Multi-physical coupling simulation method and device for millimeter wave radar radio frequency integrated microsystem

Abstract

The invention provides a multi-physical coupling simulation method and device for a millimeter wave radar radio frequency integrated microsystem, and belongs to the technical field of simulation. The method provided by the invention comprises the following steps: constructing an initial physical model; performing electromagnetic simulation on the initial physical model to obtain electromagnetic performance and component power loss; performing thermal simulation on the initial physical model by taking the power loss of the component as a heat source to obtain the thermal performance of the initial physical model; constructing a multi-objective optimization problem based on the electromagnetic performance and the thermal performance, and obtaining a parameter group of the initial physical model; establishing a Pareto frontier solution set based on the plurality of parameter groups; and determining a target parameter group from the Pareto leading-edge solution set based on an engineering target, and taking the target parameter group as an operation parameter group of the initial physical model. The invention provides a multi-physical coupling simulation method and device for a millimeter wave radar radio frequency integrated microsystem. The method and device are used for improving the design efficiency and precision of the millimeter wave radar radio frequency integrated microsystem.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of simulation, in particular to a multi-physical coupling simulation method and device for a millimeter wave radar RF integrated microsystem. BACKGROUND

[0002] With the development of modern electronic devices, power devices, microwave systems and new energy devices towards high integration, high efficiency and high reliability, the internal physical phenomena increasingly present complex characteristics of multi-field coupling. For example, in key components such as RF power amplifiers, motors, IGBT modules, the electromagnetic field distribution directly affects the energy conversion efficiency and signal transmission performance, and the heat generated by electromagnetic loss will significantly affect the temperature rise, thermal stress and long-term working stability of the device. Therefore, it is difficult to meet the high performance requirements by designing from a single physical field, and it is urgent to carry out electromagnetic-thermal multi-physical field coupling analysis and collaborative optimization.

[0003] At present, mainstream engineering simulation software such as ANSYS HFSS, ANSYS Workbench, etc. has powerful electromagnetic field and thermal field simulation capabilities, and can perform high-precision modeling and solving on electric field, magnetic field, temperature field, etc. However, the traditional design process usually completes electromagnetic simulation to obtain power loss data first, and then inputs the loss as a heat source into thermal simulation for calculation. This serial coupling method can realize basic multi-physical field analysis, but lacks the ability of global optimization of design parameters, and it is difficult to achieve optimal balance among multiple conflicting objectives. In addition, due to the high cost of multi-physical field simulation calculation, each parameter adjustment needs to repeat the complex modeling and solving process, resulting in long design iteration period and low efficiency. In order to improve the optimization efficiency, existing research attempts to introduce intelligent optimization algorithms such as genetic algorithm, particle swarm optimization, etc. to automatically search for optimal design parameters. However, most of these algorithms in the prior art run independently, and the interface between them and the simulation software is not unified, the automation degree is low, and there is a lack of effective handling mechanism for multi-objective trade-off problems. SUMMARY

[0004] Therefore, the present application provides a multi-physical coupling simulation method and device for a millimeter wave radar RF integrated microsystem to improve the design efficiency and accuracy of the millimeter wave radar RF integrated microsystem.

[0005] Specifically, the present application is realized by the following technical solutions:

[0006] The first aspect of the present application provides a multi-physical coupling simulation method for a millimeter wave radar RF integrated microsystem, the method comprising:

[0007] constructing an initial physical model according to the array antenna of the millimeter wave radar front end;

[0008] perform electromagnetic simulation on the initial physical model to obtain electromagnetic performance and component power loss;

[0009] perform thermal simulation on the initial physical model by taking the component power loss as a heat source by using an ANSYS Workbench platform to obtain thermal performance of the initial physical model;

[0010] construct a multi-objective optimization problem based on the electromagnetic performance and the thermal performance, and optimize the multi-objective optimization problem based on an optimization algorithm to obtain a parameter group of the initial physical model;

[0011] complete multiple rounds of iterative calculation by cyclically calling an Ansys simulation interface to obtain multiple parameter groups, and establish a Pareto frontier solution set based on the multiple parameter groups;

[0012] determine a target parameter group from the Pareto frontier solution set based on an engineering target, and take the target parameter group as a running parameter group of the initial physical model.

[0013] The second aspect of the application provides a multi-physical coupling simulation device for a millimeter wave radar RF integrated microsystem, the device comprising a construction module, a simulation module and a calculation module; wherein,

[0014] The construction module is configured to construct an initial physical model according to an array antenna of a millimeter wave radar front end.

[0015] The simulation module is configured to perform electromagnetic simulation on the initial physical model to obtain electromagnetic performance and component power loss.

[0016] The simulation module is further configured to perform thermal simulation on the initial physical model by taking the component power loss as a heat source by using an ANSYS Workbench platform to obtain thermal performance of the initial physical model.

[0017] The calculation module is configured to construct a multi-objective optimization problem based on the electromagnetic performance and the thermal performance, and optimize the multi-objective optimization problem based on an optimization algorithm to obtain a parameter group of the initial physical model.

[0018] The calculation module is further configured to complete multiple rounds of iterative calculation by cyclically calling an Ansys simulation interface to obtain multiple parameter groups, and establish a Pareto frontier solution set based on the multiple parameter groups.

[0019] The calculation module is further configured to determine a target parameter group from the Pareto frontier solution set based on an engineering target, and take the target parameter group as a running parameter group of the initial physical model.

[0020] The multi-physics coupling simulation method and apparatus for millimeter-wave radar RF integrated microsystems provided in this application improve the design efficiency and accuracy of millimeter-wave radar RF integrated microsystems, while solving the problems of disconnected multi-physics field analysis, reliance on manual trial and error in optimization processes, and difficulty in balancing multiple performance objectives in traditional designs. Specifically, by using the results of electromagnetic simulation as the heat source for thermal field simulation, coupled simulation is performed, breaking the limitations of single-physics field analysis. This accurately captures the interaction between electromagnetic and thermal performance during device operation, avoiding design deviations caused by ignoring inter-field correlations and ensuring consistency between simulation results and actual operating conditions. With the help of multi-objective optimization and iterative optimization mechanisms, multiple sets of candidate parameters can be efficiently generated without repeated manual parameter adjustments, significantly shortening the design iteration cycle. Furthermore, the construction of the Pareto front solution set and the engineering goal-oriented parameter selection can find the optimal balance between the mutual constraints of electromagnetic and thermal performance, ensuring that the final determined set of operating parameters not only meets the actual operating requirements of millimeter-wave radar but also provides feasible technical support for the reliable design of high-density, high-frequency, and high-power integrated systems. Attached Figure Description

[0021] Figure 1 A flowchart of an embodiment of the multi-physics coupling simulation method for millimeter-wave radar RF integrated microsystems provided in this application;

[0022] Figure 2 This is a schematic diagram of the structure of Embodiment 2 of the multi-physics coupling simulation device for millimeter-wave radar radio frequency integrated microsystems provided in this application. Detailed Implementation

[0023] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0024] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0025] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0026] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0027] Figure 1 This is a flowchart of an embodiment of the multi-physics coupling simulation method for millimeter-wave radar RF integrated microsystems provided in this application. Please refer to... Figure 1 The method provided in this embodiment may include:

[0028] S101. Construct an initial physical model based on the array antenna at the front end of the millimeter-wave radar.

[0029] Specifically, the structural parameters of the array antenna should be clearly defined, including at least the number of array elements, the spacing between array elements, the array arrangement, and the feeding topology. Based on the electromagnetic characteristics requirements of the millimeter-wave radar target's operating frequency band, the geometric model of the initial physical model should be completed using a 3D modeling tool. Key parameters of the model should be defined, including the core geometric dimensions of the array elements, the material properties and thickness of the antenna substrate, the material and dimensions of the metal radiating layer, and the boundary condition interface should be reserved for subsequent simulations.

[0030] Furthermore, constructing an initial physical model is fundamental to simulation and optimization. It needs to be based on actual engineering application scenarios to ensure the model accurately reflects the structural characteristics and physical properties of the array antenna. The determination of structural parameters must consider performance requirements such as radar detection range and resolution. For example, the number of array elements directly affects antenna gain and beamwidth, and the element spacing must avoid grating lobes. During geometric modeling, the dimensional accuracy requirements of the millimeter-wave band must be strictly followed. The definition of material properties must refer to the technical parameters of actual commercial materials to ensure consistency between the model and the physical object. The provision of boundary condition interfaces ensures that subsequent electromagnetic and thermal simulations can smoothly call upon model data, avoiding simulation interruptions due to model format incompatibility.

[0031] By precisely defining structural parameters and material properties, a realistic initial physical model is constructed, providing a reliable platform for subsequent multiphysics simulations. This avoids simulation distortions caused by oversimplification of the model or ambiguous parameter definitions, ensuring that subsequent electromagnetic-thermal coupling analysis and multi-objective optimization can be carried out based on real device characteristics, laying the foundation for the engineering implementation of the final design scheme.

[0032] Furthermore, the implementation steps for constructing the initial physical model based on the array antenna of the millimeter-wave radar front end include:

[0033] (1) Determine the structural parameters of the array antenna;

[0034] Specifically, the structural parameters include at least the number of array elements, the spacing between array elements, the array arrangement, and the feeding topology. The range of the number of array elements is determined by combining the detection range, resolution requirements, and engineering installation space of the millimeter-wave radar. The spacing between array elements is set according to the center wavelength of the target's operating frequency band. The array arrangement is selected based on the radar beam pointing requirements. The feeding topology is determined by combining the power distribution uniformity and signal loss requirements of the array antenna.

[0035] Furthermore, the structural parameters of the array antenna directly determine its electromagnetic performance and engineering applicability, and must be developed around the core functional requirements and practical application scenarios of millimeter-wave radar. Regarding the number of array elements, more elements can improve antenna gain and beamforming capabilities, but increase structural complexity and cost, requiring a balance between performance and cost. For example, automotive millimeter-wave radar commonly uses 8-16 elements to meet short-to-medium range detection needs. The element spacing must strictly match the wavelength. If the spacing is greater than the center wavelength, grating lobes (extra radiated beams) are likely to appear, leading to signal interference. If the spacing is too small, it will increase mutual coupling between elements, affecting signal transmission efficiency. The choice of arrangement method must fit the radar's scanning requirements. Linear arrays are commonly used in one-dimensional scanning scenarios such as traffic flow monitoring, while planar arrays are suitable for all-around environmental perception in autonomous driving. The feeding topology must be combined with integration requirements. Microstrip feeding is suitable for high-density integration of RF microsystems, while coaxial feeding is more suitable for high-power radar scenarios.

[0036] (2) Perform geometric modeling of the initial physical model based on the structural parameters and the target operating frequency band.

[0037] Specifically, based on the determined structural parameters, the geometric structure of the array antenna is drawn using 3D modeling tools; combined with the electromagnetic characteristics of the target operating frequency band, the material properties of the model are defined, with a material with stable dielectric constant selected for the substrate and a material with high conductivity selected for the metal radiating layer; simulation interfaces are reserved, and the locations of the feed ports and the boundaries of the radiating surface required for electromagnetic simulation, as well as the heat dissipation contact surfaces required for thermal simulation, are clearly marked in the model to ensure that subsequent simulations can directly call the data of key areas of the model; the dimensional accuracy of the geometric model is verified to ensure that the annotation error of all structural parameters is less than the preset value, and the definition of material properties is consistent with the technical parameters of actual commercial materials to avoid distortion of simulation results due to model errors.

[0038] Furthermore, the selection of 3D modeling tools must be compatible with subsequent ANSYS series simulation software to ensure that model formats (such as STEP and IGES) can be directly imported into HFSS and Workbench; the annotation of array element dimensions must be combined with the target operating frequency band. For example, the side length of a microstrip patch array element in the 77GHz band is approximately 1.8mm, and it must be calculated strictly according to the relationship between wavelength and dielectric constant to avoid antenna resonant frequency shift due to size deviation; the definition of material properties must refer to the electromagnetic and thermal characteristics of the actual materials. The dielectric constant of the substrate directly affects the impedance matching of the antenna, and the conductivity of the metal determines the magnitude of ohmic loss. If the material parameters deviate too much from reality, it will lead to power loss and temperature field distribution in the simulation that do not match the real situation; reserving simulation interfaces can reduce the workload of model modification before subsequent simulation and improve design efficiency; dimensional accuracy verification is to eliminate human error in the modeling process and ensure the consistency between the model and the actual object.

[0039] By combining structural parameters with the target operating frequency band for geometric modeling, an initial model that accurately reflects the physical characteristics of the array antenna can be generated, providing a high-quality input for subsequent electromagnetic and thermal simulations. Precise definition of material properties ensures the realism of physical phenomena such as electromagnetic loss and heat conduction in the simulation. Reserved simulation interfaces improve the continuity of the design process, and dimensional accuracy verification avoids deviations in simulation results caused by modeling errors. The final constructed initial physical model can accurately simulate the structure and material properties of actual devices and is adaptable to the needs of multi-physics simulations, providing a reliable foundation for subsequent multi-field coupling analysis and optimization design.

[0040] S102. Perform electromagnetic simulation on the initial physical model to obtain electromagnetic performance and component power loss.

[0041] Specifically, the solution frequency range for electromagnetic simulation is set based on the target operating frequency band to ensure coverage of the entire actual operating frequency band of the antenna; simulation boundary conditions such as radiation boundary, ideal electric boundary, ideal magnetic boundary, and waveport boundary are set according to the actual operating scenario and structural characteristics of the array antenna; excitation sources are configured according to the feed topology to match the voltage amplitude, phase, and modulation method of the actual operating mode; adaptive mesh generation technology is used to improve the mesh accuracy in key areas with drastic changes in electromagnetic field gradient (such as the radiation area of ​​array elements and feed ports), and appropriately reduce the accuracy in non-critical areas, before submitting the simulation task and extracting data.

[0042] Furthermore, electromagnetic simulation is conducted using professional tools such as ANSYS HFSS. The core objective is to simulate the electromagnetic response of the array antenna under operating conditions. The solution frequency range must fully cover the target frequency band to ensure that the changes in the antenna's electromagnetic performance throughout its entire operating range are captured. The definition of boundary conditions must closely match actual operating conditions. For example, radiating boundaries simulate the propagation of electromagnetic waves in free space, and ideal electrical boundaries eliminate non-physical electromagnetic interference at interfaces between different materials. The configuration of the excitation source must be consistent with the actual feeding method to ensure that the signal transmission characteristics in the simulation match the real scene. Mesh generation adopts a strategy of high precision in critical areas and low precision in non-critical areas to balance computational efficiency while ensuring simulation accuracy. This avoids excessive computational costs due to overly dense meshes or errors in results due to overly sparse meshes. The electromagnetic performance parameters extracted after simulation (such as S-parameters, gain, and field strength uniformity) reflect the antenna's signal transmission and radiation capabilities. The power loss of components (such as ohmic loss and dielectric loss) is the core heat source input for subsequent thermal simulation.

[0043] Through refined simulation settings and data extraction, the electromagnetic performance and component power loss of the array antenna can be accurately obtained, providing accurate heat source data support for thermal simulation. Compared with the traditional method of roughly estimating losses, this step directly extracts the loss distribution through electromagnetic simulation, improving the accuracy of heat source location and power values. At the same time, comprehensive electromagnetic performance parameters provide a clear optimization direction for subsequent multi-objective optimization, avoiding the problem of one-sided optimization objectives due to missing electromagnetic performance data.

[0044] Furthermore, the specific steps for performing electromagnetic simulation on the initial physical model include:

[0045] (1) Set the solution frequency range of the electromagnetic simulation based on the target operating frequency band;

[0046] Specifically, ANSYS HFSS is a professional high-frequency structural simulation software. To determine the target operating frequency band of the millimeter-wave radar array antenna, the lower and upper limits of the solution frequency range are determined based on the center frequency of the target operating frequency band. The solution frequency range must cover the entire target operating frequency band. In the ANSYS HFSS simulation environment, the set solution frequency range is entered into the analysis settings module, and the frequency scan type (such as discrete frequency scan or interpolated frequency scan) is specified to ensure that the electromagnetic response at each frequency point within the frequency band can be captured.

[0047] Furthermore, setting the solution frequency range is fundamental to electromagnetic simulation. It needs to closely match the actual operating requirements of the array antenna. Different application scenarios of millimeter-wave radar correspond to fixed operating frequency bands (such as the 77GHz band commonly used for automotive radar and the 24GHz band commonly used for security radar). Setting the solution range with the target frequency band as the core can ensure that the simulation focuses on the frequency range in which the antenna actually operates. The choice of frequency scan type needs to balance accuracy and efficiency. Discrete frequency scanning is suitable for scenarios that require accurate calculation of the performance of specific frequency points, while interpolated frequency scanning can shorten the calculation time while ensuring accuracy, adapting to the need for rapid analysis of multiple frequency points.

[0048] (2) Determine the simulation boundary conditions based on the actual working scenario and structural characteristics of the array antenna;

[0049] Specifically, for the radiating surface of the array antenna, a radiation boundary is set according to the actual free space environment. The distance between the radiation boundary and the antenna radiating surface is set to 1 / 4 of the shortest wavelength of the target operating frequency band to avoid electromagnetic wave reflection interference at the boundary. For the contact surface between the antenna substrate and the metal radiating layer, an ideal electrical boundary is set according to the structural characteristics of metal as a good conductor to eliminate electromagnetic interference from non-physical factors between the interface and simulate the real electromagnetic interaction between the metal and the dielectric. For the fixed structure (such as metal bracket or shell) on the non-radiating side of the array antenna, an ideal magnetic boundary is set according to the actual installation scenario to restrict the parallel component of the magnetic field from passing through the boundary and avoid the fixed structure from having additional impact on the electromagnetic performance. For the feed port, a wave port is set according to the port impedance matching requirements (usually a 50Ω standard impedance), and the port reference impedance and load impedance are defined to ensure that the feed signal transmission characteristics in the simulation are consistent with those in reality.

[0050] Furthermore, the setting of boundary conditions directly determines whether the electromagnetic simulation can accurately simulate the working environment and structural function of the antenna, and must be precisely defined in combination with the scene and structural characteristics. The distance of the radiation boundary is set to 1 / 4 wavelength because it can effectively absorb electromagnetic waves at this distance, simulating the characteristic of electromagnetic waves propagating without reflection in free space. If the distance is too close, it will cause electromagnetic wave reflection, resulting in false signals in the simulation results. The setting of the ideal electric boundary is based on the physical characteristic that the vertical component of the electric field on the surface of a good conductor is zero, which can accurately simulate the electromagnetic behavior of the interface between the metal radiation layer and the substrate, avoiding performance calculation deviations caused by interface interference. The ideal magnetic boundary is for fixed structures that do not participate in electromagnetic radiation. By limiting the magnetic field component, it eliminates the interference of irrelevant structures on the electromagnetic performance of the antenna. The impedance setting of the wave port must match the impedance standard of the feed network in actual engineering to ensure that the reflection and loss of signal transmission in the simulation are consistent with the real working conditions.

[0051] (3) Configure the electromagnetic simulation excitation source according to the feeding topology of the array antenna;

[0052] Specifically, based on the type of feed topology (e.g., microstrip line feed, coaxial feed, slot feed), select the corresponding excitation source type in ANSYS HFSS; configure the excitation source parameters, including the amplitude and phase of the excitation signal (set according to the array beamforming requirements, such as equal-phase excitation to achieve narrow beams), and modulation method; load the excitation source onto the feed port corresponding to the model, ensuring that the loading position of the excitation source is completely consistent with the actual feed point of the feed topology, for example, the excitation source for microstrip line feed is loaded at the connection end between the microstrip line and the array element; verify the correctness of the excitation source configuration, and check the impedance matching of the feed port through pre-simulation. If the matching is not good, adjust the excitation source parameters (e.g., amplitude, phase) or the feed port structure.

[0053] Furthermore, the excitation source is the core of the analog antenna receiving the feed signal. Its configuration must be perfectly matched with the physical characteristics of the feed topology. Different feed structures correspond to different types of excitation sources. For example, microstrip line feeds, because the signal propagates on the surface of the dielectric substrate, are suitable for voltage excitation; coaxial feeds, because the signal propagates between the inner and outer conductors, are suitable for wave excitation. The amplitude of the excitation source needs to be calculated based on the input power of the antenna design to ensure that the power input of the antenna in the simulation is consistent with that in actual operation, avoiding inaccurate loss and field strength calculations due to power deviation. The phase setting is related to the beam pointing of the array antenna. By adjusting the phase difference of the excitation signals of different array elements, beam scanning can be achieved. In the simulation, the phase needs to be set according to the designed beam requirements. The modulation method needs to match the radar's operating mode. Continuous wave excitation is suitable for velocity measurement scenarios, while pulse wave excitation is suitable for ranging scenarios. The accuracy of the excitation source loading position directly affects the signal transmission path. If the loading position is deviated, the feed signal will not be correctly transmitted to the array element, resulting in abnormal simulation results.

[0054] (4) Divide the electromagnetic simulation network according to the relationship between the region and the electromagnetic field change in the initial physical model, and perform electromagnetic simulation.

[0055] Specifically, the electromagnetic field variation gradient in different regions of the initial physical model is analyzed. Regions with drastic electromagnetic field changes (such as the element radiation region, feed port contact region, and inter-element coupling region) are divided into high-precision mesh regions, while regions with gentle electromagnetic field changes (such as the substrate non-radiation region and fixed structure region) are divided into low-precision mesh regions. Mesh generation parameters are set, with the mesh element size in the high-precision mesh region not exceeding 1 / 8 of the shortest wavelength of the target operating frequency band, and the mesh element size in the low-precision mesh region being relaxed to 1 / 4 to 1 / 2 of the wavelength. The adaptive mesh generation function of ANSYS HFSS is enabled, and mesh convergence criteria are set to ensure that the mesh accuracy meets the simulation requirements. The electromagnetic simulation task is submitted, and after the simulation is completed, electromagnetic performance parameters (S-parameters, gain, field strength uniformity) and component power loss data (ohmic loss of the metal radiating layer, substrate dielectric loss, and feed port contact loss) are extracted.

[0056] From a frequency band coverage perspective, it can fully capture the changes in electromagnetic performance of the antenna in actual operation and at the edge range, avoiding design defects caused by performance blind spots. From an environmental adaptation perspective, it can realistically simulate the interaction between the spatial environment and structure during antenna operation, eliminating false interference factors and ensuring that the simulation results closely match actual operating conditions. From a signal simulation perspective, it can accurately reproduce the process of the antenna receiving the feed signal, ensuring that key responses such as field strength distribution and signal transmission are consistent with the actual operating state. From the perspective of balancing simulation efficiency and accuracy, it can focus on the core area to ensure calculation accuracy, while avoiding the waste of efficiency caused by over-calculation in non-critical areas. Finally, the electromagnetic performance data and power loss information obtained through this set of operations can clearly reflect the antenna's signal transmission and radiation capabilities, and provide accurate heat source information for thermal simulation, effectively connecting subsequent multi-objective optimization and multi-physics coupling analysis, and helping to improve the design quality and efficiency of millimeter-wave radar RF integrated microsystems.

[0057] S103. Using the power loss of the component as a heat source, perform thermal simulation on the initial physical model using the ANSYS Workbench platform to obtain the thermal performance of the initial physical model.

[0058] Specifically, the ANSYS Workbench platform is an integrated multiphysics simulation and engineering analysis platform. Through a unified user interface, data management system, and collaborative simulation mechanism, it integrates various professional simulation modules under ANSYS (such as thermal simulation, structural mechanics, fluid mechanics, electromagnetic simulation, etc.) to realize multiphysics coupling analysis, engineering optimization, and automated design process. It is the core working environment for complex products from modeling to simulation verification.

[0059] Furthermore, through the ANSYS Workbench import interface, the initial physical model and the component power loss data output from the electromagnetic simulation are fully imported to complete the spatial mapping between the loss data and the model components. Thermal boundary conditions are defined according to the actual heat dissipation scenario. For natural convection heat dissipation, the corresponding ambient temperature and convection heat transfer coefficient are set. For forced air cooling, the fan parameters are matched. For heat dissipation substrate, the heat conduction boundary and contact thermal resistance are set. According to the simulation requirements, steady-state thermal analysis (to evaluate the long-term stable working thermal state) or transient thermal analysis (to analyze temperature changes during start-up, shutdown, and power fluctuations) is selected, the solver accuracy and time step are configured, the simulation is submitted, and the thermal performance data is extracted.

[0060] Furthermore, thermal simulation uses the power loss obtained from electromagnetic simulation as the core input. The temperature field distribution of the array antenna is simulated through the ANSYS Workbench platform. Spatial mapping between the model and the loss data is crucial; it is essential to ensure that the heat source location, power density, and actual loss distribution are completely consistent to avoid temperature field simulation distortion due to heat source misalignment. The thermal boundary conditions must be strictly matched to the actual heat dissipation scheme. For example, the natural convection environment temperature in automotive scenarios must refer to the operating temperature range of automotive electronics, and the convective heat transfer coefficient must be determined based on the airflow state. The selection of the thermal analysis type must be combined with engineering requirements; steady-state analysis is suitable for evaluating the thermal stability of long-term operation, while transient analysis can capture the dynamic temperature change process. Parameters extracted after simulation, such as the maximum temperature rise, temperature gradient, and heat flux density distribution, comprehensively reflect the antenna's thermal performance and heat dissipation bottlenecks.

[0061] By accurately mapping loss data and matching it with actual heat dissipation boundaries, the temperature field changes during antenna operation can be realistically simulated, and areas at risk of thermal stress (such as feed ports and array element centers) can be accurately identified. The acquired thermal performance data, together with the electromagnetic performance data mentioned above, constitute the core basis for multi-objective optimization, providing data support for balancing electromagnetic performance and thermal stability, and avoiding the problem of design failure due to overheating in actual operation caused by ignoring thermal effects.

[0062] Furthermore, the power loss of the aforementioned component is used as a heat source, and thermal simulation of the initial physical model is performed using the ANSYS Workbench platform; including:

[0063] (1) Import the initial physical model into the thermal simulation module based on the import interface of ANSYS Workbench, call the component power loss output by the electromagnetic simulation, and perform spatial mapping based on the component power loss;

[0064] Specifically, the initial physical model is completely imported into the thermal simulation module through the dedicated import interface of ANSYS Workbench, ensuring that the model's geometry and material properties are completely consistent with those in the electromagnetic simulation stage, thus avoiding thermal analysis distortion due to model differences. The component power loss data output by the electromagnetic simulation (such as ANSYS HFSS), including the ohmic loss of the metal radiating layer, the dielectric loss of the substrate, and the contact loss of the feed port, are called up, which are the core heat sources of the thermal simulation. Through spatial mapping technology, the loss data is accurately matched to the corresponding components and regions of the model (such as mapping the loss of the feed port to the actual feed structure location, and mapping the loss of the array element to the metal radiating layer), ensuring that the location and power density of the heat source are completely consistent with the loss distribution during actual antenna operation, laying the foundation for the accuracy of subsequent thermal analysis.

[0065] (2) Define boundary conditions based on the actual heat dissipation scenario of the array antenna during operation;

[0066] Specifically, thermal boundary conditions directly determine whether thermal simulation can accurately reflect the actual heat dissipation state of the antenna. It is necessary to consider the engineering application scenario of the array antenna, clarifying its actual heat dissipation method and environmental parameters. If the antenna is used in an automotive scenario and employs natural convection cooling, the ambient temperature and convective heat transfer coefficient must be defined. If it is used in equipment requiring forced air cooling, the fan speed and airflow direction must be matched, and a higher convective heat transfer coefficient must be set. If the antenna is mounted in close contact with a heat sink substrate, the contact thermal resistance must also be defined to simulate the heat conduction process from the antenna to the heat sink substrate. By setting realistic boundary conditions, interference from unrealistic heat dissipation factors can be eliminated, ensuring that the thermal field analysis closely reflects engineering reality.

[0067] (3) Determine the type of thermal analysis based on the simulation requirements and submit a thermal simulation task for thermal simulation analysis.

[0068] Specifically, the choice of thermal analysis type depends on the engineering requirements for evaluating thermal performance. If it is necessary to determine the thermal state of the antenna during long-term stable operation (e.g., whether the temperature rise exceeds the standard after several hours of continuous operation), steady-state thermal analysis should be selected, and a sufficiently long simulation time should be set to ensure that the temperature field reaches a stable state. If it is necessary to analyze the temperature changes under dynamic operating conditions of the antenna (e.g., the short-term temperature rise peak at the moment of power-on or during sudden power fluctuations), transient thermal analysis should be selected, and a reasonable time step should be set according to the rate of temperature change (0.1~1s during the rapid temperature rise phase, and 5~10s after the temperature stabilizes) to capture the dynamic change law of temperature over time. After determining the analysis type, the solver accuracy should be configured (e.g., default or high accuracy) to balance computational efficiency and result accuracy. Then, the thermal simulation task should be submitted. After the calculation is completed, thermal performance data such as temperature field distribution, maximum temperature rise, temperature gradient, and heat flux density should be extracted.

[0069] By mapping the model to the spatial loss, the limitation of the traditional disconnect between electromagnetic simulation and thermal simulation is broken. This ensures that the heat source in the thermal simulation comes entirely from the electromagnetic loss of the antenna during actual operation, avoiding errors caused by manual estimation of the heat source. This allows the thermal analysis results to truly reflect the intrinsic relationship between electromagnetic and thermal, providing complete performance data for electromagnetic-thermal synergy for subsequent multi-objective optimization.

[0070] S104. Based on the electromagnetic properties and the thermal properties, construct a multi-objective optimization problem, optimize the multi-objective optimization problem based on the optimization algorithm, and obtain the parameter set of the initial physical model.

[0071] Specifically, an objective function system is constructed. Electromagnetic performance objectives include minimizing return loss and insertion loss, and maximizing gain and field strength uniformity. Thermal performance objectives include minimizing maximum temperature rise and temperature gradient in key areas. Weights are assigned to each objective according to engineering requirements. Optimization variables are selected, covering geometric parameters (element side length, spacing, feed port width, substrate thickness) and material parameters (substrate dielectric constant, metal radiating layer conductivity). Process limits and performance constraints for variable values ​​are set. Algorithms are selected based on the complexity of the optimization problem. Particle swarm optimization is preferred for continuous variables, while genetic algorithms are preferred for problems with discrete variables or complex nonlinear problems. Core algorithm parameters (population size, number of iterations, crossover probability, etc.) are configured. The optimization algorithm and simulation interface are integrated. Optimization variables are iteratively adjusted, performance indicators are verified, and parameter sets are output after convergence conditions are met.

[0072] Furthermore, the core of multi-objective optimization lies in finding the optimal balance between mutually constraining electromagnetic and thermal performance. The construction of the objective function system must cover core performance indicators, and weight allocation must be combined with engineering priorities. For example, automotive radar prioritizes ensuring electromagnetic performance stability, while aerospace radar needs to focus on controlling temperature rise. The selection of optimization variables must focus on key parameters that significantly affect performance, avoiding reduced optimization efficiency due to too many variables. The setting of constraints must consider both technological feasibility and performance baselines. The selection of optimization algorithms must be adapted to the characteristics of the problem. Particle swarm optimization has a fast convergence speed and is suitable for continuous variable optimization; genetic algorithms have strong global search capabilities and are suitable for complex nonlinear problems. By integrating the algorithm with the simulation interface through Python scripts, automated iteration is achieved, ensuring efficient progress of the optimization process.

[0073] By constructing a scientific multi-objective optimization system and selecting suitable optimization algorithms, the traditional design mode relying on manual trial and error is replaced. The automated optimization process significantly shortens the design iteration cycle and avoids the subjectivity and inefficiency of manual parameter adjustment. The output parameter set can achieve an initial balance between conflicting electromagnetic and thermal properties, ensuring both the antenna's signal transmission and radiation capabilities while controlling the temperature rise, providing a high-quality initial parameter foundation for subsequent iterative optimization.

[0074] Furthermore, the specific implementation steps include:

[0075] (1) Construct an objective function system based on the evaluation indicators of the electromagnetic properties and the thermal properties;

[0076] Specifically, focusing on the core performance requirements of millimeter-wave radar array antennas, abstract electromagnetic and thermal performance is transformed into quantifiable and calculable objective functions. First, key evaluation indicators for two types of performance are identified: electromagnetic performance focuses on signal transmission and radiation capabilities, typically including minimizing return loss, minimizing insertion loss, maximizing antenna gain, and maximizing field strength uniformity; thermal performance focuses on temperature control and heat dissipation stability, typically including minimizing maximum temperature rise and minimizing temperature gradients in critical areas (feed ports, array element centers). Weights are assigned to each objective function, with the weight allocation aligning with engineering priorities (e.g., for automotive radar, prioritizing electromagnetic performance stability, with a weight of 50%–60%; for aerospace radar, prioritizing temperature rise control, with a thermal performance weight of 50%–55%), forming a multi-objective collaborative optimization computational system to ensure that the optimization direction is consistent with actual application requirements.

[0077] (2) Select optimization variables from the parameters of the initial physical model;

[0078] Specifically, the selection of optimization variables should focus on parameters that have a significant impact on electromagnetic and thermal performance and are adjustable in engineering, avoiding the reduction of optimization efficiency due to too many variables or the invalidity of optimization due to irrelevant variables. Typically, selection is made from two types of parameters in the initial physical model: geometric parameters are the core, including element side length (affecting antenna resonant frequency), element spacing (affecting beam grating lobes and mutual coupling), feed port width (affecting impedance matching), and substrate thickness (affecting signal transmission loss and heat dissipation efficiency), and their values ​​need to be limited to a specific range (e.g., element side length 5~10mm, within the limits of the bonding process); material parameters are supplementary, including substrate dielectric constant (affecting electromagnetic signal propagation speed) and metallic radiating layer conductivity (affecting ohmic loss), and their values ​​should refer to the available range of commercially available materials (e.g., substrate dielectric constant 2.2~4.5, metallic conductivity 3×10⁻⁶). 7 ~6×10 7 S / m). At the same time, parameters that cannot be adjusted (such as fixed installation structure dimensions) need to be excluded to ensure that the optimization variables are both performance sensitive and engineering feasible.

[0079] (3) Select the target optimization algorithm based on the complexity of the optimization problem and the convergence requirements;

[0080] Specifically, the choice of optimization algorithm must be adapted to the characteristics of the problem, balancing global optimization capability and convergence efficiency. If the optimization variables are mainly continuous (such as substrate thickness and dielectric constant), and a relatively optimal solution needs to be obtained quickly, the particle swarm optimization algorithm is preferred. By simulating the cooperative search of a swarm of particles, it has a fast convergence speed and is suitable for scenarios requiring rapid iteration. If the optimization variables include discrete variables (such as the number of array elements), or if the problem has complex nonlinear relationships (such as the influence of multi-parameter coupling on temperature rise), the genetic algorithm is preferred. By simulating biological evolution through crossover and mutation, it has strong global optimization capability and can avoid getting trapped in local optima, making it suitable for complex multi-constraint scenarios. If the simulation computation cost is high, the Bayesian optimization algorithm is chosen. By building a performance prediction model, it reduces the number of invalid simulations and improves optimization efficiency. At the same time, the core parameters of the algorithm need to be configured to ensure stable operation.

[0081] (4) Perform optimization calculations based on the target optimization algorithm and output the parameter set.

[0082] Specifically, the optimization algorithm is integrated with the Ansys simulation interface via a Python script. Each round of the algorithm generates a new combination of optimization variables. The script automatically calls ANSYS HFSS and Workbench to simulate the electromagnetic and thermal performance corresponding to that variable combination. The simulation results are substituted into the objective function system to calculate the comprehensive performance score of the variable set. The algorithm adjusts the variable search direction for the next round based on the score (e.g., particle swarm optimization moves closer to the optimal solution by updating particle velocity, and genetic algorithms achieve evolution by retaining high-quality individuals). When the convergence condition is met, optimization stops, and a set of parameters that satisfies all performance constraints and achieves a balance between electromagnetic and thermal performance is output. The parameter set must include the specific values ​​of each optimization variable and the corresponding objective function calculation results, providing a foundation for subsequent iterations and solution set construction.

[0083] By using a quantified objective function system, fuzzy performance optimization is transformed into calculable performance indicators, avoiding the subjectivity of adjusting parameters based on experience in traditional optimization. The selection of optimization variables focuses on key and adjustable parameters, reducing the interference of irrelevant variables on the optimization process and lowering the computational complexity of the algorithm. The differentiated selection of algorithms adapts to the characteristics of different problems, avoiding both the inefficiency caused by using complex algorithms for simple problems and the local optimum trap caused by using simple algorithms for complex problems. This ensures that a parameter set that combines performance advantages and engineering feasibility is obtained within a reasonable time. The output parameter set is not a single final solution, but a high-quality initial solution verified by the algorithm. This parameter set has met the basic performance constraints and achieved a preliminary balance between electromagnetic and thermal performance. Subsequent iterations can quickly generate multiple sets of similar high-quality parameters, providing sufficient candidate samples for constructing the Pareto front solution set. This avoids incomplete solution set coverage or performance shortcomings caused by low-quality initial parameters, ultimately helping to efficiently select the optimal operating parameter set that meets the engineering objectives.

[0084] S105. Repeatedly call the Ansys simulation interface to complete multiple rounds of iterative calculations, obtain multiple sets of parameter sets, and establish Pareto front solution sets based on the multiple sets of parameter sets.

[0085] Specifically, a parameterized simulation script is built based on the PyAnsys library, comprising three main modules: parameter input, simulation execution, and result return, to automate the simulation process. Iteration rules are set, generating N new parameter sets in each round, submitting them in batches for parallel computation, and setting the maximum value for each iteration and a performance improvement termination condition (performance change rate ≤ threshold for multiple consecutive rounds). Effective parameter sets are screened, eliminating those that do not meet the constraints, and redundant parameter sets are removed through cluster analysis, retaining M effective parameter sets. The effective parameter sets are sorted using a non-dominated sorting method, and non-dominated parameter sets are selected to form the Pareto front solution set. The front curve is plotted, and performance indicators and parameter values ​​are labeled.

[0086] Global optimization of the parameter set is achieved through iterative iteration. Closed-loop automation of simulation and optimization is realized using Python scripts. The parameterized simulation script automatically receives the parameter set output by the optimization algorithm, calls the Ansys interface to complete the electromagnetic-thermal coupling simulation, and returns performance data, avoiding the tedious manual modeling and simulation. The iteration rules must balance optimization efficiency and result validity. Parallel computing significantly reduces the time cost of multiple simulation rounds, and dual termination conditions ensure that neither the optimal solution is missed nor invalid iterations occur. The selection of effective parameter sets eliminates unsuitable, low-quality parameters and redundant parameters with overlapping performance, reducing the workload of subsequent solution set analysis. The Pareto front solution set is constructed using a non-dominated sorting method. Non-dominated parameter sets represent the optimal trade-off with no room for improvement, meaning it is impossible to improve one performance without degrading another.

[0087] Through automated iteration and global optimization, a Pareto front solution set covering different performance trade-offs is generated, providing designers with a wider range of choices compared to single parameter set outputs. The closed-loop automated process completely solves the inefficiency problem in traditional design, significantly shortening the development cycle. The construction of the Pareto front solution set clearly presents the trade-off between electromagnetic and thermal performance, avoiding design bias caused by single-objective optimization, and providing an intuitive and comprehensive basis for subsequently selecting the optimal solution in conjunction with engineering objectives.

[0088] Furthermore, the specific implementation steps include:

[0089] (1) Construct a parameterized simulation call script based on the PyAnsys library;

[0090] Specifically, the PyAnsys library is a Python programming interface provided by ANSYS, enabling automated control of ANSYS series simulation software (HFSS, Workbench, etc.). It defines an interface to receive the set of parameters to be verified from the optimization algorithm (such as element side length, substrate dielectric constant, feed port width, etc.) and automatically assigns these parameters to the corresponding variables in the ANSYS simulation model, eliminating the need for manual modification of model parameters. It automatically calls ANSYS HFSS to complete electromagnetic simulation (loading boundary conditions, mesh generation, and solution calculation) according to a preset process, and then calls ANSYS Workbench to complete thermal simulation (mapping electromagnetic loss heat sources, setting thermal boundaries, and performing thermal analysis), all without requiring manual software startup or operation. After simulation, it automatically extracts electromagnetic performance data (S-parameters, gain, field strength uniformity) and thermal performance data (maximum temperature rise, temperature gradient), converting the data into numerical matrices or tables recognizable by the optimization algorithm and synchronously feeding it back to the algorithm for subsequent evaluation. Through script-driven automation of the simulation process, it achieves a closed-loop iteration of parameter updates, batch simulations, data extraction, and optimization feedback, until a set of optimized design schemes that simultaneously satisfy excellent electromagnetic performance and good thermal stability is output.

[0091] (2) Based on the simulation call script, perform iterative calculations to obtain multiple sets of parameters;

[0092] Specifically, iteration rules are set; the simulation call script receives parameter sets in batches, and performs electromagnetic-thermal coupling simulations for each parameter set simultaneously through parallel computing, significantly reducing the simulation time for multiple parameter sets; after the simulation is completed, the script feeds back the performance data of each parameter set to the optimization algorithm, which evaluates the parameters based on the objective function system (such as the weighted score of electromagnetic and thermal performance) and selects parameter sets that meet the basic performance constraints for retention. An iteration termination condition is also set, and iteration stops when the termination condition is met.

[0093] (3) Sort the multiple sets of parameters based on the non-dominated sorting method to obtain the Pareto front solution set.

[0094] Specifically, the non-dominated ranking method is a method for selecting the optimal trade-off solution in multi-objective optimization. Its logic is to identify parameter sets that cannot improve one performance without degrading another through dominance relationships. For any two parameter sets A and B, if A's electromagnetic performance (e.g., lower return loss, higher gain) is better than or equal to B's, and its thermal performance (e.g., lower temperature rise, smaller temperature gradient) is better than B's; or A's thermal performance is better than or equal to B's, and its electromagnetic performance is better than B's, then A is said to dominate B. Conversely, if two parameter groups each have their own strengths and weaknesses (e.g., A has superior electromagnetic performance but poor thermal performance, while B has superior thermal performance but poor electromagnetic performance), then they do not dominate each other. Multiple rounds of non-dominated sorting are performed on all parameter groups. The first round selects all non-dominated parameter groups that are not dominated by other parameter groups (i.e., parameters that cannot be surpassed by any group in both electromagnetic and thermal performance), as candidates for the Pareto front. The second round repeats the selection process for the remaining dominated parameter groups until all parameter groups are sorted. Only the non-dominated parameter groups selected in the first round are retained, arranged in order of electromagnetic performance from best to worst, and thermal performance from best to worst, forming the Pareto front solution set. Simultaneously, a visualization tool (such as Matplotlib) is used to plot the front curve, visually demonstrating the trade-off between electromagnetic and thermal performance (the closer the curve is to the upper right corner, the better the overall performance), and the specific values ​​and performance indicators of each parameter group are labeled, facilitating subsequent selection based on engineering objectives.

[0095] By building a parameterized simulation script based on the PyAnsys library, the entire process of electromagnetic-thermal coupling simulation of millimeter-wave radar array antennas can be automated, replacing traditional manual operations. This significantly reduces the manpower and time consumption in parameter modification, simulation startup, and data extraction, while avoiding errors caused by manual operation. Iterative calculations based on this script can generate multiple sets of parameters that meet the basic requirements of electromagnetic and thermal performance under preset constraints and termination conditions. Parallel computing improves iteration efficiency, ensuring that the parameter sets cover different performance dimensions and possess high quality, providing a rich sample for subsequent selection. Combining this with a non-dominated sorting method to sort multiple parameter sets can accurately eliminate inferior parameter sets whose performance is dominated, retaining non-dominated parameter sets whose electromagnetic and thermal performance cannot be mutually surpassed, forming a Pareto front solution set. This clearly presents the trade-off relationship between the two types of performance, allowing designers to quickly locate the optimal trade-off parameters based on engineering objectives without blindly screening from a massive number of parameters. This effectively supports efficient design decisions for millimeter-wave radar RF integrated microsystems, balancing design efficiency, parameter quality, and engineering practicality.

[0096] Furthermore, after establishing the Pareto front solution set based on the multiple sets of parameters, the method provided in this embodiment also includes:

[0097] (1) When the engineering objective changes, based on the new engineering requirements, the parameter sets in the Pareto front solution set are re-constrained and their performance is evaluated, and parameter sets that do not meet the new engineering requirements are removed.

[0098] Specifically, a change in engineering objectives refers to adjustments in performance priorities, constraints, or quantitative indicators (for example, the original engineering objective for vehicle-mounted millimeter-wave radar was to prioritize electromagnetic performance and tolerate a maximum temperature rise of ≤50℃; the new objective, due to the shift in application scenario to a high-temperature industrial environment, is to prioritize temperature rise control while allowing for more relaxed electromagnetic performance). In this case, it is necessary to first perform constraint verification and performance evaluation on all parameter sets in the original Pareto front solution set based on the new requirements.

[0099] The hard constraints in the new engineering requirements are used as screening criteria to check whether the corresponding performance of each parameter group in the original solution set meets the standards. For example, if a parameter group has a temperature rise of 42℃ under the original requirements (meeting the original constraint ≤50℃), but does not meet the new constraint ≤35℃, it is directly eliminated. If the new requirements add a structural constraint of substrate thickness ≤1mm, the parameter group with a substrate thickness of 1.2mm in the original solution set also needs to be eliminated. For parameter groups that pass the constraint verification, their comprehensive adaptability is re-evaluated according to the performance priority of the new requirements. For example, if the weight of thermal performance in the new requirements increases from 30% to 60% and the weight of electromagnetic performance decreases from 70% to 40%, the weighted performance score of each parameter group needs to be recalculated. Even if some parameter groups meet the constraints, if their core performance (such as temperature rise control) scores too low under the new weights, they should be included in the observation or low priority range to provide a screening basis for subsequent solution set updates.

[0100] (2) Perform iterative calculations for the new engineering requirements, generate new parameter sets, and update the Pareto front solution set.

[0101] Specifically, after screening, the original Pareto front solution set may only retain a small number of parameter sets that meet the new requirements, failing to cover the full dimensions of the electromagnetic-thermal performance tradeoff under the new scenario. Therefore, new iterative calculations need to be carried out specifically. The objective function system is adjusted and optimized based on the new engineering requirements, including updating performance quantification indicators, reallocating weights, and supplementing new constraint parameters to ensure that the direction of iterative calculations is consistent with the new requirements. The range of variable values ​​is optimized according to the new requirements. A parameterized simulation script based on the PyAnsys library is called to conduct multiple rounds of iterative calculations according to the new objectives, resulting in the updated Pareto front solution set.

[0102] By using targeted iterative calculations to fill the performance gaps of the original solution set under new requirements, the updated solution set can not only retain the original high-quality parameter resources, but also meet the full-dimensional design requirements of the new scenario, providing sufficient and suitable candidate samples for the subsequent selection of target parameter sets.

[0103] S106. Based on the engineering objectives, determine the target parameter set from the Pareto front solution set, and use the target parameter set as the running parameter set of the initial physical model.

[0104] Specifically, the project prioritizes engineering objectives and defines quantitative performance indicators based on application scenarios (automotive, aerospace, security, etc.). For example, in automotive scenarios, priority is given to ensuring electromagnetic performance stability, while in aerospace scenarios, priority is given to controlling temperature rise and weight. A scoring system is established, assigning weights to indicators such as electromagnetic performance, thermal performance, and structural cost according to priority, and weighting each parameter group in the Pareto front solution set. The top-ranked candidate parameter groups are selected, and their processing feasibility (geometric parameters meet processing limits), material availability (commonly available products), and cost controllability (below budget threshold) are verified. The parameter group with the highest weighted total score and no performance shortcomings is selected as the target parameter group. The final simulation verifies its performance and compliance with engineering objectives, and the output serves as the operating parameter group for the initial physical model.

[0105] Furthermore, the determination of the target parameter set must strictly align with actual engineering needs. Prioritizing engineering objectives is crucial, as different application scenarios have significantly different performance requirements. For example, security radar needs to balance cost and performance, while aerospace radar has extremely high requirements for reliability and thermal stability. The scoring system must ensure that the weight allocation is scientific and reasonable, accurately reflecting engineering priorities and avoiding deviations from actual needs due to improper weight settings. Feasibility verification of candidate parameter sets is key to engineering implementation, requiring full consideration of practical constraints such as processing technology, material supply, and cost budget, eliminating parameter sets that are "theoretically optimal but impractical." Finally, simulation verification ensures that the performance of the target parameter set fully meets engineering requirements, avoiding performance failures due to errors in the solution set selection process.

[0106] Through engineering goal-oriented screening and verification, a set of target parameters that combines performance advantages and engineering feasibility was determined from the Pareto front solution set. Compared to blindly selecting the theoretically optimal parameter set, this approach fully considers practical application scenarios, manufacturing capabilities, and cost control, ensuring that the output operating parameter set not only balances electromagnetic performance and thermal stability but can also be directly applied to engineering production. After assigning the target parameter set to the initial physical model, the model possesses operating characteristics that fully meet engineering requirements, providing a directly implementable design solution for the development of millimeter-wave radar RF integrated microsystems, significantly improving R&D efficiency and design success rate.

[0107] Furthermore, a set of target parameters is determined from the Pareto front solution set based on engineering objectives; including:

[0108] (1) Determine the priority of the project objectives based on the aforementioned project objectives;

[0109] Specifically, the priority of engineering objectives needs to be determined based on the actual application scenario of millimeter-wave radar to clarify the necessity and importance of different performance indicators in specific scenarios. If the engineering objective is the design of a vehicle-mounted millimeter-wave radar front-end, priority should be given to ensuring electromagnetic performance stability (radar needs to accurately detect targets, and signal transmission and radiation capabilities are core), followed by controlling thermal performance (vehicle environment temperature fluctuations are large, and high temperatures need to be avoided to prevent them from affecting device lifespan). The priority order is electromagnetic performance > thermal performance > cost control. If the engineering objective is the design of aerospace millimeter-wave radar, priority should be given to controlling heat dissipation and structural size (aerospace equipment has strict limitations on weight, volume, and power consumption), followed by ensuring electromagnetic performance (the low-interference environment in space has high requirements for signal stability). The priority order is thermal performance > structural parameters > electromagnetic performance.

[0110] (2) For each set of parameters in the Pareto front solution set, set the scoring weight according to the priority of the engineering objectives;

[0111] Specifically, weight intervals are divided according to priority, scoring rules are set for the sub-indicators of each performance dimension, and a weighted calculation of sub-item scores multiplied by weights is performed for each parameter group to obtain the weight score for each parameter group.

[0112] (3) Select multiple candidate parameter groups from the Pareto front solution set based on the weighted scores;

[0113] Specifically, all parameter groups are sorted in descending order of weight score, and the top-ranked parameter groups are selected as candidate parameter groups.

[0114] (4) Verify the multiple sets of candidate parameter groups and determine the target parameter group based on the verification results.

[0115] Specifically, check whether the geometric parameters of the candidate parameter group meet the limits of the existing processing equipment to determine the feasibility of the processing technology; estimate the production cost based on the material consumption and processing complexity of the candidate parameter group to determine the cost controllability, and take the candidate parameter group that meets the processing technology feasibility and cost controllability as the target parameter group.

[0116] By considering performance diversity during the screening phase and excluding critical parameter groups during the verification phase, the final target parameter group not only has excellent overall performance but also good environmental adaptability and stability, thus ensuring the reliable application of millimeter-wave radar RF integrated microsystems.

[0117] Furthermore, after using the target parameter set as the running parameter set of the initial physical model, the method provided in this embodiment includes:

[0118] (1) Perform engineering transformation on the target parameter set;

[0119] Specifically, the abstract dimensions in the simulation are transformed into two-dimensional engineering drawings that conform to processing standards, with dimensional tolerances and geometric tolerances marked, and processing process annotations added; the material properties in the simulation are transformed into material procurement technical specifications, specifying the commercial models, performance indicators, delivery cycles, and quality standards of the materials; for key processes such as power supply structures and metal plating, detailed process parameter tables are developed. For example, the welding process of the power supply port needs to specify the welding temperature, welding time, and solder type, while the metal radiation layer coating process needs to specify the copper layer thickness and coating uniformity deviation.

[0120] (2) Prepare a sample based on the transformed target parameter set, perform performance testing on the sample, and obtain the measured data of the sample;

[0121] Specifically, based on the engineering technical documents, millimeter-wave radar array antenna prototypes were fabricated using the same process route as mass production. Test environments were set up for electromagnetic and thermal performance. Electromagnetic performance testing employed a vector network analyzer and a microwave anechoic chamber to measure return loss, insertion loss, antenna gain, beamwidth, and other indicators. Thermal performance testing used a high and low temperature test chamber and an infrared thermal imager to measure the maximum temperature rise, temperature gradient, and heat distribution under different operating conditions. The tests were repeated multiple times according to the test standards, and the average value was taken as the final measured data.

[0122] (3) Compare the data deviation between the measured data and the simulation data of the target parameter group, and correct the initial physical model based on the data deviation.

[0123] Specifically, the data deviation is calculated based on the difference between measured and simulated data. This deviation is then compared to a preset threshold. When the deviation exceeds the threshold, the cause of the deviation is identified by examining the discrepancy between the simulation assumptions and actual operating conditions. The initial physical model is adjusted based on the attribution results. If the deviation is caused by machining burrs, burr structures are added to the model. If the deviation is caused by uneven heat transfer coefficients, a spatially distributed convective heat transfer coefficient is set in the Workbench thermal simulation. If the deviation is caused by changes in material thermal conductivity, the material parameters are corrected from fixed values ​​to temperature functions. After correction, the simulation is restarted until the deviation between the simulated and measured data is less than the preset threshold.

[0124] The multi-physics coupling simulation method for millimeter-wave radar RF integrated microsystems provided in this embodiment constructs an accurate initial physical model based on an array antenna, completes electromagnetic simulation using ANSYS HFSS to obtain electromagnetic performance and component power loss, and then imports the loss as a heat source into ANSYS Workbench to achieve thermal simulation, forming a deep electromagnetic-thermal coupling analysis. This breaks the limitations of single-physics simulation and ensures that the simulation results are highly consistent with actual operating conditions. By building automated scripts using the PyAnsys library and combining intelligent algorithms such as particle swarm optimization and genetic algorithms to conduct multiple rounds of iterative calculations, Pareto front solution sets covering different performance trade-offs are generated. This replaces the inefficient traditional method of manually adjusting parameters, significantly shortening the design iteration cycle. Through engineering goal-oriented priority setting, quantitative scoring, and feasibility verification, target parameter sets that combine performance advantages and engineering feasibility are selected from the solution set. This achieves an optimal balance between electromagnetic and thermal performance while fully adapting to manufacturing processes, material supply, and cost budget constraints. Subsequent engineering transformation, prototype testing, and model correction further improve design reliability. This provides a scalable and reusable technical path for high-density, high-frequency, and high-power RF integrated microsystems, with significant engineering practical value and broad application prospects.

[0125] Corresponding to the aforementioned embodiment of a multi-physics coupling simulation method for millimeter-wave radar RF integrated microsystems, this application also provides an embodiment of a multi-physics coupling simulation device for millimeter-wave radar RF integrated microsystems.

[0126] Figure 2 This is a schematic diagram of the structure of Embodiment 2 of the multi-physics coupling simulation device for millimeter-wave radar RF integrated microsystems provided in this application. Please refer to... Figure 2 The apparatus provided in this embodiment includes a construction module 210, a simulation module 220, and a calculation module 230; wherein,

[0127] The construction module 210 is used to construct an initial physical model based on the array antenna of the millimeter-wave radar front end;

[0128] The simulation module 220 is used to perform electromagnetic simulation on the initial physical model to obtain electromagnetic performance and component power loss.

[0129] The simulation module 220 is also used to use the power loss of the component as a heat source and perform thermal simulation on the initial physical model using the ANSYS Workbench platform to obtain the thermal performance of the initial physical model.

[0130] The calculation module 230 is used to construct a multi-objective optimization problem based on the electromagnetic properties and the thermal properties, optimize the multi-objective optimization problem based on the optimization algorithm, and obtain the parameter set of the initial physical model.

[0131] The calculation module 230 is also used to repeatedly call the Ansys simulation interface to complete multiple rounds of iterative calculations, obtain multiple sets of parameter sets, and establish Pareto front solution sets based on the multiple sets of parameter sets;

[0132] The calculation module 230 is further configured to determine a target parameter set from the Pareto front solution set based on the engineering objective, and use the target parameter set as the running parameter set of the initial physical model.

[0133] The apparatus of this embodiment can be used to perform... Figure 1 The steps of the method embodiment shown are similar in principle and process, and will not be repeated here.

[0134] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0135] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0136] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A multi-physics coupling simulation method for millimeter-wave radar RF integrated microsystems, characterized in that, The method includes: An initial physical model is constructed based on the array antenna at the front end of the millimeter-wave radar; Electromagnetic simulation was performed on the initial physical model to obtain the electromagnetic performance and component power loss. Using the power loss of the component as a heat source, thermal simulation of the initial physical model was performed using the ANSYS Workbench platform to obtain the thermal performance of the initial physical model. Based on the electromagnetic and thermal properties, a multi-objective optimization problem is constructed, and the multi-objective optimization problem is optimized based on the optimization algorithm to obtain the parameter set of the initial physical model; The Ansys simulation interface is called repeatedly to complete multiple rounds of iterative calculations, resulting in multiple sets of parameter sets. Based on these multiple sets of parameter sets, a Pareto front solution set is established. Based on the engineering objectives, a set of target parameters is determined from the Pareto front solution set, and the set of target parameters is used as the set of operating parameters for the initial physical model.

2. The method according to claim 1, characterized in that, The initial physical model is constructed based on the array antenna of the millimeter-wave radar front end; including: Determine the structural parameters of the array antenna; Geometric modeling is performed on the initial physical model based on the structural parameters and the target operating frequency band.

3. The method according to claim 1, characterized in that, The electromagnetic simulation of the initial physical model includes: The solution frequency range of the electromagnetic simulation is set based on the target operating frequency band; The simulation boundary conditions are determined based on the actual working scenario and structural characteristics of the array antenna. Configure the electromagnetic simulation excitation source according to the feeding topology of the array antenna; Electromagnetic simulation networks are divided based on the relationship between the region and electromagnetic field changes in the initial physical model, and electromagnetic simulation is performed.

4. The method according to claim 1, characterized in that, The step of using the power loss of the component as a heat source and performing thermal simulation of the initial physical model using the ANSYS Workbench platform includes: The initial physical model is imported into the thermal simulation module using the ANSYS Workbench import interface. The component power loss output from the electromagnetic simulation is then called, and spatial mapping is performed based on the component power loss. Define boundary conditions based on the actual heat dissipation scenario of the array antenna during operation; Determine the type of thermal analysis based on the simulation requirements, and submit a thermal simulation task to perform the thermal simulation analysis.

5. The method according to claim 1, characterized in that, The process involves constructing a multi-objective optimization problem based on the electromagnetic and thermal properties, optimizing the multi-objective optimization problem using an optimization algorithm, and obtaining the parameter set of the initial physical model; including: An objective function system is constructed based on the evaluation indices of the electromagnetic properties and the thermal properties. Optimization variables are selected from the parameters of the initial physical model; Select the target optimization algorithm based on the complexity of the optimization problem and the convergence requirements; The optimization calculation is performed based on the target optimization algorithm, and the parameter set is output.

6. The method according to claim 1, characterized in that, The iterative calculations, performed by repeatedly calling the Ansys simulation interface, yield multiple sets of parameters. A Pareto front solution set is then established based on these parameter sets. This includes: Build a parameterized simulation call script based on the PyAnsys library; Based on the simulation script, iterative calculations are performed to obtain multiple sets of parameters. The multiple parameter sets are sorted using the non-dominated sorting method to obtain the Pareto front solution set.

7. The method according to claim 1, characterized in that, The determination of the target parameter set from the Pareto front solution set based on engineering objectives includes: The priority of the project objectives is determined based on the aforementioned project objectives; For each set of parameters in the Pareto front solution set, a scoring weight is set according to the priority of the engineering objectives; Multiple candidate parameter groups are selected from the Pareto front solution set based on weighted scores; The multiple candidate parameter sets are verified, and the target parameter set is determined based on the verification results.

8. The method according to claim 1, characterized in that, After using the target parameter set as the running parameter set of the initial physical model, it includes: The target parameter set is then transformed into an engineering form. Samples were prepared based on the transformed target parameter set, and the performance of the samples was tested to obtain the measured data of the samples. The initial physical model is corrected based on the data deviation between the measured data and the simulation data of the target parameter set.

9. The method according to claim 1, characterized in that, After establishing the Pareto front solution set based on the multiple sets of parameters, it includes: When the engineering objective changes, the parameter sets in the Pareto front solution set are re-constrained and their performance is evaluated based on the new engineering requirements, and parameter sets that do not meet the new engineering requirements are removed. Perform iterative calculations to meet the new engineering requirements, generate new parameter sets, and update the Pareto front solution set.

10. A multi-physics coupling simulation device for millimeter-wave radar radio frequency integrated microsystems, characterized in that, The device includes a construction module, a simulation module, and a calculation module; wherein... The construction module is used to construct an initial physical model based on the array antenna of the millimeter-wave radar front end; The simulation module is used to perform electromagnetic simulation on the initial physical model to obtain electromagnetic performance and component power loss. The simulation module is also used to use the power loss of the component as a heat source and perform thermal simulation on the initial physical model using the ANSYS Workbench platform to obtain the thermal performance of the initial physical model. The calculation module is used to construct a multi-objective optimization problem based on the electromagnetic properties and the thermal properties, optimize the multi-objective optimization problem based on the optimization algorithm, and obtain the parameter set of the initial physical model. The calculation module is also used to repeatedly call the Ansys simulation interface to complete multiple rounds of iterative calculations, obtain multiple sets of parameters, and establish Pareto front solution sets based on the multiple sets of parameters; The calculation module is also used to determine a target parameter set from the Pareto front solution set based on the engineering objective, and use the target parameter set as the running parameter set of the initial physical model.

Images