Adaptive Measurement and Magnetic Field Visualization Method for Spatial Magnetic Field Radiation Distribution of High-Power Semiconductor Modules

Abstract

The present invention discloses an adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules. Through the combination of experimental measurement and fitting techniques, precise measurement and visual display of the spatial magnetic field radiation around the power module are achieved. Specifically: an adaptive magnetic field intensity detection technique is adopted. By analyzing the magnetic field intensity differences at different measurement points and dynamically comparing them with a preset threshold. When the magnetic field intensity difference exceeds the threshold, the number and density of measurement points are adaptively increased to ensure comprehensive capture of the details of magnetic field changes. Further, in combination with Newton interpolation method or Lagrange interpolation method, parameter value determination such as fitting curves and visualization plotting processing are completed. The present invention provides an efficient and accurate solution for the electromagnetic compatibility design and optimization of high-power modules, guides the optimization of anti-magnetic interference design, and improves the system stability and robustness. It is of great significance to fields such as new energy vehicles, electromagnetic launch, and ejection.

Description

Technical Field

[0001] The present invention belongs to the field of electronic technology, and relates to an adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules. Background Art

[0002] The application demand for power conversion devices in fields such as photovoltaic, wind energy, new energy vehicles, and industrial frequency conversion has increased sharply, and its characteristics are high component density and compact structure. In high-power converters, the large current and high-frequency switching of power devices cause stray magnetic field radiation, which affects circuit performance and the operation of adjacent circuits. The high current change rate, voltage change rate, and peak current of power devices constitute important EMI (Electromagnetic Interference) sources. Existing research focuses on the establishment of equivalent models for magnetic field radiation of power devices, experimental verification, and circuit analysis methods, such as obtaining EMI spectra, common-mode current noise models, EMI estimation models, and reduced-order models through voltage and current changes. However, current research mainly focuses on the near-field single-point magnetic field radiation of high-power semiconductor devices, and there is limited attention to the global magnetic field radiation and rapid and accurate visualization mapping of power modules. This study aims to use an integrated experimental measurement method and mathematical fitting technology to achieve accurate and convenient visualization characterization of the spatial magnetic field radiation of high-power semiconductor device modules.

[0003] Regarding the physical process of magnetic field radiation detection, based on the conducted EMI experimental platform of a buck circuit IGBT module (H. Huang, J. Wu, W. Xu, and T. Lu, The Influence of Driving Parameters on Conducted EMI for an IGBT Module, IEEE Transactions on Electromagnetic Compatibility, vol. 62, no. 5, pp. 2285-2293, Oct. 2020, doi: 10.1109 / TEMC.2020.2971720.), by setting different driving parameters in the external circuit, the switching voltage and current of the IGBT were obtained, and the EMI noise at different driving parameters in the frequency band of 0.15 - 30 MHz was measured. However, the research on magnetic field radiation in IGBT mainly focuses on the near-field single-point magnetic field radiation of IGBT, which is achieved by measuring the radiation noise using an electromagnetic signal receiving device placed close to the IGBT circuit. The disadvantage is that the measured magnetic field distribution is not comprehensive.

[0004] Regarding the research on the magnetic field radiation distribution of devices, based on the analysis of the influence of space magnetic field coupling on the transient process of power MOSFET (Y. Hao, M. Ma and D. Xu, Analysis of space magnetic field coupling effect on transient process of power MOSFET, in proc. 2023 26th International Conference on Electrical Machines and Systems (ICEMS), Zhuhai, China, 2023, pp. 4533-4536.), the time-domain electromagnetic field simulation was completed in combination with the double-pulse test results, and the magnetic field simulation results at specific moments were given. The disadvantage is that the simulation process is extremely time-consuming, and the simulation results only show part of the magnetic field distribution, without considering the magnetic field distribution in other parts around the device.

[0005] Regarding the research on magnetic field visualization mapping, a method for predicting magnetic field radiation based on near-field scanning data (W. Abdelli, A. Frikha, X. Mininger, L. Pichon, and H. Trabelsi, Prediction of Radiation From Shielding Enclosures Using Equivalent 3-D High-Frequency Models, IEEE Transactions on Magnetics, vol. 51, no. 3, pp. 1-4, March 2015, Art no. 7001504, doi:10.1109 / TMAG.2014.2362575.) was used to make destructive openings in the magnetic field shielding cover, and single magnetic field radiation measurements were carried out at the fixed opening positions. Three opening positions were selected for magnetic field measurement in the experiment, and the magnetic field distribution map was drawn. The disadvantage is that the experimental object was locally damaged, and the measurement positions were fixed, ignoring the possible areas with dense magnetic field changes in the device to be measured, resulting in inaccurate final measurement results. Summary of the Invention

[0006] Aiming at the defects in the prior art, the purpose of the present invention is to provide an adaptive measurement and magnetic field visualization method for the space magnetic field radiation distribution of high-power semiconductor modules; this method can comprehensively capture the details of magnetic field changes and can achieve accurate and adaptive measurement of the space magnetic field radiation around complex multi-chip semiconductor high-power modules.

[0007] The technical solution adopted by the present invention is as follows:

[0008] An adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules, characterized in that through the combination of experimental measurement methods and fitting techniques, the adaptive measurement and visualization graph display of the spatial magnetic field radiation around high-power semiconductor modules are realized, including: adopting an adaptive magnetic field intensity detection technique, by analyzing the magnetic field intensity difference at different measurement points, dynamically comparing it with a preset threshold, when the magnetic field intensity difference exceeds the threshold, automatically determining the complexity of the magnetic field distribution in this area, and then adaptively increasing the number and density of measurement points, and then combining the Newton interpolation method and the Lagrange interpolation method to complete the extraction of parameters such as fitting curves and visualization drawing processing.

[0009] In the above technical solution, further, it specifically includes the following steps:

[0010] The first step: Build a magnetic field measurement platform for the high-power semiconductor module to be measured;

[0011] The second step: Based on the platform in the first step, use a magnetic field test sensor to perform equidistant measurements along two perpendicular directions of the plane at different height planes from the high-power semiconductor module;

[0012] The third step: Change the measurement frequency band of the magnetic field, reposition the magnetic field test sensor to different positions and repeat the measurement according to the method in the second step; Record the magnetic field intensity value of each point in each measurement.

[0013] The fourth step: Adopt an adaptive magnetic field intensity detection method, according to the magnetic field intensity results of the equidistant measurement points in the second and third steps, combined with the set threshold, adaptively increase the magnetic field measurement points;

[0014] The fifth step: Adopt the Newton interpolation fitting or Lagrange interpolation fitting method to obtain the magnetic field distribution fitting curve between the measurement points;

[0015] The sixth step: Extract the characteristic parameters of the fitting curve, and draw the obtained global magnetic field radiation value to obtain the visualization drawing result.

[0016] Further, through the second and third steps, the high-power semiconductor module is measured at different heights, different frequency bands, and different directions. Combined with the adaptive magnetic field intensity detection, and different fitting strategies are adopted for different magnetic field measurement points, so as to obtain the global magnetic field radiation of the high-power semiconductor module, where the different heights are any heights between 1 cm and 10 cm from the high-power semiconductor module, the different measurement directions are two perpendicular directions in the same height plane, and the different frequency bands are from 1 KHz to 50 MHz.

[0017] Further, in the fourth step, according to the adaptive magnetic field strength detection method, the magnetic field measurement points are adaptively increased according to the determination conditions, specifically including:

[0018] For two adjacent measurement points x along the same direction in the same plane n and x n+1 , let their corresponding magnetic field magnitude values be f(x n ), f(x n+1 ), compare the magnitudes of f(x n ) and f(x n+1 ). When recording the magnetic field distribution, if their magnitudes are both lower than -125 dBm, and , then no additional measurement point needs to be added between the two, that is, the threshold condition is satisfied; otherwise, the threshold condition is not satisfied, and a supplementary measurement point x b1 needs to be added at the midpoint between the corresponding two measurement points, and then the magnetic field strength is measured to obtain the magnetic field strength f(x b1 ) of the supplementary measurement point. Judge the magnetic field strength of the supplementary measurement point. If it does not satisfy , and , then at the midpoint between x n and x b1 and at the midpoint between x n+1 and x b1 , add two more supplementary measurement points x b2 , x b3 again, and so on until the condition is satisfied.

[0019] Further, based on the magnetic field strength values of the measurement points in the fourth step, for two adjacent measurement points that satisfy the threshold condition, Newton interpolation is performed and fitted into the magnetic field strength distribution function B1(x) between the measurement points:

[0020]

[0021] Further, for two adjacent measurement points that do not satisfy the threshold condition, Lagrange interpolation will be used to fit into the magnetic field strength distribution function B2(x) between the measurement points. First, construct the Lagrange basis function l i (x):

[0022]

[0023] Obtain the Lagrange fitting function B2(x):

[0024]

[0025] Further, after the fifth-step fitting process for all measurement points in the same direction, the fitting curve of this measurement direction can be obtained. By performing the above fitting process for the measurement points at different heights, different frequency bands, and different directions, the global magnetic field distribution characteristics in the studied frequency band can be obtained.

[0026] Further, equal-parameter values are taken for the fitting curve, and finally, the obtained global magnetic field radiation values are plotted in two-dimensional visualization. Specifically, if the maximum number of supplementary measurement points added due to not meeting the threshold condition in all Lagrangian interpolations is P, then P points are taken at equal distances on each fitted curve to obtain their corresponding magnetic field radiation values. Finally, two-dimensional visualization plotting is performed through the interpolation fitting method to obtain the global magnetic field distribution map.

[0027] The beneficial effects of the present invention are as follows:

[0028] The present invention proposes an adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules. By building a magnetic field radiation detection experimental platform and adopting an adaptive magnetic field intensity detection method, combined with selective Newton interpolation fitting and Lagrangian interpolation fitting methods, the global magnetic field radiation value distribution is obtained, and visualization plotting is completed. The present invention provides an efficient and accurate solution for the electromagnetic compatibility design and optimization of high-power semiconductor modules, which can improve the scientificity and pertinence of the anti-magnetic interference design of sensitive power modules, as well as the stability and robustness of the operation of the entire system. This is of great significance for promoting the application expansion of power electronics technology in new energy vehicles and electromagnetic emission, electromagnetic catapult and other fields. Description of the Drawings

[0029] Figure 1 is a flowchart of an adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules according to the present invention.

[0030] Figure 2 is a magnetic field distribution map drawn by different methods in the 15 MHz frequency band according to the present invention: (a) The figure is the distribution map measured in direction 1 at a distance of 5 cm from the DUT and drawn by interpolating and fitting the original data, and (b) The figure is the distribution map measured in direction 1 at a distance of 5 cm from the DUT and drawn by the method proposed in the present invention.

[0031] Figure 3 is a magnetic field distribution map drawn by different methods in the 15 MHz frequency band according to the present invention: (a) The figure is the distribution map measured in direction 2 at a distance of 10 cm from the DUT and drawn by interpolating and fitting the original data, and (b) The figure is the distribution map measured in direction 2 at a distance of 10 cm from the DUT and drawn by the method proposed in the present invention.

[0032] Figure 4These are magnetic field distribution diagrams drawn by different methods within the 32 MHz frequency band in the present invention: Diagram (a) is the distribution diagram measured in direction 1 at a distance of 5 cm from the DUT and drawn by interpolating and fitting the original data, and diagram (b) is the distribution diagram measured in direction 1 at a distance of 5 cm from the DUT and drawn by the method proposed in the present invention.

[0033] Figure 5 These are magnetic field distribution diagrams drawn by different methods within the 32 MHz frequency band in the present invention: Diagram (a) is the distribution diagram measured in direction 2 at a distance of 10 cm from the DUT and drawn by interpolating and fitting the original data, and diagram (b) is the distribution diagram measured in direction 2 at a distance of 10 cm from the DUT and drawn by the method proposed in the present invention. Detailed implementation manners

[0034] The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings and specific examples.

[0035] An adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules in the present invention combines experimental measurement and fitting techniques to achieve accurate and adaptive measurement of the spatial magnetic field radiation around complex power modules, and further converts the measurement results into intuitive visualization graphics for display. In the specific implementation process, first, data of some specific measurement points are obtained through experimental means. The adaptive magnetic field intensity detection technology is adopted. By analyzing the magnetic field intensity differences at different measurement points and dynamically comparing them with a preset threshold, when the magnetic field intensity difference exceeds the threshold, the system automatically determines that the magnetic field distribution in this area is relatively complex, and then adaptively increases the number and density of measurement points to ensure comprehensive capture of the details of magnetic field changes. After obtaining rich magnetic field intensity data, the present invention further selectively combines the Newton interpolation method or the Lagrange interpolation method to complete the value-taking of the fitting curve parameters and perform visualization drawing processing. As Figure 1 shown, it specifically may include the following steps:

[0036] The first step: Build an experimental platform composed of a three-phase inverter circuit, a high-power module drive and protection circuit, a system parameter acquisition circuit, and an EMI probe.

[0037] The second step: According to the experimental platform built in the first step, fix the direction of the EMI probe, and use an EMI test receiver to perform equidistant measurements in the horizontal and vertical directions within planes at different heights.

[0038] The third step: Change the measurement frequency band of the magnetic field, then reposition the EMI probe to different positions and repeat the measurement, and record the magnetic field intensity value at each point.

[0039] Step 4: Use the adaptive magnetic field intensity detection method. According to the magnetic field intensity results of the equally spaced measurement points in Step 2 and Step 3, and in combination with the set threshold, adaptively increase the magnetic field measurement points.

[0040] Step 5: Optionally use Newton interpolation fitting or Lagrange interpolation fitting method to obtain the fitting curve of the magnetic field distribution between the measurement points.

[0041] Step 6: Perform equal-parameter value taking on the fitting curve, and perform two-dimensional plotting on the obtained global magnetic field radiation values to obtain the visualization plotting result.

[0042] According to a specific embodiment of the present invention, in this method:

[0043] The magnetic field radiation evaluation of the power module involves the amplitudes at different frequency bands, different heights, and different directions. The measurements are carried out at 5 and 10 cm away from the module respectively. First, use an EMI probe to scan the measurement area above the module at equal distances, and then use the adaptive magnetic field intensity measurement method to appropriately increase the magnetic field detection points. Here, two different measurement directions are adopted: In Direction 1, the probe is positioned parallel to the measurement plane, while in Direction 2, the probe is positioned perpendicular to the measurement plane.

[0044] When performing EMI space measurement, initially use an EMI probe with a fixed Direction 1 for measurement. In the horizontal and vertical directions within each measurement plane, perform equally spaced measurements. The magnetic field measurement frequency band range is from 1 KHz to 50 MHz or even higher, and the resolution bandwidth is set to 1 KHz. Repeat the measurement by repositioning the EMI probe to different positions, and record the magnetic field intensity value at each point. Use the adaptive magnetic field intensity detection method, in combination with Newton interpolation fitting and Lagrange interpolation fitting methods, to obtain the global magnetic field distribution and perform visualization plotting.

[0045] Specifically, first complete the measurement of 13 points in the horizontal direction and 9 points in the vertical direction. Take the first measurement at the bottom leftmost as the coordinate origin (0, 0), and its magnetic field magnitude is f(x1), with the unit of dBm. Take the vertical upward direction as the positive direction of the Y-axis, and the horizontal rightward direction as the positive direction of the X-axis. The magnetic field magnitude value of the next measurement point along the positive direction of the X-axis is f(x2), with the unit of dBm. Finally, fill the obtained magnetic field intensity values of the measurement points into the initial magnetic field radiation measurement matrix R 13x9 :

[0046]

[0047] First, judge the magnitudes of f(x1) and f(x2). When recording the maximum (minimum) magnetic field distribution, if the magnitude is lower than -125 dBm, and , then directly perform Newton interpolation according to the following method and fit it into the magnetic field strength distribution function B(x) between the measurement points:

[0048]

[0049] If we let:

[0050]

[0051] Then we have:

[0052]

[0053] Among them, l1(x) and l2(x) are linear polynomials about x respectively, and they are called interpolation basis functions. They satisfy at the nodes x1 and x2:

[0054]

[0055] Therefore, a key conclusion can be drawn: The first-order interpolation polynomial B(x) that satisfies the interpolation conditions can be constructed by the linear combination of the two interpolation basis functions l1(x) and l2(x), that is, by giving the positions of two different magnetic field measurement points and their magnetic field strength values, the magnetic field strength values between the two points can be fitted.

[0056] During the specific test process, it is measured that f(x1) = -121 dBm, f(x2) = -124 dBm, x1 = 0 (m), x2 = 0.03 (m), which satisfies the above threshold conditions. Then, perform Newton fitting on it. Referring to Formula 2, the Newton interpolation fitting function B1(x) between the two points is obtained:

[0057]

[0058] If the judgment condition is not satisfied, then perform EMI measurement again in the middle of the two measurement points to obtain the magnetic field strength f(x3) of the supplementary measurement point. If it does not satisfy and , then add supplementary measurement points again between the corresponding two points until the conditions are satisfied.

[0059] Specifically, the measurement results f(x5) = -106 dBm, f(x6) = -100 dBm do not satisfy the threshold conditions. The magnetic field strength measurement point f(x b1 ) = -103 dBm is supplemented. At this time, x5 = 0.12 m, x6 = 0.15 m; at x b1 = 0.135 m, the three points at this time satisfy the above conditions.

[0060] Next, Lagrange interpolation fitting is adopted. First, the Lagrange basis function l i (x) is constructed as follows:

[0061]

[0062] l i (x) satisfies:

[0063]

[0064] The basis functions can be obtained as follows:

[0065]

[0066]

[0067]

[0068] The Lagrange fitting function B2(x) is obtained:

[0069]

[0070] Substituting the parameter values, the Lagrange fitting function B2(x) can be obtained:

[0071]

[0072] Since the fitting function satisfies that the value of B(x i ) at each point is the same as the measured value f(x i ) of the magnetic field magnitude, that is:

[0073]

[0074] It satisfies:

[0075]

[0076] This nth-degree polynomial satisfies:

[0077]

[0078] The coefficient matrix of the fitting polynomial is:

[0079]

[0080] The coefficient matrix belongs to the Vandermonde determinant. Since the original measurement points and the supplementary measurement points are different points, the corresponding fitting polynomial of the above coefficient matrix exists and is unique. Thus, the multi-point measurement results can be represented by a unique magnetic field distribution fitting function. Then, P points are equidistantly taken on the curves fitted by Newton interpolation and the curves fitted by Lagrange interpolation, and the corresponding magnetic field intensity values are added to the magnetic field radiation value matrix to obtain the final magnetic field intensity radiation value matrix for plotting.

[0081] Specifically, in all fitting processes, at most 3 additional measurement points are inserted between two points to meet the threshold condition, so the parameter P = 3. Therefore, 3 points need to be equidistantly taken on each fitting curve, and the obtained magnetic field intensity values are added to the magnetic field radiation intensity matrix. Through the two-dimensional interpolation mapping method, a visualization map of the magnetic field distribution is obtained. For different frequencies, different heights, and different test directions, the magnetic field distribution maps obtained by using the method of the present invention and directly by interpolating and fitting the original data are as Figure 2 、 3 、4 and Figure 5 shown; the distribution area of the EMI intensity of the power module can be intuitively observed. And it can be seen that the magnetic field distribution map obtained by the ordinary interpolation method is not as accurate as the distribution map obtained by the method proposed in the present invention. The former misses some regions with higher magnetic field intensity because the magnetic field distribution between the measurement points is not considered. The method of the present invention obtains a smoother and more accurate magnetic field distribution map by adaptively adding magnetic field measurement points in a timely manner.

Claims

1. An adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules, characterized in that It includes the following steps: Step 1: Build a magnetic field measurement platform for the high-power semiconductor module to be measured; Step 2: Based on the platform in Step 1, use a magnetic field test sensor to perform equidistant measurements along two perpendicular directions of the plane at different heights from the high-power semiconductor module; Step 3: Change the measurement frequency band of the magnetic field, reposition the magnetic field test sensor to different positions and repeat the measurement according to the method in Step 2; Record the magnetic field intensity values of each point in each measurement; Step 4: Adopt an adaptive magnetic field intensity detection method. According to the magnetic field intensity results of the equidistant measurement points in Step 2 and Step 3, combined with the set threshold, adaptively increase the magnetic field measurement points; Step 5: Adopt Newton interpolation fitting or Lagrange interpolation fitting method to obtain the fitting curve of the magnetic field distribution between the measurement points; Step 6: Extract the characteristic parameters of the fitting curve, and plot the obtained global magnetic field radiation value to obtain the visualization drawing result; In the fourth step, according to the adaptive magnetic field intensity detection method, adaptively increase the magnetic field measurement points according to the judgment conditions, specifically including: For two adjacent measurement points x in the same plane and in the same direction n and x n+1 , let the corresponding magnetic field magnitude values be f(x n ) and f(x n+1 ). Compare the magnitudes of f(x n ) and f(x n+1 ). When recording the magnetic field distribution, if their magnitudes are both lower than -125 dBm, and , then no additional measurement points need to be added between the two, that is, the threshold condition is satisfied; Otherwise, it does not meet the threshold condition, and a supplementary measurement point x needs to be added in the exact middle between the corresponding two measurement points. b1 , then measure the magnetic field strength again to obtain the magnetic field strength f(x b1 ) of the supplementary measurement point, and judge the magnetic field strength of the supplementary measurement point. If it does not meet , and , then at the exact middle between x n and x b1 , and at the exact middle between x n+1 and x b1 , add two more supplementary measurement points x b2 , x b3 , and so on until the condition is met.

2. The adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of a high-power semiconductor module according to claim 1, characterized in that: Through the second step and the third step, the high-power semiconductor module is measured at different heights, different frequency bands, and different directions. Combining the adaptive magnetic field intensity detection and adopting different fitting strategies for different magnetic field measurement points, the global magnetic field radiation of the high-power semiconductor module is obtained, where the different heights are any heights between 1 cm and 10 cm from the high-power semiconductor module, the different directions are two mutually perpendicular directions in the same-height plane, and the different frequency bands are from 1 KHz to 50 MHz.

3. The adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules according to claim 1, characterized in that: Based on the magnetic field intensity values of the measurement points in the fourth step, for two adjacent measurement points that meet the threshold conditions, perform Newton interpolation and fit them into the magnetic field intensity distribution function B1(x) between the measurement points.

4. The adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules according to claim 1, characterized in that: For two adjacent measurement points that do not meet the threshold conditions, use Lagrange interpolation to fit them into the magnetic field intensity distribution function B2(x) between the measurement points.

5. The adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules according to claim 1, characterized in that: After the fifth-step fitting process for all the measurement points in the same direction, the fitting curve of the measurement direction is obtained. For the measurement points at different heights, different frequency bands, and different directions, perform the above fitting process to obtain the global magnetic field distribution characteristics in the studied frequency band.

6. The adaptive measurement and magnetic field visualization method for the spatial magnetic field radiation distribution of high-power semiconductor modules according to claim 1, characterized in that: Perform equal-parameter value taking on the fitting curve, and finally perform two-dimensional visualization drawing on the obtained global magnetic field radiation value. Specifically, the maximum number of supplementary measurement points added due to not meeting the threshold conditions in all Lagrange interpolations is P. Then, take P points at equal distances on each fitted curve to obtain their corresponding magnetic field radiation values. Finally, perform two-dimensional visualization drawing through the interpolation fitting method to obtain the global magnetic field distribution map.

Images