A multi-objective optimization method for high voltage silicon carbide modules

Abstract

This invention discloses a multi-objective optimization method for high-voltage silicon carbide modules, relating to the fields of power semiconductor packaging and electromagnetic compatibility (EMC). It quantifies the transient electric field distribution vector of the switch and generates a dynamic admittance compensation scalar using a mapping function. Based on impedance offset, it calculates the resonant frequency offset of the parasitic loop and tunes the impedance topology parameters accordingly. This invention establishes the physical logic of field strength stress-driven impedance evolution, enabling local charge migration to be reflected as a controllable evolution of the resonant frequency. While suppressing electric field distortion, it actively drives interference energy to avoid sensitive frequency bands in the spectral space. This scheme unifies insulation reliability improvement and EMC performance control within the same physical response logic framework, effectively resolving the structural contradiction between insulation stress and electromagnetic interference in high-voltage power modules.

Description

Technical Field

[0001] This invention relates to the fields of power semiconductor packaging and electromagnetic compatibility technology, and in particular to a multi-objective optimization method for high-voltage silicon carbide modules. Background Technology

[0002] High-voltage silicon carbide power modules typically withstand high voltage levels and extremely high switching rise rates during operation. Under nanosecond-level switching transient conditions, significant transient electric field distortion is easily generated in the three-phase point region at the edge of the substrate. This distortion not only induces partial discharge and accelerates the aging of the insulating medium, but also excites high-frequency electromagnetic interference noise through parasitic loops inside the high-voltage silicon carbide power module.

[0003] In existing modular structure designs, the geometric layout and material configuration of the insulation structure directly determine the equivalent inductance and capacitance distribution of parasitic circuits, thereby determining the inherent resonant characteristics and electromagnetic interference spectrum distribution of high-voltage silicon carbide power modules. When the local electric field strength is reduced by increasing the insulation spacing or introducing buffer structures, parasitic parameters are often inevitably altered, causing electromagnetic interference energy to shift to higher or more sensitive frequency bands. Conversely, if the parasitic circuit is optimized with the goal of suppressing electromagnetic interference, it may lead to an increase in the local electric field stress concentration.

[0004] Therefore, within the existing design framework of high-voltage silicon carbide modules, insulation reliability and electromagnetic compatibility (EMC) performance exhibit a clear physical coupling and mutual constraint relationship. Existing technologies struggle to coordinate and regulate insulation reliability and EMC performance without introducing new performance risks. Especially under complex dynamic operating conditions and high-frequency switching conditions, the physical risk of insulation failure and excessive electromagnetic noise persists. Summary of the Invention

[0005] This invention provides a multi-objective optimization method for high-voltage silicon carbide modules, aiming to solve the technical problem that insulation stress distribution and electromagnetic interference spectrum characteristics are coupled and difficult to coordinate and control in existing high-voltage silicon carbide power modules.

[0006] In view of the above problems, the present invention provides a multi-objective optimization method for high-voltage silicon carbide modules, comprising the following steps: S1. Electric field feature quantization: Obtain the excitation parameter sequence of the high voltage silicon carbide module during the switching transient process, quantize the electric field distribution vector of the target boundary domain in the time domain, and determine the equivalent geometric capacitance of the target boundary domain. S2. Response Mapping Processing: The electric field distribution vector is input into a mapping function to generate a dynamic admittance compensation scalar; the mapping function has a preset correspondence between the admittance change and the electric field strength based on the response threshold; S3. Frequency tuning calculation: Determine the impedance offset of the target boundary region based on the dynamic admittance compensation scalar and equivalent geometric capacitance, and calculate the resonant frequency offset of the parasitic circuit in the high-voltage silicon carbide module. S4. Topology parameter tuning: Based on the matching value between the resonant frequency offset and the preset peak value of the interference spectrum in the interference spectrum characteristics, the impedance topology parameters of the target boundary region are tuned.

[0007] Furthermore, the mapping function in S2 satisfies the following formula: in, This is the dynamic admittance compensation scalar. This is the upper limit of admittance compensation. The ratio of dimensionless field strength. This is the gain coefficient. As a saturation adjustment factor, The damping coefficient; It is defined as the ratio of the magnitude of the electric field distribution vector to the response threshold.

[0008] Furthermore, the resonant frequency offset in S3 Satisfy the following formula: in, This is the resonant frequency offset. As the reference resonant frequency, The equivalent geometric capacitance of the target boundary region. The equivalent time response factor is used to compensate the dynamic admittance scalar. It is converted into an equivalent dynamic charge storage increment.

[0009] Furthermore, S1 to S4 establish the following correlation: the electric field fluctuations generated by the excitation parameter sequence trigger the impedance shift in the target boundary region, and the impedance shift drives the resonant center of the parasitic loop to move towards the peak of the interference spectrum. Further, the impedance topology parameters include the fractal dimension of the impedance distribution and the cell spacing, wherein the fractal dimension is between 1.5 and 1.6; the fractal dimension is determined by adjusting the number of sub-cells retained and the scaling ratio of the generator operator of the generalized fractal structure.

[0010] Furthermore, the method is applied to a functional area with field-controlled response characteristics, the functional area being located at the interface between the edge of the liner and the insulating medium.

[0011] This invention also provides a multi-objective optimization system for high-voltage silicon carbide modules, comprising: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the above method.

[0012] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, performs the steps of the above-described method.

[0013] The technical solution provided in this application has at least the following technical effects: By modulating the relationship between the electric field intensity distribution characteristics and admittance response within the target boundary region, the local area can form an adaptive and smooth response characteristic that changes with electric field stress under high-voltage switching transient conditions. This suppresses transient field intensity distortion near the three-phase point and ensures that the insulating medium always operates within the safe electric field range under extreme voltage rise rate conditions.

[0014] By establishing a physical relationship between the electric field stress state and the parasitic loop resonance characteristics, the local charge migration behavior inside the functional area is transformed into equivalent capacitance disturbance using the equivalent time response factor, which is then reflected as the controllable evolution of the parasitic loop resonance frequency inside the high-voltage silicon carbide module. This allows for the construction of a stable mapping logic between electric field stress and electromagnetic interference spectrum at the system level.

[0015] Without altering the main physical structure of the module, it achieves active avoidance of electromagnetic interference energy in the frequency spectrum, causing the interference energy generated by high-frequency switching to deviate from the sensitive frequency band, thereby reducing the risk of electromagnetic interference and improving the module's electromagnetic compatibility operating conditions.

[0016] The synergistic achievement of the above-mentioned technical effects unifies the process of improving the insulation reliability of high-voltage silicon carbide power modules and the process of regulating electromagnetic compatibility performance within the same physical response logic framework, and can simultaneously meet the requirements of insulation safety and electromagnetic environment adaptability under complex dynamic operating conditions. Attached Figure Description

[0017] Figure 1 The present invention provides a logic flowchart of a multi-objective optimization method for high-voltage silicon carbide modules. Detailed Implementation

[0018] The above technical solutions will now be described in detail with reference to the accompanying drawings and specific embodiments to provide a better understanding of them. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments used only to explain the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Furthermore, it should be noted that, for ease of description, only the parts related to the present invention are shown in the drawings, not all of them. Example

[0019] Please see Figure 1 The multi-objective optimization method for high-voltage silicon carbide modules is executed by a multi-objective optimization system for high-voltage silicon carbide modules. The multi-objective optimization method for high-voltage silicon carbide modules includes the following steps: S1. Electric field feature quantization: For the multi-objective optimization system of high voltage silicon carbide module, the excitation parameter sequence of high voltage silicon carbide module during the switching transient process is obtained, and the electric field distribution vector of the target boundary domain is quantized in the time domain, and the equivalent geometric capacitance of the target boundary domain is determined.

[0020] S2. Response Mapping Processing: For the multi-objective optimization system of the high-voltage silicon carbide module, the electric field distribution vector is input into the mapping function to generate a dynamic admittance compensation scalar. The mapping function has a pre-defined correspondence between the admittance change and the electric field strength based on the response threshold.

[0021] S3. Frequency tuning calculation: For the multi-objective optimization system of high voltage silicon carbide module, the impedance offset of the target boundary region is determined based on the dynamic admittance compensation scalar and equivalent geometric capacitance, and the resonant frequency offset of the parasitic circuit in high voltage silicon carbide module is calculated.

[0022] S4. Topology parameter tuning: For the multi-objective optimization system of high-voltage silicon carbide module, the impedance topology parameters of the target boundary region are tuned according to the matching value between the resonant frequency offset and the preset interference spectrum peak value in the interference spectrum characteristics.

[0023] By executing steps S1 to S4 above, the multi-objective optimization system for high-voltage silicon carbide modules achieves coordinated control of the insulation strength and electromagnetic compatibility performance of the high-voltage silicon carbide modules.

[0024] The physical application environment, target boundaries, and logical execution process of the multi-objective optimization method for high-voltage silicon carbide modules are revealed through the following specific embodiments.

[0025] A multi-objective optimization method for high-voltage silicon carbide (HCC) modules is applied to power electronic operating environments with high-voltage, high-frequency switching. In the actual operation of the HCC module, the voltage level is set to be greater than or equal to 3.3 kV. At the instant of switching, the rate of change of voltage over time is set to be greater than or equal to 100 V / ns. The charge surge generated by this high-voltage transient condition constitutes the external excitation condition for the multi-objective optimization method for the HCC module. The power semiconductor chips inside the HCC module frequently switch under these conditions, causing the target interface to experience dynamically changing electric field pressure. The multi-objective optimization method for the HCC module obtains the excitation parameter sequence of the HCC module under operating conditions, providing a logical starting point for subsequent impedance adjustment.

[0026] The physical space targeted by the multi-objective optimization method for high-voltage silicon carbide modules is defined as the target boundary region. The target boundary region is located in the edge region of the substrate inside the high-voltage silicon carbide module. The specific geometric range of the target boundary region is formed by the intersection of the edge of the substrate metal layer, the surface of the ceramic substrate, and the insulating medium used for packaging. Because substances with different dielectric constants converge at the sharp corners of the substrate metal layer edge, this region forms a three-phase point. During switching transients, electric field distortion occurs at the three-phase point. The electric field distribution vector exhibits a highly concentrated state in the three-phase point region, manifested as a local electric field modulus higher than the average field strength of the surrounding insulating medium. This physical distribution characteristic makes the three-phase point a core source of insulation failure and electromagnetic interference. The multi-objective optimization method for high-voltage silicon carbide modules reshapes the electric field distribution at the three-phase point by configuring a functional area with field-controlled response characteristics within the target boundary region. [Mapping: Weight 6] The multi-objective optimization method for high-voltage silicon carbide modules is uniformly managed by a multi-objective optimization system for high-voltage silicon carbide modules. The execution of this method begins with the quantization of electric field characteristics, specifically including the real-time acquisition of the switching excitation parameter sequence and the time-domain calculation of the electric field distribution vector in the boundary region.

[0027] The multi-objective optimization system for high-voltage silicon carbide (HCC) modules first initiates a real-time acquisition process for the switching excitation parameter sequence. A high-bandwidth voltage acquisition channel, either external to the HCC module or integrated into the gate drive unit, synchronously captures the gate drive voltage signal and the DC bus voltage signal. To fully capture voltage surges with a slope greater than 100V / ns, the sampling frequency of the high-bandwidth voltage acquisition channel is set to above 1GHz. The acquired analog voltage waveforms are processed by a high-precision analog-to-digital converter and converted into a time-series sequence of excitation parameters. Each data point in the excitation parameter sequence includes a timestamp and the corresponding voltage amplitude. Using a digital filtering algorithm, the multi-objective optimization system for HCC modules preprocesses the excitation parameter sequence to remove random thermal noise, ensuring that the excitation parameter sequence reflects the potential fluctuations of the HCC module during switching transients.

[0028] With the generation of the excitation parameter sequence, the multi-objective optimization system for the high-voltage silicon carbide module then enters the time-domain calculation stage of the electric field distribution vector in the boundary region. The multi-objective optimization system for the high-voltage silicon carbide module internally loads a geometric topological discretization model for the target boundary region. This model divides the physical boundaries of the metal layer edge, the ceramic substrate, and the insulating dielectric into micro-grid units. Using the geometric topological discretization model, combined with the relative position information of each micro-grid unit within the target boundary region and the dielectric constitutive parameters, the multi-objective optimization system for the high-voltage silicon carbide module calculates the equivalent geometric capacitance using the static electric field integration method. The equivalent geometric capacitance characterizes the inherent charge storage capacity of the target boundary region in the non-response state.

[0029] Simultaneously, the multi-objective optimization system for the high-voltage silicon carbide module loads the excitation parameter sequence as boundary conditions into a pre-defined quasi-static electromagnetic field model. The quasi-static electromagnetic field model establishes a three-dimensional spatial field intensity mapping on each micro-grid unit by numerically solving the transient solution of the Laplace equation. Through this three-dimensional spatial field intensity mapping, the multi-objective optimization system for the high-voltage silicon carbide module separates the normal electric field component distributed along the edge of the metal layer and the tangential electric field component distributed along the surface of the ceramic substrate. The magnitude and direction information of the normal and tangential electric field components together form the electric field distribution vector. This vector quantifies the spatial geometric distribution of the electric field intensity within the target boundary region and characterizes the dynamic evolution trajectory of the electric field stress on the nanosecond scale.

[0030] After the electric field distribution vector and equivalent geometric capacitance are constructed, the response mapping process is triggered. The multi-objective optimization system for the high-voltage silicon carbide module provides the electric field distribution vector as input data to the mapping function to generate the dynamic admittance compensation scalar required for subsequent steps. This mapping function pre-defines the correspondence between admittance change and electric field strength. Thus, the multi-objective optimization system for the high-voltage silicon carbide module completes the logical transformation from physical field characteristics to admittance response data and structural impedance parameters.

[0031] Driven by the electric field distribution vector, the multi-objective optimization system for the high-voltage silicon carbide module enters the response mapping processing stage. The first step in response mapping processing is the execution of gating logic based on the response threshold. In this stage, the multi-objective optimization system for the high-voltage silicon carbide module extracts the magnitude of the electric field distribution vector and establishes a division relationship between the magnitude of the electric field distribution vector and the preset response threshold, thereby calculating the dimensionless field strength ratio. The mapping function processes the difference between the dimensionless field strength ratio and the value of 1 through a linear rectification function. When the magnitude of the electric field distribution vector is lower than the response threshold, the dimensionless field strength ratio is less than 1, and the output of the linear rectification function is confirmed to be 0. At this time, the mapping function does not generate an effective admittance increment, ensuring that the functional areas inside the high-voltage silicon carbide module maintain a high-resistance insulation state under low field strength conditions. This gating logic constitutes the silent mechanism of the multi-objective optimization method for the high-voltage silicon carbide module, avoiding ineffective adjustment actions within the safe voltage range.

[0032] When the magnitude of the electric field distribution vector exceeds the response threshold, the multi-objective optimization system for the high-voltage silicon carbide module immediately invokes the mapping function to perform saturation mapping calculation of the dynamic admittance compensation scalar. The mapping function determines the dynamic admittance compensation scalar using the following formula: In this formula, It is a scalar for dynamic admittance compensation. The preset compensation upper limit saturation value, The ratio of dimensionless field strength. This is the gain coefficient. As a saturation adjustment factor, Damping coefficient. Dimensionless electric field ratio. Defined as the ratio of the magnitude of the electric field distribution vector to a preset material activation threshold. The hyperbolic tangent function term simulates the carrier mobility saturation characteristics of nonlinear conductive materials under real physical conditions, while the linear damping term is used to suppress numerical divergence under high field strength excitation. The dynamic admittance compensation scalar calculated by this mapping function for the multi-objective optimization system of the high-voltage silicon carbide module reflects the physical response of the functional region under transient field strength in real time. The generation of the dynamic admittance compensation scalar directly drives the multi-objective optimization system for the high-voltage silicon carbide module into the frequency tuning calculation stage.

[0033] For the multi-objective optimization system of high-voltage silicon carbide modules, the gain coefficient was determined by performing sensitivity analysis and stability tests on the mapping function. Saturation regulation factor and damping coefficient The optimal range was determined. To verify the selection of the mapping function coefficients, response characteristic data under different parameter combinations were recorded for the multi-objective optimization system of the high-voltage silicon carbide module. The specific comparison results are shown in the table below:

[0034] The above comparative data shows that the gain coefficient Choosing a value of 35 ensures that the multi-objective optimization method for high-voltage silicon carbide modules has sufficient dynamic sensitivity while suppressing noise interference.

[0035] Saturation regulation factor Choosing 1.0 ensures the continuity of the dynamic admittance compensation scalar near the material activation threshold, avoiding the deterioration of the electromagnetic compatibility environment caused by impedance abrupt changes.

[0036] Damping coefficient Choosing a value of 0.1 provides a negative feedback constraint for the mapping function under high field strength at the physical logic level, ensuring the numerical convergence of the multi-objective optimization system for high-voltage silicon carbide modules under extreme overload conditions. The coordinated selection of the above coefficients enables the mapping function to accurately and stably transform the abstract electric field distribution vector into physically meaningful admittance compensation data.

[0037] Based on the output of the dynamic admittance compensation scalar, frequency tuning calculations are initiated for a multi-objective optimization system of a high-voltage silicon carbide module. The causal relationship between impedance shift and charge migration is established during the frequency tuning calculation. When the dynamic admittance compensation scalar indicates that the functional region has entered the response state, a tunneling effect occurs within the nonlinear conductive material. This tunneling effect drives micro-conduction behavior of charges within the target boundary region. This micro-conduction behavior directly alters the equivalent complex impedance of the target boundary region. The change in equivalent complex impedance is defined as impedance shift. Specifically, impedance shift manifests as the synchronous evolution of the imaginary and real parts of the equivalent admittance, thereby physically altering the charge storage and energy loss characteristics of the parasitic loops within the high-voltage silicon carbide module.

[0038] After the impedance offset is quantized, a coupled solution for the resonant frequency offset is performed for the multi-objective optimization system of the high-voltage silicon carbide module. The tuning logic for the resonant frequency offset is implemented through the following coupling equations: In this equation, The center offset, The system's inherent resonant frequency, The equivalent geometric capacitance of the fractal microelectrode structure. The equivalent time response factor is used for multi-objective optimization of high-voltage silicon carbide modules. Establishing a dimensional transformation between admittance response and charge storage effect allows for the dynamic admittance compensation scalar. With equivalent geometric capacitance Achieve dimensional uniformity in the denominator.

[0039] Equivalent time response factor used in multi-objective optimization system for high-voltage silicon carbide modules This was determined through a pre-executed offline calibration process. In this process, the experimental platform used electromagnetic field simulation tools to obtain the rate of change of energy storage under different admittance states. The contribution coefficient of admittance change to equivalent capacitance was determined through regression analysis. For the multi-objective optimization system of the high-voltage silicon carbide module, the dynamic admittance compensation scalar was dynamically adjusted. This drives the resonant frequency downward, thereby enabling dynamic tracking of electromagnetic interference spectrum shifts.

[0040] Guided by the quantization results of the resonant frequency offset, topology parameter tuning is initiated for the multi-objective optimization system of the high-voltage silicon carbide module. A parameter search process for fractal dimension and cell spacing is performed for the multi-objective optimization system of the high-voltage silicon carbide module. The fractal dimension is limited to a value range of 1.5 to 1.6 to achieve a physical balance between electric field smoothing effectiveness and parasitic dielectric losses.

[0041] The microelectrode arrays corresponding to the impedance topology parameters adopt either a generalized Sierpinski carpet structure or a generalized Koch snowflake structure. The multi-objective optimization system for the high-voltage silicon carbide module controls the fractal dimension by adjusting the iteration number of the selected fractal structure, the scaling ratio of the generator operator, and the retention density of the sub-cells.

[0042] As a specific implementation path, fractal dimension Determined through the following generator logic: in, This represents the number of sub-units retained in each recursion. This is the scaling factor for the sub-unit.

[0043] When using the generalized Sierpinski carpet structure, the multi-objective optimization system for high-voltage silicon carbide modules adjusts the vacancy distribution of sub-cells in the 3×3 grid to reduce the number of retained sub-cells. With scaling factor The combination satisfies Constraints (e.g., when scaling factor) When the value is one-third, by introducing a random or deterministic spatial truncation algorithm, the equivalent number of retained sub-units of the microelectrode array can be reduced. (The fractal dimension is between 5 and 6, and continuous control of the fractal dimension is achieved by combining non-integer iterative weights). When using a generalized Koch snowflake structure, the multi-objective optimization system for the high-voltage silicon carbide module adjusts the polygonal deflection angle of the generator to change the effective path length of the operator, thereby increasing the fractal dimension to the range of 1.5 to 1.6. The multi-objective optimization system for the high-voltage silicon carbide module adjusts the physical size of the minimum self-similar unit to control the unit spacing, ensuring that the impedance distribution and interference spectrum characteristics achieve logical matching.

[0044] A quantitative comparative test was performed on the physical performance of the multi-objective optimization system for high-voltage silicon carbide modules under different topology parameters. The experimental results are shown in the table below:

[0045] The comparative data above show that there is a specific physical performance window between the fractal dimension and 1.5 and 1.6. When the fractal dimension is below 1.5, the admittance compensation generated by the microelectrode array is insufficient, resulting in substandard smoothing of the electric field gradient at the three-phase point, and the resonant centroid of the parasitic loop cannot avoid interference spectrum peaks. When the fractal dimension is above 1.6, although the field strength suppression effect tends to stabilize, the local parasitic loss increases significantly.

[0046] Through the aforementioned iterative parameter search, the multi-objective optimization system for the high-voltage silicon carbide module seeks the optimal solution within the constraints. When the frequency difference reaches a preset convergence threshold, i.e., when the centroid of the resonant frequency matches the peak value of the interference spectrum, the multi-objective optimization system for the high-voltage silicon carbide module locks the current impedance topology parameters. Locking the impedance topology parameters signifies the achievement of collaborative optimization at the logical level for the multi-objective optimization method for the high-voltage silicon carbide module. Subsequently, the relevant impedance topology parameters will be output to guide the physical construction of the functional areas with field-controlled response characteristics within the high-voltage silicon carbide module.

[0047] After the impedance topology parameters are locked, the multi-objective optimization system for the high-voltage silicon carbide module guides the physical implementation of the functional region with field-controlled response characteristics. This functional region is positioned at the interface between the substrate edge and the insulating medium. Inside this functional region, a nonlinear conductive material with a nonlinear constitutive relation is filled. The critical electric field strength of the nonlinear conductive material is set between 3 kV / mm and 5 kV / mm. When the transient electric field distribution vector magnitude of the target interface exceeds the critical electric field strength, the charge carriers inside the nonlinear conductive material are stimulated to undergo nonlinear migration, causing the functional region with field-controlled response characteristics to transition from a high-resistivity state to a controlled conduction state.

[0048] In a specific implementation, the nonlinear conductive material is selected from epoxy resin-based composite dielectrics doped with zinc oxide (ZnO) microparticles, or silicone rubber composite dielectrics doped with silicon carbide (SiC) particles. The multi-objective optimization system for high-voltage silicon carbide modules adjusts the doping volume fraction of the microparticles to ensure that the critical conduction field strength of the nonlinear conductive material falls between 3 kV / mm and 5 kV / mm.

[0049] The field reshaping effect at the three-phase point at the liner edge was verified through comparison of numerical simulation and measured data. In a high-voltage silicon carbide module without the application of a multi-objective optimization method for the module, the maximum electric field mode at the three-phase point at the liner edge reached 22 kV / mm. After applying the multi-objective optimization method for the high-voltage silicon carbide module and completing the impedance topology parameter tuning, the maximum electric field mode at the three-phase point at the liner edge was suppressed to below 14 kV / mm, and the electric field intensity distribution showed a significant homogenization trend. This field intensity homogenization effect reduces the risk of interfacial partial discharge.

[0050] The programmed implementation of the multi-objective optimization method for high-voltage silicon carbide modules is achieved through logical interaction between a computer-readable storage medium and a processor. The computer-readable storage medium stores computer programs that implement electric field characteristic quantization, response mapping processing, frequency tuning calculation, and topology parameter tuning. Upon receiving the power branch excitation parameter sequence, the processor retrieves the mapping function and frequency coupling equation from the computer-readable storage medium. The mapping model and tuning algorithm are stored in the instruction register in machine code form and are calculated using the processor's arithmetic logic unit. During execution, the processor allocates dedicated data buffers in memory for the dynamic admittance compensation scalar and resonant frequency offset. Through the collaboration between the computer-readable storage medium and the processor, the multi-objective optimization method for high-voltage silicon carbide modules achieves automated operation from physical field perception to topology parameter control. Thus, the multi-objective optimization method for high-voltage silicon carbide modules achieves synergistic optimization under the dimensions of insulation strength and electromagnetic compatibility.

[0051] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-objective optimization method for high-voltage silicon carbide modules, characterized in that, Includes the following steps: S1. Electric field feature quantization: Obtain the excitation parameter sequence of the high voltage silicon carbide module during the switching transient process, quantize the electric field distribution vector of the target boundary domain in the time domain, and determine the equivalent geometric capacitance of the target boundary domain. S2. Response Mapping Processing: The electric field distribution vector is input into a mapping function to generate a dynamic admittance compensation scalar; the mapping function has a preset correspondence between the admittance change and the electric field strength based on the response threshold; S3. Frequency tuning calculation: Determine the impedance offset of the target boundary region based on the dynamic admittance compensation scalar and equivalent geometric capacitance, and calculate the resonant frequency offset of the parasitic circuit in the high-voltage silicon carbide module. S4. Topology parameter tuning: Based on the matching value between the resonant frequency offset and the preset peak value of the interference spectrum in the interference spectrum characteristics, the impedance topology parameters of the target boundary region are tuned.

2. The method according to claim 1, characterized in that, The mapping function in S2 satisfies the following formula: in, This is the dynamic admittance compensation scalar. This is the upper limit of admittance compensation. The ratio of dimensionless field strength. This is the gain coefficient. As a saturation adjustment factor, The damping coefficient; It is defined as the ratio of the magnitude of the electric field distribution vector to the response threshold.

3. The method according to claim 1, characterized in that, The resonant frequency offset in S3 Satisfy the following formula: in, This is the resonant frequency offset. As the reference resonant frequency, The equivalent geometric capacitance of the target boundary region. The equivalent time response factor is used to compensate the dynamic admittance scalar. It is converted into an equivalent dynamic charge storage increment.

4. The method according to claim 1, characterized in that, The impedance topology parameters include the fractal dimension of the impedance distribution and the cell spacing, wherein the fractal dimension is between 1.5 and 1.6; the fractal dimension is determined by adjusting the number of sub-cells retained and the scaling ratio of the generator operator of the generalized fractal structure.

5. The method according to claim 1, characterized in that, The method is applied to a functional area with field-controlled response characteristics, which is located at the junction of the liner edge and the insulating medium.

6. A multi-objective optimization system for high-voltage silicon carbide modules, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the method as claimed in any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images