A simulation optimization method for irregular heat dissipation structures of aviation DC-DC power supplies

Abstract

This invention discloses a simulation optimization method for irregular heat dissipation structures in aerospace DC-DC power supplies, belonging to the field of computer-aided engineering design. The method includes: partitioning the heat flux density of the heating elements and establishing a non-uniform toothed model containing geometric operators; coupling the geometric operators of adjacent regions through boundary transition functions and applying periodic turbulence excitation to generate adaptive flow field simulation data; constructing a response surface surrogate model and introducing a contradiction elimination factor to automatically identify contradictory regions and trigger local compensation, outputting the optimal parameter combination, and finally performing high-precision simulation verification and fine-tuning. This invention employs multi-physics coupled simulation and intelligent optimization algorithms to resolve the contradiction between lightweight design and heat dissipation performance, improve the R&D efficiency of high-performance aerospace power supplies, and achieve extreme weight reduction and precise thermal control under harsh aerospace operating conditions.

Description

Technical Field

[0001] This invention relates to the field of computer-aided engineering design, and in particular to a simulation optimization method for irregular heat dissipation structures of aviation DC-DC power supplies. Background Technology

[0002] Aviation DC-DC power supplies are core power supply units in avionics systems, and their performance directly affects the reliability and safety of aircraft. With the development of aviation technology, airborne equipment places increasingly stringent demands on the power density, lightweight design, and environmental adaptability of power supplies. Thermal design is a crucial aspect of ensuring the stable operation of aviation DC-DC power supplies under high power density conditions; its quality directly determines the power supply's temperature rise, lifespan, and overall weight. Against this backdrop, computer-aided engineering (CAE) technology is widely used in the simulation analysis and design verification of thermal structures.

[0003] In existing technologies, the design of heat dissipation structures for aerospace DC-DC power supplies typically relies on engineers' experience, combined with verification using computer-aided design and simulation software. The design process generally involves engineers first designing a preliminary heat dissipation structure based on experience or traditional design principles, such as using uniformly arranged straight-tooth or finned heat sinks. Then, commercial simulation software is used to build a physical model of the structure and perform thermal-fluid coupling simulation analysis to evaluate its heat dissipation performance. If the simulation results do not meet the design requirements, the engineer returns to the design phase, manually adjusts the structural dimensions or shape, and repeats the simulation verification. This cyclical process of "design-simulation-manual modification" constitutes the main technical means of current heat dissipation structure design.

[0004] However, the aforementioned existing technical solutions have significant drawbacks. First, based on the simple geometry of traditional straight-tooth structures, it is difficult to effectively organize airflow under forced air cooling conditions, easily forming dead zones in critical areas such as the tooth root, leading to localized heat accumulation and low overall heat transfer efficiency. Second, when introducing new lightweight materials such as magnesium alloys, existing simulation processes only treat them as simple heat-conducting media, failing to fully consider and utilize their nonlinear heat conduction characteristics that may be exhibited in complex irregular structures. Finally, the manual iterative design method is inefficient, time-consuming, and labor-intensive, and the optimization process is limited by the designer's experience and intuition, making it difficult to find a globally optimal balance point among multiple conflicting design objectives such as "weight-thermal resistance-flow resistance." The design results are often locally optimal solutions, with the risk of heat dissipation redundancy or exceeding weight limits. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a simulation optimization method for irregular heat dissipation structures in aviation DC-DC power supplies. It employs a closed-loop design logic that integrates heat flux density partitioning, non-uniform geometric modeling, dynamic adaptive simulation, and a conflict resolution optimization strategy. This logic can automatically deduce the optimal heat dissipation structure that balances extreme lightweight design with high-efficiency heat dissipation performance.

[0006] To achieve the above objectives, this application adopts the following technical solution: Firstly, a simulation optimization method for irregular heat dissipation structures of aviation DC-DC power supplies is provided, comprising: acquiring heat flux density distribution data of each heat-generating element inside the aviation DC-DC power supply, and dividing the heat dissipation area on the heat sink into a high-density heat flux region, a medium-density heat flux region, and a low-density heat flux region based on the heat flux density distribution data, thereby obtaining partitioned heat flux region data; for each heat flux region, establishing independent geometric operators composed of tooth cutoff rate parameters, wave trajectory amplitude parameters, and dislocation angle parameters, and nonlinearly coupling the geometric operators of adjacent regions through a boundary transition function for achieving smooth transition, thereby generating a self-compensating non-uniform tooth structure model; applying periodic turbulence excitation to the self-compensating non-uniform tooth structure model for pre-disrupting the laminar boundary layer, and under forced... In a wind-cooled simulation environment, the flow field turbulence and local temperature difference of each tooth cross section are monitored in real time. Based on the flow field turbulence and local temperature difference, the parameters of the corresponding geometric operators are dynamically adjusted to generate adaptive flow field simulation data. Based on the adaptive flow field simulation data, a response surface surrogate model containing the mapping relationship between the parameters of each geometric operator and the heat dissipation performance index is constructed. During the optimization process, a dynamically variable contradiction elimination factor is introduced to resolve design contradictions. The contradiction elimination factor automatically identifies the regions that exacerbate the contradiction between weight and thermal resistance and triggers local compensation operators to output the globally optimal combination of structural parameters. The globally optimal combination of structural parameters is verified by high-precision full-physics field simulation. If there are deviations, local nested fine-tuning is performed on the deviation regions to lock the final irregular heat dissipation structure shape.

[0007] Based on the above technical solution, in the simulation optimization method of the irregular heat dissipation structure of the aviation DC-DC power supply provided in this application, a closed-loop design logic that integrates heat flux density partitioning, non-uniform geometric modeling, dynamic adaptive simulation and contradiction resolution optimization strategy is adopted, which can automatically deduce the best heat dissipation structure that takes into account both extreme lightweight and high-efficiency heat dissipation performance.

[0008] In conjunction with the first aspect mentioned above, in one possible implementation, the geometric operators of adjacent regions are nonlinearly coupled through a boundary transition function used to achieve a smooth transition, generating a self-compensating non-uniform tooth structure model. This includes: determining the exponential decay coefficient of the boundary transition function based on the heat flux density ratio between the high-density heat flux region and the medium-density heat flux region; using the exponential decay coefficient as a weight to perform weighted interpolation on the tooth cutoff rate parameter and wave trajectory amplitude parameter corresponding to the adjacent regions, generating a continuous transition geometric operator; and associating the continuous transition geometric operator with the nonlinear thermal conductivity of a magnesium alloy material whose thermal conductivity varies with temperature, so that the weight reduction effect in the low heat flux region assists in heat dissipation in the high heat flux region through the adjustment of the material's thermal conduction direction, thereby generating a self-compensating non-uniform tooth structure model.

[0009] In conjunction with the first aspect mentioned above, one possible implementation involves applying periodic turbulence excitation to the self-compensating non-uniform toothed structure model to pre-disrupt the laminar boundary layer, and monitoring the flow field turbulence and local temperature difference at each tooth cross-section in real time within a forced air-cooled simulation environment. The parameters of the corresponding region's geometric operators are dynamically adjusted based on the flow field turbulence and local temperature difference. This includes: setting a micro-protrusion array at the tooth root of the self-compensating non-uniform toothed structure model as a periodic turbulence excitation; during the forced air-cooled simulation, collecting the vorticity, turbulent kinetic energy, and velocity gradient at each tooth cross-section through a set of flow field sensor nodes to calculate the flow field turbulence, while simultaneously collecting the local temperature difference; increasing the wave trajectory amplitude parameter in the corresponding region when the flow field turbulence is below the activation threshold used to determine the flow field health; and decreasing the tooth cutoff rate parameter in the corresponding region when the local temperature difference exceeds the upper temperature threshold used to determine the urgency of heat dissipation. Real-time dynamic intervention is performed during the simulation to generate adaptive flow field simulation data.

[0010] In conjunction with the first aspect mentioned above, one possible implementation involves constructing a response surface proxy model based on adaptive flow field simulation data. This model includes the mapping relationship between various geometric operator parameters and heat dissipation performance indicators. During the optimization process, a dynamically variable conflict resolution factor is introduced to resolve design contradictions. This conflict resolution factor automatically identifies regions that exacerbate the weight-thermal resistance conflict and triggers local compensation operators. The process includes: using geometric operator parameters from the adaptive flow field simulation data as input, and using the highest shell temperature, device node temperature, fluid pressure loss, and overall radiator weight as heat dissipation performance indicators to establish a response surface proxy model; during the optimization algorithm iteration, when a set of parameters is detected that causes a decrease in the overall radiator weight while the highest shell temperature exceeds the upper temperature threshold used to determine over-temperature, the conflict resolution factor is automatically activated and located in the region causing the temperature to exceed the limit; within this located region, a local compensation operator is triggered. This local compensation operator independently increases the geometric parameters of the microfins within this region while keeping the geometric operator parameters of other regions unchanged, causing the optimization algorithm to automatically converge towards resolving the conflict and outputting the globally optimal combination of structural parameters.

[0011] In conjunction with the first aspect mentioned above, one possible implementation involves performing high-precision full-physics simulation verification on the optimal combination of global structural parameters. If deviations exist, local nested fine-tuning is performed on the deviation region to lock the final irregular heat dissipation structure form. This includes: substituting the optimal combination of global structural parameters into the three-dimensional physical model corresponding to the forced air-cooling simulation environment for full-field coupled simulation to obtain the verified maximum shell temperature and device node temperature; comparing the deviation between the verified maximum shell temperature and the predicted value of the response surface surrogate model; if the deviation exceeds the allowable range used to measure prediction accuracy, identifying the region where the deviation occurs; within the region where the deviation occurs, fine-tuning the corresponding geometric operator parameters with a step size limited by the original parameter value, repeating the high-precision simulation until the deviation is eliminated, and locking the final irregular heat dissipation structure form.

[0012] In conjunction with the first aspect mentioned above, in one possible implementation, the heat flux density distribution data of each heat-generating element inside the aviation DC-DC power supply is obtained, and the heat dissipation area on the heat sink base is divided into a high-density heat flux zone, a medium-density heat flux zone, and a low-density heat flux zone based on the heat flux density distribution data. The partitioned heat flux zone data includes: obtaining the rated power and layout position of the SiC surface mount device, the IMS substrate, and the high-frequency magnetic integrated transformer; calculating the heat flux density per unit area of ​​the base area directly below each heat-generating element using a thermal resistance network model used to describe the heat transfer path; and dividing the base area into a high-density heat flux zone, a medium-density heat flux zone, and a low-density heat flux zone in descending order of the heat flux density distribution data, and recording the boundary coordinates of each zone to obtain the partitioned heat flux zone data.

[0013] In conjunction with the first aspect mentioned above, one possible implementation involves associating a continuously transitioning geometric operator with the nonlinear thermal conductivity of a magnesium alloy material whose thermal conductivity varies with temperature. This allows the weight reduction effect in the low heat flux region to assist heat dissipation in the high heat flux region by controlling the material's thermal conduction direction. The generation of a self-compensating non-uniform tooth structure model includes: embedding the nonlinear thermal conductivity of the magnesium alloy material as a material property parameter into the continuously transitioning geometric operator; reducing the tooth height by decreasing the tooth profile cutoff ratio parameter in the low heat flux region to achieve weight reduction, and using the nonlinear thermal conductivity to guide heat conduction along the base plane towards the high-density heat flux region, forming a self-compensating heat conduction path; and coupling the self-compensating heat conduction path with the geometric operator parameters to generate a self-compensating non-uniform tooth structure model.

[0014] In conjunction with the first aspect mentioned above, in one possible implementation, during the optimization algorithm iteration process, when it is detected that a certain set of parameters causes the overall weight of the heat sink to decrease while the maximum temperature of the casing exceeds the upper temperature threshold used to determine whether the temperature exceeds the limit, the contradiction elimination factor is automatically activated and locates the region causing the temperature to exceed the limit. This includes: setting a threshold for the reduction of the overall weight of the heat sink and an upper temperature threshold; the contradiction elimination factor monitors the weight reduction and temperature value corresponding to each set of parameters during the iteration process; when the weight reduction exceeds the reduction threshold and the temperature value exceeds the upper temperature threshold, the contradiction elimination factor maps the temperature difference between the maximum temperature of the casing and the upper temperature threshold to an identifier of the region; the identifier drives the local compensation operator to independently add the geometric parameters of the micro fins in the region, thereby achieving local resolution of the contradiction between weight and thermal resistance.

[0015] In conjunction with the first aspect mentioned above, in one possible implementation, during the forced air-cooling simulation, a set of flow field sensor nodes for monitoring the flow field state collects the vorticity, turbulent kinetic energy, and velocity gradient of each tooth section to calculate the flow field turbulence. Simultaneously, the local temperature difference is collected, including: collecting the instantaneous values ​​of vorticity, turbulent kinetic energy, and velocity gradient through the flow field sensor nodes and performing a weighted average to obtain the flow field turbulence; simultaneously collecting the temperature values ​​at each sampling point and calculating the difference between these values ​​and the inlet temperature to obtain the local temperature difference; using the flow field turbulence and local temperature difference together as feedback signals to drive the dynamic adjustment of the geometric operator parameters, forming a collaborative feedback closed loop of flow field turbulence and local temperature difference.

[0016] Secondly, a simulation optimization system for irregular heat dissipation structures of aviation DC-DC power supplies is provided, comprising: a region division module, used to acquire heat flux density distribution data of each heat-generating element inside the aviation DC-DC power supply, and divide the heat dissipation area on the heat dissipation base into high-density heat flux region, medium-density heat flux region, and low-density heat flux region according to the heat flux density distribution data, to obtain partitioned heat flux region data; a model generation module, used to establish independent geometric operators composed of tooth cutoff rate parameters, wave trajectory amplitude parameters, and dislocation angle parameters for each heat flux region, and to nonlinearly couple the geometric operators of adjacent regions through boundary transition functions for achieving smooth transition, to generate a self-compensating non-uniform tooth structure model; and a simulation adjustment module, used to apply periodic turbulence excitation to the self-compensating non-uniform tooth structure model for pre-disrupting the laminar boundary layer. The system monitors the flow field turbulence and local temperature difference at each tooth cross section in real time within a forced air-cooled simulation environment. Based on the flow field turbulence and local temperature difference, it dynamically adjusts the parameters of the corresponding geometric operators to generate adaptive flow field simulation data. The optimization output module constructs a response surface proxy model based on the adaptive flow field simulation data, which includes the mapping relationship between the parameters of each geometric operator and the heat dissipation performance index. During the optimization process, a dynamically variable contradiction elimination factor is introduced to resolve design contradictions. The contradiction elimination factor automatically identifies the regions that exacerbate the contradiction between weight and thermal resistance and triggers local compensation operators to output the globally optimal combination of structural parameters. The locking verification module performs high-precision full-physics simulation verification of the globally optimal combination of structural parameters. If deviations exist, it performs local nested fine-tuning on the deviation regions to lock the final irregular heat dissipation structure shape.

[0017] Compared with the prior art, the present invention has the following advantages: The method provided by this invention combines heat flux density partitioning, non-uniform geometric modeling, dynamic adaptive simulation, and contradiction resolution optimization strategies. Under the premise of ensuring that the temperature of the shell and core components meets the stringent aerospace standards, it can break through the traditional design limitations in balancing weight and heat dissipation performance and obtain a heat dissipation structure with better lightweight performance.

[0018] This invention transforms the traditional "design-verification" process, which relies on manual iteration, into an intelligent design mode that is "goal-driven and automatically generated." The optimization process requires no manual intervention and can automatically explore and converge to the global optimal solution, shortening the development cycle of high-performance aerospace power supply cooling systems and reducing reliance on designers' experience.

[0019] The simulation and optimization logic used in this invention can deeply couple the nonlinear physical properties of materials, the geometric shape of structures, and the dynamic flow field under forced air cooling environment. This makes the final output irregular heat dissipation structure not a simple geometric patchwork, but the optimal product under the synergistic effect of multiple physical fields. This allows for a fuller exploration of the heat dissipation potential of new lightweight aerospace materials and the realization of structural-functional integration.

[0020] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims, and drawings. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A structural architecture diagram of a simulation optimization system for an aerospace DC-DC power supply with an irregular heat dissipation structure provided in this application embodiment; Figure 2 A flowchart illustrating a simulation optimization method for an aerospace DC-DC power supply with an irregular heat dissipation structure, provided in an embodiment of this application. Figure 3 This is a schematic diagram of the evolution of geometric parameters driven by the boundary transition function provided in the embodiments of this application.

[0023] Figure 4 This is a schematic diagram of conflict intensification area identification and location mapping provided in the embodiments of this application. Detailed Implementation

[0024] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0025] The simulation optimization method for irregular heat dissipation structure of aviation DC-DC power supply provided in this application embodiment can be applied to, for example... Figure 1 In the simulation optimization system 100 of an aerospace DC-DC power supply with an irregular heat dissipation structure, as shown, Figure 1 As shown, the system includes: The region division module is used to obtain the heat flux density distribution data of each heat-generating element inside the aviation DC-DC power supply, and divide the heat dissipation area on the heat sink into high-density heat flux area, medium heat flux area and low heat flux area according to the heat flux density distribution data, so as to obtain the regional heat flux area data. The model generation module is used to establish independent geometric operators consisting of tooth cutoff rate parameters, wave trajectory amplitude parameters, and dislocation angle parameters for each heat flow region. The geometric operators of adjacent regions are nonlinearly coupled through boundary transition functions to achieve smooth transition, thereby generating a self-compensating non-uniform tooth structure model. The simulation adjustment module is used to apply periodic turbulence excitation to the self-compensating non-uniform tooth structure model to pre-disrupt the laminar boundary layer, and monitor the flow field turbulence and local temperature difference of each tooth section in real time in a forced air-cooled simulation environment. Based on the flow field turbulence and local temperature difference, the parameters of the corresponding region's geometric operators are dynamically adjusted to generate adaptive flow field simulation data. The optimization output module is used to construct a response surface proxy model based on the adaptive simulation data of the flow field, which includes the mapping relationship between the parameters of various geometric operators and the heat dissipation performance index. During the optimization process, a dynamic and variable contradiction elimination factor is introduced to resolve design contradictions. The contradiction elimination factor automatically identifies the region that intensifies the contradiction between weight and thermal resistance and triggers the local compensation operator to output the global optimal combination of structural parameters. The locking verification module is used to perform high-precision full-physics simulation verification of the optimal combination of global shape and structural parameters. If there is a deviation, it performs local nested fine-tuning on the deviation area to lock the final irregular heat dissipation structure shape.

[0026] like Figure 2 As shown in the figure, this application provides a simulation optimization method for irregular heat dissipation structures of aviation DC-DC power supplies, including: Acquire the heat flux density distribution data of each heat-generating component inside the aviation DC-DC power supply, and divide the heat dissipation area on the heat sink into a high-density heat flux zone, a medium-density heat flux zone, and a low-density heat flux zone based on the heat flux density distribution data to obtain the zoned heat flux zone data. For each heat flow region, an independent geometric operator consisting of tooth cutoff rate parameter, wave trajectory amplitude parameter and dislocation angle parameter is established. The geometric operators of adjacent regions are nonlinearly coupled through the boundary transition function used to achieve smooth transition, generating a self-compensating non-uniform tooth structure model. Periodic turbulence excitation is applied to the self-compensating non-uniform tooth structure model to pre-disrupt the laminar boundary layer, and the flow field turbulence and local temperature difference of each tooth section are monitored in real time in a forced air-cooled simulation environment. The parameters of the corresponding region geometric operators are dynamically adjusted according to the flow field turbulence and local temperature difference to generate adaptive flow field simulation data. Based on adaptive flow field simulation data, a response surface proxy model is constructed, which includes the mapping relationship between various geometric operator parameters and heat dissipation performance indicators. In the optimization process, a dynamic and variable contradiction elimination factor is introduced to resolve design contradictions. The contradiction elimination factor automatically identifies the region that intensifies the contradiction between weight and thermal resistance and triggers the local compensation operator, outputting the global optimal combination of structural parameters. The optimal combination of global structural parameters is verified by high-precision full-physics simulation. If there is a deviation, local nested fine-tuning is performed on the deviation area to lock the final irregular heat dissipation structure shape.

[0027] It should be noted that this invention provides a simulation optimization method for irregular heat dissipation structures of aerospace DC-DC power supplies. Its core technical principle lies in firstly, by accurately partitioning the heat source into heat flux density zones, and based on this, establishing a special self-compensating non-uniform toothed structure model. This model utilizes nonlinearly coupled geometric operators and the physical properties of materials to enable the heat dissipation structure to exhibit different heat dissipation behaviors in different regions. Next, in a dynamically interactive simulation environment, the shape of the heat dissipation teeth is adaptively adjusted by pre-applying turbulence excitation and monitoring the flow and thermal field states in real time. Then, using an intelligent optimization strategy that can proactively identify and resolve design contradictions, a global search is performed on a response surface proxy model constructed from simulation data. This strategy can automatically trigger a local compensation mechanism to solve potential heat dissipation problems when pursuing weight reduction goals. Finally, the theoretical optimal solution is verified through high-precision simulation, and errors are corrected through local nested fine-tuning, thereby locking in the final heat dissipation structure shape after deep coupling of multiple physical fields and iterative optimization by intelligent algorithms.

[0028] In one possible implementation of the embodiments of this application, combined with Figure 2 By nonlinearly coupling the geometric operators of adjacent regions through boundary transition functions to achieve a smooth transition, a self-compensating non-uniform tooth structure model is generated, including: The exponential decay coefficient of the boundary transition function is determined based on the ratio of heat flux density between the high-density heat flux region and the medium-density heat flux region. Using the exponential decay coefficient as a weight, the tooth cutoff rate parameter and wave trajectory amplitude parameter corresponding to adjacent regions are weighted and interpolated to generate a geometric operator for continuous transition. By associating a continuous transition geometric operator with the nonlinear thermal conductivity of a magnesium alloy material whose thermal conductivity varies with temperature, the weight reduction effect in the low heat flux region is assisted by the heat conduction direction of the material to dissipate heat in the high heat flux region, thereby generating a self-compensating non-uniform tooth structure model.

[0029] In some implementations, the heat flux density per unit area of ​​the high-density heat flux region inside the aviation DC-DC power supply is first obtained. The unit is With respect to the heat flux density per unit area in the medium heat flux region The ratio of these values ​​is used to determine the exponential decay coefficient of the boundary transition function. The calculation formula is: ; in, The width of the transition zone between adjacent areas, in units of In this embodiment, The value is set to 15% of the characteristic length of the high-density heat flux region, such as 15% of the shortest distance from the geometric center to the boundary. This parameter ensures a smooth heat flux transition. Subsequently, this exponential decay coefficient is used to construct the boundary transition function: ; in, This represents the normal distance from any point within the transition zone to the boundary of the high-density heat flux region. As weights, parameters derived from a pre-defined initial geometric operator library are weighted and interpolated to generate geometric operators with continuous transitions. The specific interpolation formula is: Tooth cutoff rate parameter: ; Wave trajectory amplitude parameters: ; Dislocation angle parameters: ; in, and Preset initial values ​​for the high-density heat flux region. and Initial values ​​were preset for the medium heat flux region. By performing synchronous weighted interpolation on the dislocation angles, it was ensured that the deflection degree of adjacent heat dissipation teeth in the transition zone between regions evolved nonlinearly and smoothly, eliminating the abrupt changes in the geometric operator at the boundary. To achieve self-compensation at the thermal level, the aforementioned geometric operator was compared with the nonlinear thermal conductivity of the magnesium alloy material. To link, the unit is The thermal conductivity varies with the real-time simulation temperature. The unit is The changes follow the formula: ; in, Reference temperature The basic thermal conductivity is given by the following values. Temperature sensitivity coefficient Value In the low heat flux region, by setting a smaller tooth cutoff ratio parameter... Reducing tooth height to achieve weight reduction, and in simulation engines such as ANSYS or COMSOL, by defining a local coordinate system, the thermal conductivity is... Configured as an anisotropic thermal conductivity tensor matrix, where the thermal conductivity in the base plane direction, i.e., the XY plane, is set to... The thermal conductivity in the direction perpendicular to the base, i.e., the Z direction, is set to [value missing]. This definition of anisotropic material properties, by suppressing vertical heat dissipation and enhancing planar heat conduction, utilizes the physical property of magnesium alloys having stronger thermal conductivity in the low-temperature region to induce heat to spontaneously migrate from the high-temperature region to the low-heat-flux region, forming a self-compensating heat conduction path to assist heat dissipation in the high-density region, ultimately generating a complete self-compensating non-uniform toothed structure model. For example... Figure 3 As shown, the tooth cutoff ratio parameter is displayed. Distance of calculation point from region boundary The evolution curves visually demonstrate the smooth transition characteristics of the boundary transition function during cross-regional coupling.

[0030] For example, in the process of generating a self-compensating non-uniform tooth structure model, the nominal heat flux density of the high-density heat flux region is first obtained. Nominal value of medium heat flow region Set the span of the boundary transition zone Calculate the exponential decay coefficient by substituting the values ​​into the formula. At the boundary of the high-density area At the transition point, the boundary transition function weights are calculated. If the initial tooth cutoff rate of the high-density region... Wave trajectory amplitude The initial value corresponding to the medium heat flow region Then the geometric operator parameters after weighted interpolation at that position are: tooth cutoff rate. Wave trajectory amplitude This achieved nonlinear smooth coupling of the parameters. Subsequently, the above parameters were correlated with the nonlinear thermal conductivity of the magnesium alloy material at the simulation temperature. Below, from the reference temperature The basic thermal conductivity at that time and sensitivity coefficient The real-time thermal conductivity was calculated. By configuring an anisotropic tensor matrix in the simulation, the thermal conductivity of the base plane is set as... ,and Set only to 20% Thus, by utilizing the physical properties of the low heat flux region to assist in heat dissipation in the high heat flux region, the self-compensating non-uniform tooth structure model was constructed.

[0031] In one possible implementation, combining Figure 2 Periodic turbulence excitations to pre-disrupt the laminar boundary layer are applied to a self-compensating non-uniform toothed structure model. The flow field turbulence and local temperature difference at each tooth cross-section are monitored in real time within a forced air-cooled simulation environment. The parameters of the corresponding region's geometric operators are dynamically adjusted based on the flow field turbulence and local temperature difference, including: A micro-protrusion array is set at the root of the tooth in the self-compensating non-uniform tooth structure model as a periodic turbulence excitation. During the forced air cooling simulation, a set of flow field sensor nodes for monitoring the flow field state are used to collect the vorticity, turbulent kinetic energy and velocity gradient of each tooth section, calculate the flow field turbulence, and collect the local temperature difference at the same time. When the flow field turbulence is lower than the activation threshold used to judge the health of the flow field, increase the wave trajectory amplitude parameter in the corresponding region; When the local temperature difference exceeds the upper temperature threshold used to determine the urgency of heat dissipation, the tooth cutoff ratio parameter of the corresponding region is reduced, and real-time dynamic intervention is performed in the simulation to generate adaptive simulation data of the flow field.

[0032] In some implementations, a micro-protrusion array composed of hemispherical mutants is first set at the root of the tooth in the self-compensating non-uniform tooth structure model as a periodic turbulence excitation. The radius r of each hemispherical mutant is set to 10% of the average height of the heat dissipation tooth, and the spacing d is set to 3 times the radius r. During the forced air cooling simulation, the physical quantities of each tooth section are collected by flow field sensors placed at key nodes of the flow field, and the flow field turbulence degree is calculated according to the formula. : ; in, This is the normalized average vorticity value. Turbulent kinetic energy, unit: , Main air velocity at the inlet, unit: , For local velocity gradient, The maximum velocity gradient across the entire field is represented by the weighting ratios of 0.4, 0.3, and 0.3, derived from empirical calibration of laminar boundary layer disruption efficiency. Simultaneously, local temperature differences are collected in real-time. ; in, This represents the current temperature at the monitoring point on the inter-tooth cross section. The inlet ambient temperature is set to 288.15 K. When the flow field turbulence is monitored... Activation threshold for flow field health hour, A value of 0.55 is typically used, indicating significant laminar flow characteristics in the region, which drives the wave trajectory amplitude parameter in the corresponding region. By step size Make adjustments to increase the size: ; in, This is the initial amplitude reference value. When the local temperature difference... Temperatures exceeding the upper limit threshold used to determine the urgency of heat dissipation At that time, the value is 70K, corresponding to the maximum operating temperature rise limit of aviation power supplies, and the tooth cutoff rate parameter of the corresponding driving region. By offset Make a reduction adjustment: ; in, The initial set value of the cutoff rate for the corresponding region in the initial geometric operator library is used as a dimensional benchmark for parameter perturbation. This can be achieved by increasing the effective heat dissipation area to forcibly reduce local temperature rise. The aforementioned feedback adjustment logic is executed in real-time during simulation iterations. Through adaptive parameter correction, it ensures that the flow field state and temperature rise index tend to be balanced, ultimately generating adaptive flow field simulation data that reflects real complex working conditions.

[0033] For example, in a forced air-cooled simulation environment, a set of non-uniform toothed structures in the medium heat flux region of an aerospace DC-DC power supply uses a hemispherical array of micro-protrusions at the tooth roots as periodic turbulence excitation to pre-disrupt the laminar boundary layer. During the simulation, the flow field sensor nodes collect the normalized average vorticity value of a certain tooth cross section in real time. The turbulent kinetic energy is 0.45. for Mainstream air intake speed for The ratio of the local velocity gradient to the maximum velocity gradient across the entire field is 0.4. A weighted average is calculated using a weighting ratio of 0.4:0.3:0.3 to obtain the flow field turbulence degree in this region. Because of this If the value is lower than the preset flow field health activation threshold of 0.55, the laminar flow characteristics at that location are determined to be significant, and parameter adjustment is triggered, adjusting the wave trajectory amplitude parameter in that area. From the initial Increase to This enhances the flow field disturbance. Simultaneously, the sensor monitors the real-time temperature at that point. for Inlet air temperature for The local temperature difference was calculated. Because this temperature difference exceeds the upper limit threshold for emergency heat dissipation. Automatic intervention is triggered, adjusting the tooth cutoff ratio parameter of the corresponding region. The value was reduced from 0.6 to approximately 0.557 to increase the heat dissipation area. Through this real-time dynamic feedback adjustment based on the characteristics of the flow field and thermal field, highly environmentally adaptable flow field adaptive simulation data was finally generated.

[0034] In one possible implementation, combining Figure 2Based on adaptive flow field simulation data, a response surface surrogate model is constructed, which includes the mapping relationship between various geometric operator parameters and heat dissipation performance indicators. During the optimization process, a dynamically variable conflict resolution factor is introduced to resolve design contradictions. This conflict resolution factor automatically identifies regions that exacerbate the conflict between weight and thermal resistance and triggers local compensation operators, including: Using the geometric operator parameters in the adaptive flow field simulation data as input, and the maximum shell temperature, device node temperature, fluid pressure loss and overall weight of the heat sink as heat dissipation performance indicators, a response surface proxy model is established. During the optimization algorithm iteration process, when it is detected that a certain set of parameters causes the overall weight of the heat sink to decrease while the highest temperature of the casing exceeds the upper temperature limit threshold used to determine whether the temperature is over-temperature, the contradiction elimination factor is automatically activated and located to the area that causes the temperature to exceed the limit. Within the defined region, a local compensation operator is triggered. This operator independently increases the geometric parameters of the microfins within the region while keeping the geometric parameters of other regions unchanged. This allows the optimization algorithm to automatically converge toward resolving the contradictions and output the globally optimal combination of structural parameters.

[0035] In some implementations, a response surface surrogate model is first constructed using adaptive flow field simulation data, and a dynamic contradiction elimination factor is introduced: ; in, This represents the total weight of the heat dissipation structure in the current iteration step, in grams. The preset target weight limit is set to 1500g; The equivalent thermal resistance of the entire machine is extracted from the simulation, and the unit is 1. , The preset target thermal resistance threshold is set to 0.12 K / W; weighting coefficient and The values ​​are set to 0.6 and 0.4 respectively. To achieve the "automatic recognition" function, the physical triggering conditions are monitored in real time, specifically whether the reduction in the overall weight of the radiator in the current iteration step exceeds a preset threshold, and whether the highest temperature of the casing exceeds the upper temperature limit threshold. When the above physical conditions are simultaneously met, it indicates that the contradiction between weight reduction and heat dissipation has emerged. At this point, the contradiction elimination factor is initially activated, and further calculations are performed to determine the global contradiction factor. The contribution gradient is determined by summing the partial derivatives of the geometric operator parameters within each partition. : ; in, Representing the Each heat flow partition. When a certain partition's... When the activation threshold of 0.15 is exceeded for three consecutive iterations, it indicates that parameter fine-tuning in this region can no longer simultaneously achieve weight reduction and thermal resistance reduction. This region is identified as an area of ​​heightened conflict, triggering a local compensation operator. The specific execution logic of this local compensation operator is to independently add a set of microfin arrays while maintaining the original wave trajectory in this region. The geometric parameters of the newly added microfins are defined as follows: Fin height: ; in, The original reference height for the heat dissipation fins is 40mm; fin thickness... Set to 20% of the average thickness of the heat dissipation fins in the current area; the fins are arranged at equal intervals along the normal direction of the wave trajectory, with the distribution spacing... This is 1 / 4 of the wavelength of the wave trajectory in the current region. In terms of integration, this compensation process serves as the inner-loop correction logic of the Particle Swarm Optimization (PSO) algorithm. Before each particle position update, i.e., parameter evaluation, the algorithm first determines whether compensation is triggered. If triggered, the geometric properties of the particle are corrected according to the compensation operator, and then it is fed into a high-precision surrogate model to calculate the fitness value. This integration method ensures that the optimization process is performed on a physical structure after "local contradiction resolution," ultimately outputting a combination of irregular structural parameters that satisfies the global optimal objective.

[0036] For example, in the process of constructing a response surface surrogate model that includes the mapping relationship between geometric operator parameters and heat dissipation performance indicators, the tooth cutoff rate parameter of each partition in the adaptive flow field simulation data is used. Wave trajectory amplitude parameters and dislocation angle parameters The highest temperature of the casing and the overall weight of the radiator are used as inputs. When the optimization algorithm iterates to a certain step, a threshold for the weight reduction is set. Preset target weight limit for upper temperature threshold for The current overall weight of the radiator has been monitored. From the initial Reduce to Calculate the decrease Exceeded the threshold, and The weight reduction target was met, but the predicted maximum shell temperature reached [a certain value]. Calculate the local temperature difference using the formula. ,because The contradiction elimination factor is automatically activated and the centroid coordinates of the region are extracted. This is mapped to a spatial identifier. To accurately resolve the conflict, the parameters of this region are further calculated to influence the global conflict factor. Sensitivity gradient If the current iteration step is and Then the calculation yields Because this value exceeded the activation threshold of 0.15 for three consecutive iterations, the local compensation operator was officially activated, increasing the local heat transfer area coefficient of the overheated region while keeping other region parameters constant. Adjustments will be made according to the compensation ratio, such as the new area. By locally increasing the thickness or density of microfins, the optimization process automatically shifts towards mitigating thermal resistance while satisfying the ultimate weight reduction, ultimately outputting the globally optimal combination of irregular structural parameters.

[0037] In one possible implementation, combining Figure 2 High-precision full-physics simulation was performed to verify the optimal combination of global structural parameters. If deviations were found, local nested fine-tuning was performed on the deviation areas to lock in the final irregular heat dissipation structure form, including: The optimal combination of global structural parameters was substituted into the three-dimensional physical model corresponding to the forced air-cooling simulation environment to perform full-field coupling simulation, and the verified maximum shell temperature and device node temperature were obtained. The deviation between the verified maximum shell temperature and the predicted value of the response surface surrogate model is compared. If the deviation exceeds the allowable range used to measure the prediction accuracy, the region where the deviation is located is identified. Within the region where the deviation occurs, the corresponding geometric operator parameters are finely adjusted by perturbation with step size limited by the original parameter values. High-precision simulation is repeated until the deviation is eliminated, and the final irregular heat dissipation structure shape is locked.

[0038] In some implementations, the globally optimal combination of structural parameters output by the optimization algorithm is re-imported into the finite element simulation software to establish a high-precision full-physics coupling verification model with a fine mesh, and the temperature deviation of each heat flux zone is calculated based on the simulation results. ; in, The highest temperature in the region obtained from high-precision simulation, in units of , The upper limit of the target temperature for the project is set to a value of [value to be filled in]. If the temperature deviation in a certain zone Exceeding the allowable threshold The value is Then, the local nested fine-tuning procedure is initiated, targeting the wave trajectory amplitude parameters in that region. Perform gradient-direction-based perturbation fine-tuning. New parameters after perturbation. The calculation formula is: ; in, This is the step size factor, which is fixed at 3% of the original parameter value to limit the perturbation range and maintain global optimality. The perturbation direction operator is calculated as follows: ; That is, when the temperature exceeds the standard A value of 1 drives an increase in amplitude to enhance heat transfer. After fine-tuning, the deviation elimination evaluation index for this region is calculated: ; in, The temperature deviation was obtained from a second simulation after fine-tuning. This represents the initial deviation before fine-tuning. When the convergence condition is met... or At this point, it is determined that the local deviation has been eliminated and the iteration is terminated, ultimately locking the morphological parameters of the irregular heat dissipation structure. The aforementioned fine-tuning step size factor, allowable range threshold, and convergence criterion are all derived from the accuracy requirements of the airborne power supply heat dissipation reliability verification stage, ensuring the consistency between the simulation model and the physical prototype under the extreme weight reduction target.

[0039] For example, in the high-precision full-physics simulation verification stage, the globally optimal combination of structural parameters output by optimization is substituted into a three-dimensional physical model containing a fine mesh for coupled verification. If the engineering target temperature upper limit... The temperature was set to 368.15K, while the simulation showed the highest temperature in a certain zone. If the temperature is 370.15K, then the temperature deviation of this zone can be calculated using the formula. Because this deviation exceeded the permissible threshold. The system automatically identifies the region as a deviation area and initiates local nested fine-tuning. Set the step size factor. The value is 0.03, at which point the perturbation direction operator... If the amplitude parameters of the original wave trajectory in this area If it is 1.5mm, then the new parameters after fine-tuning After fine-tuning, perform a high-precision simulation again. If the new temperature deviation is obtained... Reduce to 0.8K and calculate the deviation elimination evaluation index. Because the convergence condition is met. and After determining that the local deviation had been eliminated, the morphological parameters of the irregular heat dissipation structure were finally locked.

[0040] In one possible implementation, combining Figure 2 The heat flux density distribution data of each heat-generating component inside the aviation DC-DC power supply was obtained. Based on the heat flux density distribution data, the heat dissipation area on the heat sink was divided into high-density heat flux zone, medium-density heat flux zone, and low-density heat flux zone. The resulting zoned heat flux zone data includes: The rated power and layout of SiC surface mount devices, IMS substrates and high-frequency magnetic integrated transformers are obtained. The heat flux density per unit area in the base area directly below each heat-generating element is calculated using a thermal resistance network model to describe the heat transfer path, and the heat flux density distribution data is obtained. Based on the heat flux density distribution data in descending order of value, the base area is divided into high-density heat flux zone, medium-density heat flux zone, and low-density heat flux zone, and the boundary coordinates of each zone are recorded to obtain the heat flux zone data of each zone.

[0041] In some implementations, the coordinates of the boundary region between the high-density heat flux region and the medium-density heat flux region are first determined, and the aforementioned boundary transition function based on exponential decay characteristics is introduced to achieve a smooth evolution of parameters between adjacent zones. The exponential decay coefficient of this function is determined by the ratio of the average heat flux densities of the high and medium zones, reflecting the constraint of the thermal gradient on geometric evolution. This function is then used as a weight to adjust the tooth-shaped cutoff rate parameters of adjacent regions. Perform nonlinear coupling operations, and the resulting evolution parameters Preset cutoff rate in the medium heat flux region With the preset cutoff rate of the high-density heat flux region The resulting continuous transition geometry is obtained through weighted interpolation using this function. Subsequently, the generated continuous transition geometry is correlated with the aforementioned nonlinear thermal conductivity of the magnesium alloy, which varies with temperature. In practice, a user-defined function (UDF) is called in the finite element simulation software to assign dynamically updated material thermal conductivity characteristics to the corresponding geometric grid points. This allows excess heat dissipation capacity in low heat flux regions to assist in cooling high heat flux regions, thereby forming a non-uniform toothed structure model with self-compensation capabilities at the physical level. The above coupling logic ensures that while the irregular structure possesses thermal compensation capabilities, the continuity of the derivatives at its geometric edges meets the requirements of CNC machining for tool path smoothness, avoiding abrupt changes in physical properties at cross-regional connections.

[0042] For example, in the stage of acquiring the heat flux density distribution data of each heat-generating component in an aviation DC-DC power supply, the rated power of the SiC patch device, IMS substrate, and high-frequency magnetic integrated transformer are first obtained as 120W, 45W, and 85W, respectively. The heat flux density is then calculated using a thermal resistance network model based on the layout position. If the projected area of ​​the base directly below the SiC patch is 10cm², the area below the magnetic integrated transformer is 15cm², and the area below the IMS substrate is 20cm², then the heat flux density per unit area for each region is calculated as follows: , The base area is divided into three regions in descending order of heat flux: a high-density heat flux region (SiC region), a medium-density heat flux region (transformer region), and a low-density heat flux region (IMS substrate region). The coordinates of the four corner boundaries of each region are recorded. For example, the boundary of the high-density heat flux region is recorded as follows: to The complete partitioned heat flow region data is output as the input basis for subsequent geometric operator modeling.

[0043] In one possible implementation, combining Figure 2 By associating a continuously transitioning geometric operator with the nonlinear thermal conductivity of a magnesium alloy material whose thermal conductivity varies with temperature, the weight reduction effect in the low heat flux region is assisted in heat dissipation in the high heat flux region by controlling the material's thermal conduction direction. This generates a self-compensating non-uniform toothed structure model, including: The nonlinear thermal conductivity of magnesium alloy is embedded as a material property parameter into a geometric operator with continuous transition. In the low heat flux region, weight reduction is achieved by decreasing the tooth height by reducing the tooth profile cutoff ratio parameter, and heat is guided to the high-density heat flux region along the base plane by using the nonlinear thermal conductivity to form a self-compensating heat conduction path. The self-compensating heat conduction path is coupled with the geometric operator parameters to generate a self-compensating non-uniform tooth structure model.

[0044] In some implementations, the physical morphology of the heat dissipation teeth for each zone is first defined through parametric modeling in the local coordinate system of the heat sink base, based on the zoned heat flow data. Among these, the tooth cutoff ratio parameter... Used to characterize the inclination rate of the heat dissipation tooth side. By adjusting... The side slope of the heat dissipation fins can be varied within the range of 0.5 to 0.8, thereby achieving structural weight reduction while maintaining the heat transfer area. Wave trajectory amplitude parameters. The mathematical expression used to describe the amplitude of the undulation of the heat dissipation teeth along the centerline of the fluid flow direction is as follows: ; in, The coordinates are for the flow direction. The preset fluctuation period is set to 10mm. The value range is set to be 10% to 30% of the average thickness of the heat dissipation fins. Dislocation angle parameter. It is defined as the angle between the center lines of two adjacent rows of heat dissipation fins in the plane of the base, and its calculation formula is: ; in, The angle represents the lateral dislocation displacement between adjacent teeth, and it controls the degree of deflection of the flow field when passing through the gap between the teeth. The set of independent geometric operators consisting of the above three parameters serves as the basis for subsequent nonlinear coupling. Their initial values ​​and adjustment ranges are derived from the manufacturability limit test data of the SiC device heat sink base, ensuring that the generated heat dissipation model meets the requirements of physical simulation while also possessing the possibility of actual manufacturing.

[0045] For example, in the process of generating a self-compensating non-uniform tooth structure model, the nominal heat flux density of the high-density heat flux region is first obtained. Nominal value of medium heat flow region Set the span of the boundary transition zone Calculated according to the exponential decay coefficient formula At the boundary of the high-density area At the transition point, calculate the boundary transition function weights. If the initial tooth cutoff rate of the high-density region Wave trajectory amplitude The initial value corresponding to the medium heat flow region Then the geometric operator parameters after weighted interpolation at this position are the tooth cutoff rate. and wave trajectory amplitude This achieves smooth nonlinear coupling of the parameters. The parameters are then correlated with the nonlinear thermal conductivity of the magnesium alloy material at the simulation temperature. Below, based on reference temperature The basic thermal conductivity at that time and sensitivity coefficient The real-time thermal conductivity was calculated. By configuring an anisotropic tensor matrix in the simulation, the thermal conductivity of the base plane is set as... ,and The thermal conductivity is set to 20%. By reducing the tooth cutoff ratio in the low heat flux region The tooth height was reduced to achieve weight reduction, and the enhanced thermal conductivity of the material in the plane of the base was used to induce heat conduction to the high-density heat flow area, forming a self-compensating heat conduction path to assist heat dissipation, thus completing the model construction.

[0046] In one possible implementation, combining Figure 2 During the optimization algorithm iteration process, when it is detected that a certain set of parameters causes the overall weight of the heat sink to decrease while the highest temperature of the casing exceeds the upper temperature threshold used to determine whether it is overheating, the contradiction elimination factor is automatically activated and locates the areas that cause the temperature to exceed the limit, including: Set thresholds for the reduction of the overall weight of the radiator and the upper limit of the temperature. The contradiction elimination factor monitors the weight reduction and temperature value corresponding to each set of parameters during the iteration process. When the weight reduction exceeds the reduction threshold and the temperature exceeds the upper temperature threshold, the contradiction elimination factor maps the temperature difference between the shell's highest temperature and the upper temperature threshold to the identifier of the region. Identifiers drive local compensation operators to independently increase the geometric parameters of microfins, thereby achieving a local solution to the contradiction between weight and thermal resistance.

[0047] In some implementations, the surface temperature matrix of the heat sink obtained from high-precision full-physics simulation is first acquired, and the upper limit of the temperature of each heated zone relative to the engineering target temperature is obtained using the aforementioned temperature difference calculation formula. Local temperature difference A region identification mapping algorithm is introduced to positively map local temperature differences into spatial location identifiers. Specifically, the logic is as follows: when a local temperature difference in a certain region is detected... And the total weight corresponding to this area The preset target weight limit has been reached. When the value is 1500g, the coordinates of the centroid of the region in the base coordinate system are automatically extracted. This was defined as a potential "conflict intensification zone." To verify the accuracy of the identification, the geometric parameters within this zone were further calculated in relation to the global conflict factor. Sensitivity gradient When the sensitivity gradient Exceeding the preset activation threshold for three consecutive iterations Upon final confirmation that the physical region pointed to by the mapping identifier is the conflict intensification region, an interrupt command is immediately sent to the inner loop of the optimization algorithm to trigger subsequent local compensation operators. The aforementioned region identifier mapping method and gradient activation criterion are derived from sensitivity experimental data of multi-objective optimization algorithms under extreme weight reduction constraints, ensuring that the algorithm can accurately locate the heat dissipation bottleneck region. Figure 4 The image shows the results of identifying conflicting regions on the surface of the heat sink. The black areas in the image represent regions identified based on local temperature differences. Furthermore, the area requiring compensation identified after the weight meets the standard, center The identifier is the centroid coordinate of the locked region.

[0048] For example, during the optimization algorithm iteration process, a threshold for reducing the overall weight of the heat sink is set at 10%, and a preset target weight upper limit is established. and temperature upper limit threshold When the current parameters are monitored, causing the radiator weight to... From the initial Reduce to At that time, the weight reduction was calculated as follows: The reduction exceeded the 10% threshold and met the weight reduction target, but the predicted highest shell temperature at this point... Reached .because The contradiction elimination factor is automatically activated, and the local temperature difference in the area is calculated. The positive temperature difference is mapped into a spatial location identifier using a region identification mapping algorithm, that is, the centroid coordinates of the overheated region are extracted. This identifier then drives the local compensation operator, independently increasing the geometric parameters of the microfins for this specific region while keeping the geometric operator parameters of other regions unchanged; if the original reference height of this region... Current tooth cutoff rate Then calculate the height of the compensated microfins. Furthermore, by arranging such fins at equal intervals in the normal direction in this region, the local heat transfer area is increased, enabling the optimization algorithm to automatically converge towards resolving the contradiction between weight and thermal resistance, and finally outputting the best combination of global structural parameters.

[0049] In one possible implementation, combining Figure 2 During the forced air-cooling simulation, a set of flow field sensor nodes was used to monitor the flow field state, collecting vorticity, turbulent kinetic energy, and velocity gradient at each tooth cross section to calculate the flow field turbulence. Simultaneously, local temperature differences were also collected, including: Instantaneous values ​​of vorticity, turbulent kinetic energy, and velocity gradient are collected by flow field sensor nodes, and the flow field turbulence is obtained by weighted averaging. At the same time, the temperature values ​​at each sampling point are collected, and the difference between the temperature at the sampling point and the temperature at the air inlet is calculated to obtain the local temperature difference; The flow field turbulence and local temperature difference are used together as feedback signals to drive the dynamic adjustment of geometric operator parameters, forming a collaborative feedback closed loop of flow field turbulence and local temperature difference.

[0050] In some implementations, instantaneous values ​​of vorticity, turbulent kinetic energy, and velocity gradient at each tooth cross-section are collected using flow field sensor nodes. Subsequently, a weighted average algorithm is introduced to calculate the flow field turbulence degree. At the same time, real-time temperature values ​​at each sampling point are collected simultaneously. And calculate its relationship with the inlet air temperature. The difference yields the local temperature difference. The above calculations yielded... and Together, they form a set of multidimensional feedback signals that drive the dynamic adjustment of geometric operator parameters. This logic, which transforms instantaneous physical quantities into a disorder index through weighted calculation and interacts with local temperature differences, forms a simulation adaptive closed loop.

[0051] For example, during forced air cooling simulation and dynamic adjustment, sensor nodes placed at key locations in the flow field are used to collect the vorticity of the inter-tooth cross section in real time. Turbulent kinetic energy and local velocity gradient Set the main airflow velocity at the inlet. Maximum velocity gradient in the entire field If the normalized average eddy current is detected at a certain sampling point... Turbulent kinetic energy Local velocity gradient The turbulence degree of the flow field is then calculated using the weighted formula. Meanwhile, the real-time temperature was monitored at this point. If the initial temperature of the air inlet Then calculate the local temperature difference. When the system determines that the flow field turbulence level of 0.395 is lower than the activation threshold of 0.5, it automatically triggers a feedback signal to increase the wave trajectory amplitude parameter in the corresponding region. To enhance turbulence; if local temperature difference If the temperature exceeds the upper limit of the heat dissipation urgency, reduce the tooth cutoff ratio. To free up heat exchange space, adaptive simulation data of the flow field is ultimately generated through this collaborative feedback loop of flow field turbulence and local temperature difference.

[0052] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0053] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A simulation optimization method for irregular heat dissipation structures of aviation DC-DC power supplies, characterized in that, The method includes: The heat flux density distribution data of each heat-generating element inside the aviation DC-DC power supply is obtained, and the heat dissipation area on the heat sink is divided into a high-density heat flux area, a medium-density heat flux area and a low-density heat flux area based on the heat flux density distribution data to obtain the partitioned heat flux area data. For each heat flow region, an independent geometric operator consisting of tooth cutoff rate parameter, wave trajectory amplitude parameter and dislocation angle parameter is established. The geometric operators of adjacent regions are nonlinearly coupled through a boundary transition function to achieve a smooth transition, thereby generating a self-compensating non-uniform tooth structure model. Periodic turbulence excitation for pre-disrupting the laminar boundary layer is applied to the self-compensating non-uniform tooth structure model, and the flow field turbulence and local temperature difference of each tooth section are monitored in real time in a forced air-cooled simulation environment. The parameters of the geometric operator in the corresponding region are dynamically adjusted according to the flow field turbulence and the local temperature difference to generate adaptive flow field simulation data. Based on the adaptive simulation data of the flow field, a response surface proxy model is constructed, which includes the mapping relationship between the parameters of each geometric operator and the heat dissipation performance index. In the optimization process, a dynamic and variable contradiction elimination factor is introduced to resolve the design contradiction. The contradiction elimination factor automatically identifies the region that intensifies the contradiction between weight and thermal resistance and triggers the local compensation operator, outputting the global optimal combination of structural parameters. The optimal combination of global shape and structural parameters is verified by high-precision full-physics simulation. If there is a deviation, local nested fine-tuning is performed on the deviation area to lock the final irregular heat dissipation structure shape.

2. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 1, characterized in that, The step of nonlinearly coupling the geometric operators of adjacent regions using a boundary transition function to achieve a smooth transition, and generating a self-compensating non-uniform tooth structure model, includes: The exponential decay coefficient of the boundary transition function is determined based on the ratio of the heat flux density of the high-density heat flux region to that of the medium-density heat flux region. Using the exponential decay coefficient as a weight, the tooth cutoff rate parameter and the wave trajectory amplitude parameter corresponding to the adjacent regions are weighted and interpolated to generate the geometric operator with continuous transition; The geometric operator of the continuous transition is associated with the nonlinear thermal conductivity of a magnesium alloy material whose thermal conductivity varies with temperature. This allows the weight reduction effect in the low heat flux region to assist in heat dissipation in the high heat flux region by controlling the thermal conductivity direction of the material, thereby generating a self-compensating non-uniform tooth structure model.

3. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 1, characterized in that, The self-compensating non-uniform toothed structure model is subjected to periodic turbulence excitation to pre-disrupt the laminar boundary layer, and the flow field turbulence and local temperature difference of each tooth section are monitored in real time in a forced air-cooled simulation environment. The parameters of the geometric operator in the corresponding region are dynamically adjusted according to the flow field turbulence and the local temperature difference, including: A micro-protrusion array is set at the root of the tooth in the self-compensating non-uniform tooth structure model as a periodic turbulence excitation. During the forced air cooling simulation, a set of flow field sensor nodes for monitoring the flow field state are used to collect the vorticity, turbulent kinetic energy and velocity gradient of each tooth section, calculate the flow field turbulence, and collect the local temperature difference at the same time. When the flow field turbulence is lower than the activation threshold used to determine the health of the flow field, the wave trajectory amplitude parameter in the corresponding region is increased. When the local temperature difference exceeds the upper temperature threshold used to determine the urgency of heat dissipation, the tooth cutoff ratio parameter in the corresponding region is reduced, and real-time dynamic intervention is performed in the simulation to generate adaptive flow field simulation data.

4. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 1, characterized in that, Based on the adaptive flow field simulation data, a response surface proxy model is constructed, which includes the mapping relationship between the parameters of each geometric operator and the heat dissipation performance index. During the optimization process, a dynamically variable conflict resolution factor is introduced to resolve design contradictions. This conflict resolution factor automatically identifies regions that exacerbate the conflict between weight and thermal resistance and triggers local compensation operators, including: Using the geometric operator parameters in the adaptive flow field simulation data as input, and the highest shell temperature, device node temperature, fluid pressure loss and overall weight of the radiator as heat dissipation performance indicators, a response surface proxy model is established. During the optimization algorithm iteration process, when it is detected that a certain set of parameters causes the overall weight of the heat sink to decrease while the highest temperature of the casing exceeds the upper temperature threshold used to determine whether the temperature is over-temperature, the contradiction elimination factor is automatically activated and located to the area that causes the temperature to exceed the standard. The local compensation operator is triggered within the located area. The local compensation operator independently increases the geometric parameters of the microfins within the area, while keeping the geometric operator parameters of other areas unchanged. This allows the optimization algorithm to automatically converge in the direction of resolving the contradictions and output the globally optimal combination of shape and structure parameters.

5. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 1, characterized in that, The optimal combination of global shape and structural parameters is verified by high-precision full-physics simulation. If deviations exist, local nested fine-tuning is performed on the deviation areas to lock in the final irregular heat dissipation structure shape, including: The optimal combination of global structural parameters is substituted into the three-dimensional physical model corresponding to the forced air-cooling simulation environment to perform full-field coupling simulation, and the verified maximum shell temperature and device node temperature are obtained. The deviation between the verified maximum temperature of the shell and the predicted value of the response surface surrogate model is compared. If the deviation exceeds the allowable range used to measure the prediction accuracy, the region where the deviation is located is identified. Within the region where the deviation occurs, the corresponding geometric operator parameters are finely adjusted with a step size limited by the original parameter values. High-precision simulation is repeated until the deviation is eliminated, and the final irregular heat dissipation structure shape is locked.

6. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 1, characterized in that, The process involves acquiring heat flux density distribution data for each heat-generating element inside the aviation DC-DC power supply, and dividing the heat dissipation area on the heat sink base into high-density heat flux zone, medium-density heat flux zone, and low-density heat flux zone based on the heat flux density distribution data, resulting in zoned heat flux zone data including: The rated power and layout of SiC surface mount devices, IMS substrates and high-frequency magnetic integrated transformers are obtained. The heat flux density per unit area in the base area directly below each heat-generating element is calculated using a thermal resistance network model to describe the heat transfer path, and the heat flux density distribution data is obtained. According to the heat flux density distribution data in descending order of value, the base area is divided into a high-density heat flux zone, a medium-density heat flux zone, and a low-density heat flux zone, and the boundary coordinates of each zone are recorded to obtain the zoned heat flux zone data.

7. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 2, characterized in that, The geometric operator of the continuous transition is correlated with the nonlinear thermal conductivity of a magnesium alloy material whose thermal conductivity varies with temperature. This allows the weight reduction effect in the low heat flux region to assist heat dissipation in the high heat flux region by controlling the material's thermal conduction direction. The resulting self-compensating non-uniform tooth structure model includes: The nonlinear thermal conductivity of the magnesium alloy material is embedded as a material property parameter into the geometric operator of the continuous transition. In the low heat flux region, weight reduction is achieved by decreasing the tooth height by reducing the tooth profile cutoff ratio parameter, and the nonlinear thermal conductivity is used to guide heat to be conducted along the base plane direction to the high-density heat flux region, forming a self-compensating heat conduction path; The self-compensating heat conduction path is coupled with the geometric operator parameters to generate a self-compensating non-uniform tooth structure model.

8. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 4, characterized in that, During the iteration of the optimization algorithm, when it is detected that a certain set of parameters causes the overall weight of the heat sink to decrease while the highest temperature of the casing exceeds the upper temperature threshold used to determine whether the temperature is over-limited, the contradiction elimination factor is automatically activated and locates the region causing the temperature to exceed the limit, including: The threshold for the reduction of the overall weight of the radiator and the upper temperature threshold are set, and the contradiction elimination factor monitors the weight reduction and temperature value corresponding to each set of parameters during the iteration process. When the weight reduction exceeds the reduction threshold and the temperature exceeds the upper temperature threshold, the contradiction elimination factor maps the temperature difference between the highest temperature of the casing and the upper temperature threshold to an identifier of the region. The identifier drives the local compensation operator to independently increase the geometric parameters of the microfins in the region, thereby achieving a local resolution of the contradiction between weight and thermal resistance.

9. The simulation optimization method for an irregular heat dissipation structure of an aviation DC-DC power supply according to claim 3, characterized in that, In the forced air-cooling simulation, a set of flow field sensor nodes for monitoring the flow field state collects the vorticity, turbulent kinetic energy, and velocity gradient of each tooth cross section to calculate the flow field turbulence. Simultaneously, local temperature differences are collected, including: The instantaneous values ​​of vorticity, turbulent kinetic energy, and velocity gradient are collected by the flow field sensor nodes and then weighted and averaged to obtain the flow field turbulence degree. At the same time, the temperature values ​​at each sampling point are collected, and the difference between the temperature at the sampling point and the temperature at the air inlet is calculated to obtain the local temperature difference; The flow field turbulence and the local temperature difference are used together as feedback signals to drive the dynamic adjustment of the geometric operator parameters, forming a collaborative feedback closed loop of flow field turbulence and local temperature difference.

10. A simulation optimization system for an irregularly shaped heat dissipation structure of an aviation DC-DC power supply, characterized in that, The system is used for a simulation optimization method of an aerospace DC-DC power supply irregular heat dissipation structure as described in any one of claims 1-9, the system comprising: The region division module is used to acquire the heat flux density distribution data of each heat-generating element inside the aviation DC-DC power supply, and divide the heat dissipation area on the heat dissipation base into a high-density heat flux area, a medium-density heat flux area and a low-density heat flux area according to the heat flux density distribution data, so as to obtain the partitioned heat flux area data. The model generation module is used to establish independent geometric operators consisting of tooth cutoff rate parameters, wave trajectory amplitude parameters, and dislocation angle parameters for each heat flow region, and to nonlinearly couple the geometric operators of adjacent regions through boundary transition functions to achieve smooth transition, thereby generating a self-compensating non-uniform tooth structure model. The simulation adjustment module is used to apply periodic turbulence excitation to the self-compensating non-uniform tooth structure model to pre-disrupt the laminar boundary layer, and to monitor the flow field turbulence and local temperature difference of each tooth section in real time in a forced air-cooled simulation environment. Based on the flow field turbulence and local temperature difference, the module dynamically adjusts the parameters of the geometric operator in the corresponding region to generate adaptive flow field simulation data. The optimization output module is used to construct a response surface proxy model based on the adaptive simulation data of the flow field, which includes the mapping relationship between the parameters of each geometric operator and the heat dissipation performance index. In the optimization process, a dynamic and variable contradiction elimination factor is introduced to resolve the design contradiction. The contradiction elimination factor automatically identifies the region that intensifies the contradiction between weight and thermal resistance and triggers the local compensation operator to output the global optimal combination of structural parameters. The locking verification module is used to perform high-precision full-physics field simulation verification on the optimal combination of global shape structure parameters. If there is a deviation, local nested fine-tuning is performed on the deviation area to lock the final irregular heat dissipation structure shape.

Images