clusters is greater than the number of equivalent dipoles, then adjacent clusters are merged in order of radiation contribution from low to high until the number of clusters is reduced to the number of equivalent dipoles.

[0040] If the initial number of clusters is less than the number of equivalent dipoles, the corresponding clusters are split in descending order of radiation contribution until the number of clusters increases to the number of equivalent dipoles.

[0041] If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters.

[0042] In one possible implementation, the device further includes:

[0043] The acquisition module is used to acquire phase-free magnetic field amplitude data within a preset fixed distance range between the scanning plane and the surface of the device under test as the near-field scanning data.

[0044] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0045] The memory stores computer-executed instructions;

[0046] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0047] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0048] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0049] The near-field radiation source spatial pre-calibration method, apparatus, electronic device, and storage medium provided in this application's embodiments screen strong radiation grid points in near-field scanning data whose electromagnetic field intensity is higher than the background noise threshold to form a strong radiation contribution point set. Based on the radiation energy distribution characteristics of the strong radiation contribution point set and a preset number of equivalent dipoles, the number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles. K-means clustering is performed with the adjusted number of clusters to automatically divide the strong radiation region into spatial clusters with the same number of equivalent dipoles, and the geometric boundary of each spatial cluster is output. The above method accurately solves the core technical problems of traditional near-field radiation source reconstruction, such as the lack of effective spatial pre-calibration, low efficiency of blind full-plane search, poor reconstruction accuracy, difficulty in adapting to low-cost phase-free testing scenarios, and inability to fit the radiation distribution characteristics of different devices under test. It achieves efficient, accurate, and highly adaptable near-field radiation source spatial pre-calibration results. Attached Figure Description

[0050] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0051] Figure 1 A flowchart illustrating the spatial precalibration method for near-field radiation sources provided in this application. Figure 1 ;

[0052] Figure 2 A flowchart illustrating the spatial precalibration method for near-field radiation sources provided in this application. Figure 2 ;

[0053] Figure 3 A flowchart illustrating the spatial precalibration method for near-field radiation sources provided in this application. Figure 3 ;

[0054] Figure 4 A schematic diagram of the near-field radiation source spatial pre-calibration device provided in this application;

[0055] Figure 5 A schematic diagram of the structure of the electronic device provided in this application.

[0056] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0057] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0058] With the rapid development of electronic devices towards miniaturization, high frequency, and high integration, electromagnetic interference (EMI) problems are becoming increasingly prominent, directly affecting the normal operation of equipment and electromagnetic compatibility (EMC) compliance. Near-field radiation source reconstruction and far-field prediction technologies have become core methods for EMC design and optimization of electronic devices due to their elimination of the need for large microwave anechoic chambers and low testing costs. Existing solutions typically select magnetic or electric dipoles as equivalent sources, using inversion algorithms such as least squares methods and genetic algorithms to solve for the equivalent source parameters, thereby deriving far-field radiation characteristics. This approach achieves equivalent source reconstruction based on near-field data. However, the initial position of the equivalent dipole in these methods is often randomly set within the full scanning plane, relying solely on the algorithm's global search capability to find the optimal solution. The initial position of the equivalent dipole lacks spatial constraints, requiring numerous ineffective iterations in low-radiation or even non-radiation regions, resulting in low reconstruction efficiency, significant waste of computational resources, and impact on the accuracy of subsequent far-field predictions.

[0059] To address the aforementioned issues, this application provides a method, apparatus, electronic device, and storage medium for spatial pre-calibration of near-field radiation sources. This provides precise spatial constraints for equivalent source reconstruction, improving both reconstruction efficiency and accuracy while adapting to phase-free data scenarios. Specifically, existing solutions typically select magnetic or electric dipoles as equivalent sources, using inversion algorithms such as least squares or genetic algorithms to solve for equivalent source parameters and derive far-field radiation characteristics. This approach reconstructs equivalent sources based on near-field data. However, the initial positions of the equivalent dipoles in these methods are often randomly set within the full-scan plane, relying solely on the algorithm's global search capability to find the optimal solution. The initial positions of the equivalent dipoles lack spatial constraints, requiring numerous ineffective iterations in low-radiation or even non-radiation regions, leading to low reconstruction efficiency, significant waste of computational resources, and negatively impacting the accuracy of subsequent far-field predictions. Considering the above problems, the inventors investigated whether it is possible to first screen strong radiation grid points, then dynamically determine the number of clusters based on the radiation energy distribution characteristics, divide multiple high radiation contribution spatial clusters through K-means clustering, and output the cluster center and geometric boundary, thereby reducing invalid iterations, improving reconstruction efficiency and model matching degree, without the need for phase data, and dynamically adapting to the radiation distribution differences of different devices under test.

[0060] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0061] Figure 1 A flowchart illustrating the spatial precalibration method for near-field radiation sources provided in this application. Figure 1 ,like Figure 1 As shown, the method includes:

[0062] S101: Select strong radiation grid points in the near-field scanning data whose electromagnetic field intensity is higher than the background noise threshold to form a set of strong radiation contribution points.

[0063] To remove background noise grid points from the near-field scanning data and retain only valid data that reflects the true radiation source of the device under test (DUT), thereby reducing subsequent clustering calculations, avoiding noise interference with clustering results, and ensuring that cluster division focuses on the true radiation region, the near-field scanning data is filtered.

[0064] Specifically, the input is phase-free magnetic field amplitude data collected within a preset fixed distance range between the scanning plane and the DUT surface. The data is in units of "grid points" (each grid point contains spatial coordinates (x, y) and the corresponding magnetic field amplitude).

[0065] The background noise threshold is determined by statistically analyzing the magnetic field amplitude data of the entire scan area (e.g., 1.5-3 times the average amplitude of the entire area), without relying on phase information, making it suitable for low-cost scenarios.

[0066] Traverse all scan grid points, mark grid points with magnetic field amplitude ≥ background noise threshold as "strong radiation grid points", and classify grid points with amplitude < threshold as background noise and remove them. Retain the (x,y) coordinates and amplitude information of strong radiation grid points to form an effective dataset for clustering, namely the radiation contribution point set.

[0067] S102: Based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles, dynamically adjust the number of clusters to be the same as the number of equivalent dipoles.

[0068] K-means clustering with a fixed number of clusters cannot adapt to the differences in radiation distribution among different devices under test, thus requiring dynamic adjustment. If the initial number of clusters is too small, radiation sources that should be independent will be merged, causing a single dipole to represent multiple physical sources during subsequent equivalent source reconstruction, resulting in decreased reconstruction accuracy. If the initial number of clusters is too large, the radiation region will be over-segmented, preventing the number of equivalent sources from corresponding one-to-one with the spatial clusters. The final number of clusters must equal the number of equivalent dipoles. The purpose is to establish a one-to-one mapping relationship between each spatial cluster and an equivalent dipole, ensuring that each dipole is optimized only within its corresponding spatial cluster. This achieves effective dimensionality reduction of the location search space and avoids situations where multiple dipoles are trapped in the same radiation region or distributed in a non-radiative region.

[0069] Based on the actual radiation distribution characteristics of the device under test, a reasonable number of clusters is automatically determined to be equal to the preset number of equivalent dipoles, thus establishing a mapping relationship between the equivalent dipoles and the radiation space region.

[0070] Specifically, the characteristics of radiation energy distribution are analyzed to determine the initial number of clusters. K-means clustering is performed on the set of strong radiation contribution points to calculate the radiation contribution of each cluster. Based on all radiation contribution values, the initial number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles.

[0071] Among them, the radiation energy distribution characteristics include at least one of the radiation energy entropy value of the strong radiation contribution point set and the spatial density distribution, wherein the radiation energy entropy value is used to characterize the degree of concentration of radiation energy, and the spatial density distribution is used to characterize the aggregation pattern of the strong radiation region in the scanning plane.

[0072] It should be noted that the equivalent dipole number refers to the number of equivalent magnetic or electric dipoles used when modeling the electromagnetic radiation characteristics of the device under test (DUT) using the equivalent source method. In near-field to far-field transformation techniques in the field of electromagnetic compatibility (EMC), engineers typically use several discrete dipole sources to equivalently characterize the complex electromagnetic radiation behavior of the DUT. Each dipole has specific position coordinates and a dipole moment (including amplitude and phase information). The near-field radiation generated by these dipoles can fit the measured near-field scan data and thus be used to predict far-field radiation characteristics.

[0073] The number of equivalent dipoles reflects the precision with which the complexity of the radiation sources of the device under test (DUT) is characterized. Physically, the electromagnetic radiation from the DUT is generated by various internal noise sources, such as clock signals from digital chips, switching noise from power modules, differential signals from high-speed interfaces, and high-frequency currents on PCB traces. These physical radiation sources are spatially distributed across different locations within the device. The number of equivalent dipoles is a discretized approximation of this complex radiation source distribution; each equivalent dipole represents the combined effect of one or more physical radiation sources.

[0074] When the number of equivalent dipoles is set too small, each dipole needs to cover a large spatial region and undertake the task of representing multiple physical radiation sources, resulting in an oversimplified model that cannot accurately reproduce the detailed features of the near-field distribution, thus affecting the accuracy of far-field prediction. When the number of equivalent dipoles is set too large, although the model's fitting accuracy will improve, too many unknown parameters will be introduced, making the reconstruction problem more ill-conditioned. The search space of the optimization algorithm will expand sharply, the convergence difficulty will increase, and overfitting will easily occur, meaning that the model can perfectly fit the near-field data but has a large deviation in far-field prediction.

[0075] Therefore, the selection of the number of equivalent dipoles needs to strike a balance between modeling accuracy and computational complexity. Typically, engineers will pre-determine a reasonable number of equivalent dipoles based on factors such as the size of the device under test, the distribution density of the radiation source, the wavelength of the target frequency, and the expected reconstruction accuracy. For example, for a small, simple circuit board, 20 to 50 equivalent dipoles may be sufficient; while for a large, complex system-level circuit board, 100 to 200 or even more equivalent dipoles may be required.

[0076] In this embodiment, the number of equivalent dipoles plays a crucial role. It is not only the final target cluster number in the clustering process, but also a key parameter for establishing the mapping relationship between equivalent sources and radiation spatial regions.

[0077] Specifically, the preset number of equivalent dipoles, n, directly determines the number of clusters in the subsequent K-means clustering. After dynamic adjustment in step S102, the final number of clusters will be equal to n. This means that after clustering, exactly n spatial clusters will be obtained, each corresponding to a sub-region with concentrated radiant energy. Subsequently, these n spatial clusters will establish a one-to-one mapping relationship with the n equivalent dipoles in the subsequent reconstruction process, that is, the search range for the position of the i-th equivalent dipole will be strictly limited to within the geometric boundary of the i-th spatial cluster.

[0078] In traditional methods, the optimal positions of n equivalent dipoles are randomly searched across the entire scanning plane, and each dipole can appear anywhere, resulting in an extremely large search space. This invention, however, structurally binds n spatial clusters to n dipoles, allowing each dipole to optimize only within its own dedicated subspace, achieving a fundamental shift from global search to local search. The benefits of this shift are obvious: the search space changes from a free combination of n dipoles across the entire plane to a parallel search of n dipoles within their respective independent subspaces, significantly reducing the dimensionality of the search space and dramatically improving the convergence speed.

[0079] S103: Perform K-means clustering with the adjusted number of clusters to automatically divide the strong radiation region into spatial clusters with the same number of equivalent dipoles, and output the geometric boundary of each spatial cluster.

[0080] The K-means clustering algorithm is suitable for the technical scenario of this invention because radiative hotspots are typically clustered in space, conforming to the K-means clustering assumption. Furthermore, this algorithm has low complexity and fast convergence speed, making it suitable as a pre-calibration method. The geometric boundary of each spatial cluster defines the legal position range of the corresponding equivalent dipole. Subsequent differential evolution algorithms must check whether the position coordinates are within this boundary during initialization, mutation, and crossover operations. This constraint mechanism reduces the dimensionality of the full-plane search to a local subspace search, significantly shrinking the search space. The necessity of expanding the boundary lies in the fact that radiative energy may diffuse beyond the cluster boundary, and strict boundary constraints may exclude the true optimal solution. An appropriate expansion coefficient maintains the constraint effect while providing sufficient search freedom.

[0081] The core objective of this step is to perform K-means clustering with the adjusted number of clusters, accurately delineate the radiation region, and output the geometric boundary of each spatial cluster, providing spatial constraints for subsequent equivalent source reconstruction. These geometric boundaries will serve as constraints on the position search range of each equivalent dipole in the subsequent differential evolution algorithm. Specifically, the geometric boundaries are used to constrain the position search range of each equivalent dipole during the subsequent equivalent source reconstruction process from the full scan plane to its corresponding spatial cluster.

[0082] Specifically, the standard K-means clustering algorithm is first applied to the set of strong radiation contribution points, using the target cluster number obtained in step S102 as the cluster number. The algorithm execution process includes four stages: initialization, allocation, update, and iteration. In the initialization stage, the same number of points as the target cluster number are randomly selected from the set of strong radiation contribution points as initial cluster centers. The choice of initial centers affects the convergence speed but the final result is relatively stable. In the allocation stage, the Euclidean distance from each grid point to each cluster center is calculated, and each grid point is assigned to the nearest cluster center. In the update stage, the centroid coordinates of all grid points within each cluster are recalculated as new cluster centers. In the iteration stage, the allocation and update steps are repeated until the cluster centers no longer change or the preset maximum number of iterations is reached, ultimately resulting in several stable spatial clusters.

[0083] After clustering, the geometric boundary information of each spatial cluster is output. The geometric boundary includes the center coordinates of the cluster, which is the average of the coordinates of all grid points within the cluster, representing the geometric center of the cluster. The minimum and maximum values ​​of the cluster in the x-axis direction and the y-axis direction are also output. These four boundary values ​​together define the rectangular range of the cluster within the scan plane.

[0084] As a preferred approach, the geometric boundary can also be extended. The purpose of extension is to avoid overly restrictive boundaries that prevent the optimization algorithm from finding the optimal solution near the boundary. The extension coefficient can be adaptively determined based on the discreteness of the grid points within the cluster; the higher the discreteness, the larger the extension coefficient. The extended boundary expands outward by a certain proportion from the original boundary, forming an extended constraint boundary with a safety margin.

[0085] The output of this step is the geometric boundary information of several spatial clusters, forming a position constraint mapping table. Each spatial cluster corresponds to a number and includes its center coordinates, minimum and maximum values ​​in the x-axis direction, minimum and maximum values ​​in the y-axis direction, and the extended boundary after expansion processing. These geometric boundaries will serve as constraints on the position search range of each equivalent dipole in the subsequent equivalent source reconstruction process, realizing a structured binding between the equivalent dipole position optimization and the radiation physical space.

[0086] Optionally, before step S101, the method further includes:

[0087] Phase-free magnetic field amplitude data within a preset fixed distance range between the scanning plane and the surface of the device under test are collected as near-field scanning data.

[0088] This step is the foundational data acquisition stage of the entire near-field radiation source spatial pre-calibration method. It provides effective, consistent, and accurate near-field raw data that matches the true radiation characteristics of the device under test (DUT) for subsequent steps. The core is to lock the near-field range of electromagnetic radiation and ensure the validity and consistency of the data by "presetting a fixed distance range and acquiring phase-free magnetic field amplitude". At the same time, it fits the application scenario of low-cost testing and is a prerequisite for all subsequent pre-calibration operations.

[0089] Collect effective near-field data that accurately reflects the spatial distribution of radiation sources on the DUT surface, strictly distinguish between near-field and far-field, and exclude invalid data that is distorted after far-field attenuation;

[0090] Ensure that all collected data is based on a unified distance benchmark to avoid distortion of magnetic field amplitude data due to fluctuations in scanning distance, and ensure the accuracy of subsequent grid point selection and clustering.

[0091] It only collects phase-free magnetic field amplitude data, reducing the requirements for testing equipment, and is suitable for low-cost EMC testing scenarios in small and medium-sized enterprises and early-stage R&D, thus getting rid of the rigid dependence of traditional technologies on phase data.

[0092] Data is collected in a gridded manner to form a standardized dataset of coordinates and amplitudes, which is then directly matched with the gridded data processing logic of subsequent steps, achieving seamless integration of technical steps.

[0093] Specifically, the hardware system components for implementing this method are as follows:

[0094] Near-field scanning platform: Includes a three-dimensional precision mechanical motion control arm for carrying a magnetic field probe to perform high-precision displacement above the device under test (DUT).

[0095] Magnetic field probe: A high-frequency, high-spatial-resolution magnetic field sensing element responsible for acquiring phase-free magnetic field strength data within the plane to be measured.

[0096] Spectrum analyzer or receiver: Connected to the probe, used to record the magnetic field amplitude at a specific frequency point.

[0097] Central processing unit: responsible for running the near-field radiation source spatial precalibration method provided in this application.

[0098] Based on the type of DUT, radiation frequency, and industry-standard near-field testing, a pre-defined fixed distance range (such as 5-10mm, 8-12mm, etc.) between the scanning plane and the DUT's radiation surface is set in advance. This distance range represents the typical range of near-field radiation, ensuring that the scanning plane is completely within the DUT's near-field range. The narrow range of fluctuations meets the requirements of actual engineering operations (avoiding the difficulty of high-precision operation at absolutely fixed distances), while ensuring that the acquisition distance deviation of all scanning points is within an acceptable range, and that the data has a unified comparison benchmark.

[0099] The near-field probe is fixed within a preset distance range, and a surface scanning method (such as rectangular line-by-line scanning) is used to scan the radiation surface of the DUT point by point. For each physical point scanned, the amplitude data of the phaseless magnetic field at that point is collected, and the two-dimensional spatial coordinates (x, y) of that point are recorded to form a one-to-one correspondence between "x coordinate - y coordinate - phaseless magnetic field amplitude". The grid density of the scan can be preset according to the radiation accuracy requirements of the DUT to ensure complete coverage of the radiation area of ​​the DUT.

[0100] The collected data from all scanning points are standardized and organized to form a gridded near-field scanning dataset. Each data unit in the dataset is a scanning grid point, containing a unique two-dimensional spatial coordinate (x, y) and the corresponding phaseless magnetic field amplitude value. This dataset is the direct input for the subsequent S101 step and can be directly used for the screening operation of strong radiation grid points.

[0101] The near-field radiation source spatial pre-calibration method provided in this application selects strong radiation grid points in near-field scanning data whose electromagnetic field intensity is higher than the background noise threshold to form a strong radiation contribution point set. Based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles, the number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles. K-means clustering is performed with the adjusted number of clusters to automatically divide the strong radiation region into spatial clusters with the same number of equivalent dipoles, and the geometric boundary of each spatial cluster is output. The above method accurately solves the core technical problems of traditional near-field radiation source reconstruction, such as lack of effective spatial pre-calibration, low efficiency of blind full-plane search, poor reconstruction accuracy, difficulty in adapting to low-cost phase-free testing scenarios, and inability to fit the radiation distribution characteristics of different devices under test. It achieves efficient, accurate, and highly adaptable near-field radiation source spatial pre-calibration.

[0102] Figure 2 A flowchart illustrating the spatial precalibration method for near-field radiation sources provided in this application. Figure 2 ,like Figure 2 As shown, step S102 specifically includes:

[0103] S201: Analyze the characteristics of radiation energy distribution to determine the initial number of clusters.

[0104] Based on the actual and objective distribution characteristics of the radiation energy of the DUT, rather than subjective human preset, an initial cluster number K0 that fits the spatial distribution of the radiation source is given. This avoids blindly setting the initial cluster number, which would lead to significant subsequent adjustments and increased computational costs. At the same time, it ensures that the initial cluster number fits the actual radiation of the DUT from the source, improving the accuracy of subsequent cluster division.

[0105] Specifically, the initial number of clusters is determined based on the radiation energy distribution characteristics and the preset mapping relationship, where the mapping relationship includes the relationship between the radiation energy distribution characteristics and the number of clusters.

[0106] For the strong radiation grid points selected only by S101, the radiation amplitude of each grid point is used as the basis for characterizing the radiation energy. No other parameters need to be introduced, ensuring the simplicity of the analysis and engineering feasibility. Two types of features that are easy to calculate and can intuitively reflect the spatial distribution of radiation energy are selected:

[0107] Spatial density of radiation energy: Calculate the spatial distribution density of strong radiation grid points, that is, the number of strong radiation grid points per unit area. The higher the density, the more concentrated the radiation source is, and the lower the density, the more dispersed the radiation source is.

[0108] Radiative energy entropy: This measures the dispersion of radiation amplitude distribution at strong radiation grid points. A higher entropy value indicates a more dispersed radiation amplitude distribution (i.e., the DUT has multiple dispersed strong radiation regions), while a lower entropy value indicates a more concentrated radiation amplitude distribution (i.e., the DUT has only one or a few concentrated strong radiation regions).

[0109] Determining the initial cluster number K0: Establish the correspondence between the radiation energy distribution characteristics and the initial cluster number, and determine K0 based on objective characteristics:

[0110] If the spatial density is high and the entropy value is low (radiation is concentrated), then a smaller K0 is set to match the distribution characteristics of the concentrated radiation source;

[0111] If the spatial density is low and the entropy value is high (radiation dispersion), a larger K0 is set to match the distribution characteristics of the dispersed radiation source. The initial cluster number K0 that closely matches the actual radiation energy distribution of the DUT is obtained, which serves as the base cluster number for subsequent S202 clustering.

[0112] S202: Perform K-means clustering on the set of strong radiation contribution points and calculate the radiation contribution of each cluster.

[0113] Based on the initial cluster number K0 determined in S201, preliminary K-means clustering is performed on the strong radiation contribution points concentrated in the strong radiation grid to obtain K0 preliminary spatial clusters. At the same time, the actual radiation value of each preliminary cluster is quantified by radiation contribution, providing a unique and objective quantitative basis for the subsequent dynamic adjustment of the cluster number in S203, avoiding the subjectivity of the cluster number adjustment, and ensuring that the core radiation region of the DUT is retained after the adjustment.

[0114] Specifically, for each cluster, the radiation contribution of the cluster is calculated by the ratio of the sum of the radiation intensities of all grid points within the cluster to the sum of the radiation intensities of all grid points with full radiation.

[0115] Using K0 as the cluster number, K-means clustering is performed on the spatial coordinates (x, y) and radiation amplitude of the strong radiation grid points in two dimensions, and iterated until the cluster centers are stable (such as when the change of the cluster center coordinates between two adjacent iterations meets the preset stability requirements, or when the preset number of iterations is reached). The two-dimensional clustering ensures that the initial clusters are both spatially adjacent and have similar radiation amplitudes, which is consistent with the physical distribution characteristics of the actual radiation sources of the DUT.

[0116] Radiation contribution calculation: For each preliminary spatial cluster obtained by clustering, its radiation contribution is calculated separately. The calculation method is the sum of the radiation intensities of all strong radiation grid points in a single cluster / the sum of the radiation intensities of all strong radiation grid points. The result is presented in the form of a percentage (e.g., the contribution of a certain cluster is 25%). This calculation method is simple, can be directly implemented in engineering, and can accurately quantify the actual value of each cluster to the overall radiation of the DUT.

[0117] K0 preliminary spatial clusters and the radiative contribution percentage of each cluster are obtained, forming a one-to-one correspondence dataset between preliminary clusters and radiative contributions.

[0118] S203: Dynamically adjust the initial number of clusters to be the same as the number of equivalent dipoles based on all radiation contributions.

[0119] The initial cluster number K0 determined in S201 is dynamically adjusted to the target cluster number K_target (i.e., K_target=n) that strictly matches the preset number of equivalent dipoles, based solely on the radiation contribution. This achieves a one-to-one correspondence between the number of clusters and the number of equivalent dipoles, ensuring that each spatial cluster subsequently divided can accurately correspond to the optimization range of an equivalent dipole. This provides a clear spatial constraint quantity basis for subsequent equivalent source reconstruction, while ensuring that the adjusted K_target not only fits the actual radiation distribution of the DUT, but also seamlessly connects with the preset parameters of the equivalent dipoles in the engineering end.

[0120] Specifically, if the initial number of clusters is greater than the number of equivalent dipoles, adjacent clusters are merged in order of increasing radiation contribution until the number of clusters is reduced to the number of equivalent dipoles.

[0121] If the initial number of clusters is less than the number of equivalent dipoles, the corresponding clusters are split in descending order of radiation contribution until the number of clusters increases to the number of equivalent dipoles.

[0122] If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters.

[0123] The near-field radiation source spatial pre-calibration method provided in this application analyzes the radiation energy distribution characteristics, determines the initial number of clusters, performs K-means clustering on the set of strong radiation contribution points, calculates the radiation contribution of each cluster, and dynamically adjusts the initial number of clusters to be the same as the number of equivalent dipoles based on all radiation contribution values. This method accurately solves the core technical problems in traditional near-field radiation source pre-calibration, such as the subjective pre-setting of the number of clusters, which fails to match the actual radiation energy distribution characteristics of different devices under test, and the low matching degree between the number of clusters and the number of equivalent dipoles, which easily leads to distortion of the radiation region division and thus affects the efficiency and accuracy of subsequent equivalent source reconstruction. This method achieves adaptive, quantitative, and engineered determination of the number of clusters, laying a precise quantitative foundation for the subsequent accurate division of spatial clusters with high radiation contributions.

[0124] Figure 3 A flowchart illustrating the spatial precalibration method for near-field radiation sources provided in this application. Figure 3 ,like Figure 3 As shown, step S203 specifically includes:

[0125] S301: If the initial number of clusters is greater than the number of equivalent dipoles, then adjacent clusters are merged in order of increasing radiation contribution until the number of clusters is reduced to the number of equivalent dipoles.

[0126] When the initial number of clusters exceeds the number of equivalent dipoles, the number of clusters is reduced to the number of equivalent dipoles by directional merging of low-value, spatially adjacent initial clusters. At the same time, it is ensured that the core clusters with high radiation contribution are retained after merging, and the spatial distribution of the clusters still conforms to the physical distribution of the actual radiation source of the device under test (DUT). This avoids the problem of core radiation clusters being split and spatial cluster distribution being fragmented due to random merging.

[0127] Specifically, first confirm the numerical relationship between the initial number of clusters K0 determined in S201 and the preset number of equivalent dipoles n. This step is triggered when K0 > n.

[0128] Retrieve all preliminary spatial clusters output by S202, and the radiation contribution and spatial location information (xmin / xmax, ymin / ymax) of each cluster.

[0129] Sort all preliminary clusters from low to high according to their radiation contribution, and mark the clusters with the lowest contribution as priority for merging;

[0130] Starting with the cluster with the lowest contribution, only the initial clusters that are spatially adjacent are merged (adjacency is determined by whether the geometric boundaries of the clusters overlap or are continuously connected without spatial gaps). After merging, a new spatial cluster is generated. The geometric boundary of the new cluster is the boundary extreme value of each adjacent cluster before merging, and the radiation contribution of the new cluster is the sum of the radiation contributions of each adjacent cluster before merging.

[0131] Each time a merge is completed, the number of clusters decreases by 1, and the current number of clusters is recounted. If the number of clusters is still greater than n, the contribution ranking-directed merge operation is repeated until the number of clusters equals n.

[0132] The number of clusters at this point is the target cluster number K_target, and at the same time, the spatial location-radiative contribution data of K_target radiative clusters are obtained.

[0133] S302: If the initial number of clusters is less than the number of equivalent dipoles, the corresponding clusters are split in descending order of radiation contribution until the number of clusters increases to the number of equivalent dipoles.

[0134] When the initial number of clusters is less than the number of equivalent dipoles, the number of clusters is increased to n by targeted splitting of high-value core initial clusters. The core logic of the splitting is that high-contribution clusters are likely to be the core radiation regions of the DUT, which contain multiple sub-radiation sources. After splitting, the cluster division can better fit the spatial distribution of actual radiation sources, while avoiding meaningless splitting of low-contribution clusters, ensuring that all newly added clusters are regions with actual radiation value.

[0135] Confirm the numerical relationship between K0 and n; trigger this step when K0 < n.

[0136] Retrieve all preliminary spatial clusters output by S202, as well as the radiation contribution of each cluster and the coordinates and amplitude information of all strong radiation grid points within the cluster;

[0137] Sort all preliminary clusters in descending order of radiation contribution, and mark the cluster with the highest contribution as the priority for splitting;

[0138] The initial cluster with the highest contribution is split into two secondary K-means clusters. The split is based on the spatial coordinates and amplitude distribution of the strong radiation grid points within the cluster, ensuring that the two secondary clusters are independent sub-regions with high intra-cluster similarity and low inter-cluster similarity. After the split, two new spatial clusters are generated, and their geometric boundaries and radiation contributions are calculated (the radiation contribution of a secondary cluster is equal to the sum of the radiation intensities of the grid points within the cluster / the sum of the radiation intensities of all strong radiation grid points).

[0139] Each time a split is completed, the cluster count is increased by 1, and the current cluster count is recalculated. If the cluster count is still less than n, the contribution ranking-targeted splitting operation is repeated until the cluster count equals n.

[0140] The number of clusters at this point is the target cluster number K_target, and at the same time, the spatial location-radiative contribution data of K_target radiative clusters are obtained.

[0141] S303: If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters.

[0142] When the initial number of clusters naturally matches the number of equivalent dipoles, the process is simplified by directly using the existing data, avoiding meaningless merging / splitting operations, reducing computational load, and improving the efficiency of the entire pre-calibration process. At the same time, the preliminary cluster distribution and radiation contribution data output by S202 are retained, directly providing a basis for the subsequent final clustering.

[0143] Confirm the numerical relationship between K0 and n; trigger this step when K0 = n.

[0144] Without any merging or splitting operations, the initial number of clusters K0 is directly determined as the target number of clusters K_target;

[0145] The spatial location-radiative contribution data of the K_target preliminary spatial clusters output by S202 are retained and used directly as the basis data for the subsequent final K-means clustering.

[0146] The near-field radiation source spatial precalibration method provided in this application embodiment, if the initial number of clusters is greater than the number of equivalent dipoles, then adjacent clusters are merged in order of radiation contribution from low to high until the number of clusters is reduced to the number of equivalent dipoles. If the initial number of clusters is less than the number of equivalent dipoles, then the corresponding clusters are split in order of radiation contribution from high to low until the number of clusters increases to the number of equivalent dipoles. If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters. The above method solves the core problems of traditional pre-calibration, such as subjective manual pre-setting of the number of clusters, mismatch between the number of clusters and the number of equivalent dipoles, and lack of quantitative basis for adjusting the number of clusters. Through standardized and quantitative merging / splitting operations, the cluster number determination not only conforms to the actual radiation energy distribution characteristics of the DUT, but also accurately matches the preset parameters of the equivalent dipoles on the engineering side. This avoids problems such as incomplete radiation area coverage, spatial fragmentation, and excessive redundant clusters caused by unreasonable cluster numbers. At the same time, it ensures that the adjusted radiation clusters are all in the core radiation area of ​​the DUT, providing key quantitative support for reducing invalid iterations of equivalent source reconstruction and improving reconstruction accuracy.

[0147] Figure 4 This is a schematic diagram of the near-field radiation source space pre-calibration device provided in this application, as shown below. Figure 4 As shown, the near-field radiation source spatial pre-calibration device 400 provided in this embodiment specifically includes:

[0148] The filtering module 401 is used to filter strong radiation grid points in near-field scanning data whose electromagnetic field intensity is higher than the background noise threshold, forming a set of strong radiation contribution points.

[0149] The dynamic adjustment module 402 is used to dynamically adjust the number of clusters to be the same as the number of equivalent dipoles based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles.

[0150] The division output module 403 is used to perform K-means clustering with the adjusted number of clusters, automatically dividing the strong radiation region into spatial clusters with the same number of equivalent dipoles, and outputting the geometric boundary of each spatial cluster. The geometric boundary is used to constrain the position search range of each equivalent dipole in the subsequent equivalent source reconstruction process from the full scan plane to its corresponding spatial cluster.

[0151] In one possible implementation, the dynamic adjustment module 402 includes:

[0152] Analyze the characteristics of radiation energy distribution to determine the initial number of clusters;

[0153] Perform K-means clustering on the set of strong radiation contribution points and calculate the radiation contribution of each cluster;

[0154] The initial number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles based on all radiation contributions.

[0155] In one possible implementation, the radiation energy distribution characteristics include at least one of the radiation energy entropy value of the set of strong radiation contribution points and the spatial density distribution, wherein the radiation energy entropy value is used to characterize the degree of concentration of radiation energy, and the spatial density distribution is used to characterize the aggregation pattern of strong radiation regions in the scanning plane.

[0156] In one possible implementation, the dynamic adjustment module 402 analyzes the radiation energy distribution characteristics of the strong radiation contribution point set and determines the initial number of clusters, specifically including:

[0157] The initial number of clusters is determined based on the radiation energy distribution characteristics and the preset mapping relationship. The mapping relationship includes the correspondence between the radiation energy distribution characteristics and the initial number of clusters.

[0158] In one possible implementation, the dynamic adjustment module 402 performs K-means clustering on the strongly radiating grid points to calculate the radiative contribution of each cluster, specifically including:

[0159] For each cluster, the radiative contribution of the cluster is calculated by the ratio of the sum of the radiative intensities of all grid points within the cluster to the sum of the radiative intensities of all grid points with full radiative intensity.

[0160] In one possible implementation, the dynamic adjustment module 402 dynamically adjusts the initial number of clusters to be the same as the number of equivalent dipoles based on all radiation contributions, including:

[0161] If the initial number of clusters is greater than the number of equivalent dipoles, then adjacent clusters are merged in order of radiation contribution from low to high until the number of clusters is reduced to the number of equivalent dipoles.

[0162] If the initial number of clusters is less than the number of equivalent dipoles, the corresponding clusters are split in descending order of radiation contribution until the number of clusters increases to the number of equivalent dipoles.

[0163] If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters.

[0164] In one possible implementation, the near-field radiation source spatial precalibration device 400 further includes:

[0165] The acquisition module 404 is used to acquire phase-free magnetic field amplitude data within a preset fixed distance range between the scanning plane and the surface of the device under test as near-field scanning data.

[0166] The near-field radiation source spatial precalibration device provided in this embodiment can perform the near-field radiation source spatial precalibration method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0167] Figure 5 A schematic diagram of the structure of the electronic device provided in this application. Figure 5 As shown, the electronic device 500 provided in this embodiment includes at least one processor 501 and a memory 502. Optionally, the electronic device 500 further includes a communication component 503. The processor 501, memory 502, and communication component 503 are connected via a bus 504.

[0168] In a specific implementation, at least one processor 501 executes computer execution instructions stored in memory 502, causing at least one processor 501 to perform the above-described method.

[0169] The specific implementation process of processor 501 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0170] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0171] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0172] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0173] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0174] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0175] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0176] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0177] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0178] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0179] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0180] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0181] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0182] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A spatial pre-calibration method for near-field radiation sources, characterized in that, include: Strong radiation grid points with electromagnetic field strength higher than the background noise threshold in near-field scanning data are selected to form a set of strong radiation contribution points. Based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles, the number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles. K-means clustering is performed with the adjusted number of clusters to automatically divide the strong radiation region into spatial clusters with the same number of equivalent dipoles, and the geometric boundary of each spatial cluster is output. The geometric boundary is used to constrain the position search range of each equivalent dipole in the subsequent equivalent source reconstruction process from the full scan plane to its corresponding spatial cluster.

2. The method according to claim 1, characterized in that, The step of dynamically adjusting the number of clusters to be the same as the number of equivalent dipoles based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles includes: Analyze the radiation energy distribution characteristics to determine the initial number of clusters; Perform K-means clustering on the set of strong radiation contribution points and calculate the radiation contribution of each cluster; The initial number of clusters is dynamically adjusted to be the same as the number of equivalent dipoles based on all radiation contributions.

3. The method according to claim 1, characterized in that, The radiation energy distribution characteristics include at least one of the radiation energy entropy value and spatial density distribution of the strong radiation contribution point set, wherein the radiation energy entropy value is used to characterize the degree of concentration of radiation energy, and the spatial density distribution is used to characterize the aggregation pattern of the strong radiation region in the scanning plane.

4. The method according to claim 2, characterized in that, The analysis of the radiation energy distribution characteristics of the strong radiation contribution point set and the determination of the initial number of clusters include: The initial number of clusters is determined based on the radiation energy distribution characteristics and a preset mapping relationship, wherein the mapping relationship includes the correspondence between the radiation energy distribution characteristics and the initial number of clusters.

5. The method according to claim 2, characterized in that, The process of performing K-means clustering on the set of strong radiation contribution points to calculate the radiation contribution of each cluster includes: For each cluster, the radiation contribution of the cluster is calculated by the ratio of the sum of the radiation intensities of all grid points within the cluster to the sum of the radiation intensities of all grid points with full intensity radiation.

6. The method according to claim 2, characterized in that, The step of dynamically adjusting the initial cluster number to be the same as the equivalent dipole number based on all radiation contributions includes: If the initial number of clusters is greater than the number of equivalent dipoles, then adjacent clusters are merged in order of radiation contribution from low to high until the number of clusters is reduced to the number of equivalent dipoles. If the initial number of clusters is less than the number of equivalent dipoles, the corresponding clusters are split in descending order of radiation contribution until the number of clusters increases to the number of equivalent dipoles. If the initial number of clusters is equal to the number of equivalent dipoles, then the initial number of clusters is determined as the final number of clusters.

7. The method according to claim 1, characterized in that, The method further includes: Phase-free magnetic field amplitude data within a preset fixed distance range between the scanning plane and the surface of the device under test are collected as the near-field scanning data.

8. A spatial pre-calibration device for a near-field radiation source, characterized in that, include: The filtering module is used to filter out strong radiation grid points in near-field scanning data whose electromagnetic field intensity is higher than the background noise threshold, forming a set of strong radiation contribution points. The dynamic adjustment module is used to dynamically adjust the number of clusters to be the same as the number of equivalent dipoles based on the radiation energy distribution characteristics of the strong radiation contribution point set and the preset number of equivalent dipoles. The division output module is used to perform K-means clustering with the adjusted number of clusters, automatically dividing the strong radiation region into spatial clusters with the same number of equivalent dipoles, and outputting the geometric boundary of each spatial cluster. The geometric boundary is used to constrain the position search range of each equivalent dipole in the subsequent equivalent source reconstruction process from the full scan plane to its corresponding spatial cluster.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images