Interactive modeling and optimization platform and method for electromagnetic devices

Abstract

The present invention discloses an interactive modeling and optimization platform and method for electromagnetic devices, comprising a view layer, a control layer, and a model layer. The platform and method can display options corresponding to modeling and optimization algorithms through a visual interactive interface, obtain the modeling and optimization algorithms selected by the user, receive observation variables and targets set by the user, execute parameterized modeling and optimization processes based on the modeling and optimization algorithms selected by the user and the set observation variables and targets, and display the execution results of the parameterized modeling and optimization processes in the visual interactive interface. The present invention can provide users with a visual interactive interface, through which users can select the modeling and optimization algorithms used in the modeling and optimization processes, thereby reducing repetitive programming and algorithm debugging work and improving the efficiency of electromagnetic device modeling and optimization. The present invention has broad application in the field of electromagnetic device modeling and optimization technology.

Description

Technical Field

[0001] The present invention relates to the technical field of electromagnetic device modeling and optimization, and in particular to an interactive modeling and optimization platform and method for electromagnetic devices. Background Art

[0002] Electromagnetic components such as filters, power splitters, duplexers, couplers, and antennas are widely used in wireless communication systems and other electronic systems. The performance of these components directly impacts the quality of the entire communication system, necessitating the use of electromagnetic simulation to optimize their performance. As the demand for electromagnetic components in wireless communication systems increases, the structure and scale of electromagnetic simulations are becoming increasingly complex, making the modeling and optimization of these components extremely time-consuming. Currently, manual modeling and optimization processes are inefficient, making it difficult to achieve device modeling and optimization in a short period of time.

[0003] The use of optimization algorithms can effectively accelerate the modeling and optimization process of electromagnetic devices. However, when faced with different device structures or optimization targets, dedicated algorithms need to be implemented according to design requirements. This leads to problems such as repeated programming and time-consuming algorithm debugging, which increases the time and human resources required for device optimization. For example, when optimizing the performance curves or performance indicators of specific devices, multidisciplinary optimization software such as OptiSLang only supports optimization of the S-parameter data of electromagnetic devices and only supports modeling of the performance indicators of electromagnetic devices. Due to the lack of modeling of electromagnetic device performance curves and optimization of directional pattern data, it is unable to provide more direct and rich information. On the other hand, compared with customized electromagnetic device-specific algorithms, traditional intelligent optimization algorithms also have disadvantages such as low modeling and optimization efficiency, which further limits the scope of use and development of such electromagnetic device optimization software. Summary of the Invention

[0004] In response to the current technical problems of low efficiency in electromagnetic device modeling and optimization and the lack of a software platform with both dedicated and general algorithms, the purpose of the present invention is to provide an interactive modeling and optimization platform and method for electromagnetic devices.

[0005] In one aspect, an embodiment of the present invention includes an interactive modeling and optimization platform for electromagnetic devices, comprising:

[0006] View layer; the view layer is used to display a visual interactive interface, the visual interactive interface includes at least one sub-interface;

[0007] Control layer; the control layer is used to store data such as performance curves and performance indicators corresponding to various electromagnetic devices, perform interactive parameter settings and automatic error detection functions, display options corresponding to the modeling algorithm and optimization algorithm through the visual interactive interface, obtain the modeling algorithm and optimization algorithm selected by the user through the visual interactive interface, receive observation variables and targets set by the user, display the execution results of the parameterized modeling process and the optimization process in the visual interactive interface, and import and update existing data files;

[0008] Model layer; the model layer is used to execute the modeling algorithm and the optimization algorithm, and execute the parameterized modeling process and the optimization process according to the modeling algorithm and the optimization algorithm selected by the user and the set observation variables and objectives.

[0009] Furthermore, the at least one sub-interface includes an algorithm setting sub-interface, a path setting sub-interface, a parameter setting sub-interface, a simulation setting sub-interface, a target setting sub-interface, a debugging setting sub-interface, an automatic phase extraction sub-interface, an automatic expression generation sub-interface, and a proxy model sub-interface.

[0010] Furthermore, the interactive parameter setting and automatic error detection functions are performed, including:

[0011] A parameter setting interactive interface is displayed through the visual interactive interface, wherein the parameter setting interactive interface is used for interactive input of various operating parameters, wherein the operating parameters include algorithm parameters, path parameters, design variable parameters, electromagnetic simulation software parameters, custom observation variables and target parameters, and debugging parameters;

[0012] Performing error detection on information input by the user through the visual interactive interface; the error detection includes character string validity detection, data type detection and value range detection;

[0013] Capture error information generated during the operation of the view layer, the control layer, and the model layer.

[0014] Furthermore, the electromagnetic device includes a filter, a power splitter, a duplexer, a coupler, an antenna unit and an array antenna;

[0015] The performance curves include an S-parameter curve, a directivity pattern curve, an amplitude curve and a phase curve;

[0016] The performance indicators include return loss, insertion loss, antenna gain and sidelobe level.

[0017] Furthermore, the modeling algorithm and optimization algorithm include:

[0018] A Kriging proxy model is constructed based on different regression models and correlation models, and the output of the proxy model is a performance curve or performance index of the electromagnetic device;

[0019] Running a dedicated optimization algorithm or a universal optimization algorithm on the Kriging proxy model; the dedicated optimization algorithm includes a VF algorithm applicable to filters, power splitters, duplexers, and couplers, as well as a VF algorithm, FFT algorithm, and VF-FFT algorithm applicable to antenna units and array antennas; the universal optimization algorithm includes an NSGA-II-I algorithm, an NSGA-II-D algorithm, an NSDE-I algorithm, an NSDE-D algorithm, an OCDE-I algorithm, and an OCDE-D algorithm applicable to all electromagnetic devices;

[0020] The optimization algorithm includes a search algorithm running on the Kriging agent model, and the search algorithm includes a GA / NSGA-II algorithm, an NSDE algorithm, and an OCDE algorithm;

[0021] The optimization algorithm includes classification and sorting algorithms for design variable combinations. During the operation of the platform optimization program, different classification and sorting algorithms are allowed to be switched according to actual conditions.

[0022] Furthermore, the visual interactive interface can be used as a numerical interface for displaying numerical values, a picture interface for displaying simulation curves, a console interface for displaying optimization progress, or a parametric modeling interface.

[0023] Furthermore, the parametric modeling process includes:

[0024] Based on electromagnetic simulation data and a user-selected algorithm, a Kriging proxy model is constructed to predict user-set observation variables and target loss function values, and the observation variables and target loss function values are displayed in a visual interactive interface of the proxy model sub-interface; wherein the observation variables correspond to the performance curve, and the target loss function values correspond to the performance indicators;

[0025] Update the parametric modeling interface of design variables and update the output of the proxy model corresponding to the design variables in real time;

[0026] The optimization process includes:

[0027] After the user completes the platform parameter settings, the cyclic processing process is executed, and each cyclic processing process generates a generation of data files;

[0028] The cyclic processing process includes:

[0029] When the executed loop processing process is the first round of loop processing, an initial design variable combination is generated according to the algorithm selected by the user; when the executed loop processing process is not the first round of loop processing, the parametric modeling process is automatically executed according to the existing simulation results and the algorithm selected by the user, and the search algorithm selected by the user searches around the search center on the proxy model to generate a current design variable combination;

[0030] Automatically generate vbs files with relevant settings and call electromagnetic simulation software HFSS for simulation;

[0031] After the simulation is finished, the simulation results are automatically read to obtain the values of the observation variables set by the user;

[0032] Automatically calculate the loss function value of the user-set target;

[0033] Automatically classify and rank design variable combinations;

[0034] Automatically update the platform's visual interactive interface;

[0035] When the loop end condition is met, the execution of all the loop processing processes is ended.

[0036] Furthermore, the loop end condition includes:

[0037] The optimization result of the last round of cyclic processing meets the goal set by the user;

[0038] or

[0039] The cumulative number of rounds of the cyclic processing process reaches the total number of generations set by the user.

[0040] Furthermore, the function of importing and updating existing data files includes:

[0041] Import global data files of different projects;

[0042] Import the data file of any round of the cyclic processing process in any project;

[0043] Based on the imported data files and combined with the debugging setting function provided by the platform, the data files can be updated;

[0044] Continue running based on the optimization results of any round of the cyclic processing process in any project.

[0045] On the other hand, an embodiment of the present invention further includes an interactive modeling and optimization method for an electromagnetic device, the interactive modeling and optimization method for an electromagnetic device comprising:

[0046] Displaying a visual interactive interface, wherein the visual interactive interface includes at least one sub-interface;

[0047] Store the performance curves and performance indicators corresponding to various electromagnetic devices;

[0048] Perform interactive parameter setting and automatic error detection functions;

[0049] Displaying options corresponding to the modeling algorithm and the optimization algorithm through the visual interactive interface;

[0050] Obtaining the modeling algorithm and optimization algorithm selected by the user through the visual interactive interface, and receiving the observation variables and targets set by the user;

[0051] Executing a modeling algorithm and an optimization algorithm, and executing a parameterized modeling process and a loop processing process according to the modeling algorithm and the optimization algorithm selected by the user and the set observation variables and objectives;

[0052] The execution results of the parameterized modeling process and the optimization process are displayed in the visual interactive interface.

[0053] The beneficial effects of the present invention are as follows: the interactive electromagnetic device automatic modeling and optimization platform in the embodiment can automatically implement a series of processes such as electromagnetic device modeling, optimization, simulation, and data reading for the performance curves and performance indicators of various electromagnetic devices, and provide a visual interactive interface to the user, so that the user can select the modeling algorithm and optimization algorithm used in the modeling and optimization process through the visual interactive interface, and view the modeling and optimization results through the visual interactive interface, thereby reducing the repetitive programming and algorithm debugging work that the user needs to perform in the modeling and optimization process, lowering the user's usage threshold, and improving the modeling and optimization efficiency of electromagnetic devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic diagram of the structure and logic of an interactive modeling and optimization platform for electromagnetic devices in an embodiment;

[0055] Figure 2 This is a schematic diagram of the main interface of the platform interface in the embodiment;

[0056] Figure 3 、 Figure 4 、 Figure 5 A schematic diagram of a visual interactive interface within the platform interface in the embodiment;

[0057] Figure 6 、 Figure 7 This is an example of a pop-up window display for checking the validity of platform parameter settings and capturing error information in the embodiment;

[0058] Figure 8 、 Figure 9 This is a schematic diagram of the algorithm setting sub-interface of the platform in the embodiment;

[0059] Figure 10 This is a schematic diagram of the proxy model sub-interface of the platform in the embodiment;

[0060] Figure 11 is a flow chart of the optimization process described in the embodiment;

[0061] Figure 12 、 Figure 13 This is a schematic diagram of the target setting sub-interface of the platform in the embodiment;

[0062] Figure 14 、 Figure 15 This is an example of a shortcut key for importing data files into the platform in the embodiment;

[0063] Figure 16 This is a schematic diagram of the debugging setting sub-interface of the platform in the embodiment;

[0064] Figure 17 A schematic structural diagram of a filtering antenna device used in the platform of the embodiment;

[0065] Figure 18 is an initial simulation curve of the filtering antenna device applied to the platform in the embodiment;

[0066] Figure 19 Schematic diagram of the numerical values of the optimal design variables of the filtering antenna device applied to the platform in the embodiment. DETAILED DESCRIPTION

[0067] In this embodiment, refer to Figure 1 The interactive modeling and optimization platform for electromagnetic devices (hereinafter referred to as the "platform") includes a view layer, a control layer and a model layer. Part of the platform's functions are respectively executed through the view layer, the control layer and the model layer, thereby realizing the overall function of the platform.

[0068] In this embodiment, "electromagnetic device" may specifically refer to a filter, a power splitter, a duplexer, a coupler, an antenna unit, or an array antenna, or may be a general term for multiple such devices. "Performance curve" may specifically refer to an S-parameter curve, a directional pattern curve, an amplitude curve, or a phase curve, or may be a general term for multiple such curves. "Performance indicator" may specifically refer to a return loss, insertion loss, antenna gain, or sidelobe level, or may be a general term for multiple such indicators.

[0069] In this embodiment, the view layer is used to display a visual interactive interface, which includes at least one sub-interface. In the view layer, an interactive interface can be provided, and the modeling and optimization results of the visualization platform can be visualized.

[0070] In this embodiment, the control layer is used to store data such as performance curves and performance indicators corresponding to various electromagnetic devices, perform interactive parameter settings and automatic error detection, display modeling and optimization algorithm options through a visual interactive interface, obtain user-selected modeling and optimization algorithms through the visual interactive interface, receive user-defined observation variables and objectives, display the results of the parameterized modeling and optimization processes in the visual interactive interface, and import and update existing data files. The control layer can implement interactive logic and data processing, receive parameter settings from the interactive interface, and import existing settings.

[0071] In this embodiment, the model layer is used to execute modeling and optimization algorithms, executing parameterized modeling and optimization processes based on the user-selected modeling and optimization algorithms and the set observation variables and objectives. This layer enables modeling and optimization, simulation, and result reading. Common electromagnetic simulation software operations can be written as corresponding API functions and called to execute modeling and optimization algorithms. Without loss of generality, multiple sub-interfaces can be constructed using Matlab or Python, based on the platform's basic logic, to facilitate interactive parameter setting.

[0072] Reference Figure 2 , which is the main interface of the platform interface in this embodiment. The left side of the main interface can provide multiple sub-interfaces. Each sub-interface can have multiple settings pages, allowing users to configure settings based on different structures. The right side of the main interface can use an image display box to display the convergence curve of the optimization results, or use a text box to display the optimization progress and results. The interface is updated when a round of the cycle process is completed.

[0073] Reference Figure 3 、 Figure 4 、 Figure 5 , is an example of a visual interactive interface of the platform interface in the embodiment, which can be further divided into a workspace page and a console page. In the workspace page, you can use the value page interface and the image page interface to display the numerical value and simulation result curves respectively. At the same time, interactive interfaces such as drop-down boxes and list boxes can be provided. When the user selects the corresponding data object, the platform automatically generates the corresponding data table or image and displays it. In the console page, you can use a text box to display the optimization progress during the program execution.

[0074] In this embodiment, the platform's interactive parameter setting function can be implemented through interactive interfaces such as buttons, text boxes, drop-down boxes, tables, check boxes, list boxes, and list trees in the platform interface and its various sub-interfaces. Without loss of generality, parameters related to modeling and optimization algorithms can be set in the algorithm setting sub-interface, parameters related to file paths can be set in the path setting sub-interface, parameters related to design variables of electromagnetic devices can be set in the variable setting sub-interface, parameters related to electromagnetic simulation software can be set in the simulation setting sub-interface, parameters related to optimization targets can be set in the target setting sub-interface, parameters related to debugging and updating programs can be set in the debugging setting sub-interface, and parameters related to modeling can be set in the proxy model sub-interface. The results of parametric modeling can also be displayed.

[0075] In this embodiment, the platform's automatic error detection function can be implemented based on the validity detection performed after each sub-interface is set up, and during the program execution, the error information of the program console is captured and displayed in a pop-up window. Figure 6 、 Figure 7 By detecting the data type and value range of the text box, such as floating-point numbers, positive integers, strings, and the size relationship between values, if the data type filled in by the user is inconsistent with the data type required by the platform, or the value range is wrong, or there is any error during the operation of the platform program, a pop-up window will display the corresponding information to facilitate the user to detect and correct it.

[0076] In this embodiment, the visual interactive interface can function as a numerical interface for displaying numerical values, a graphical interface for displaying simulation curves, a console interface for displaying optimization progress, or a parametric modeling interface. The sub-interfaces of the visual interactive interface include an algorithm setting sub-interface, a path setting sub-interface, a parameter setting sub-interface, a simulation setting sub-interface, a target setting sub-interface, a debugging setting sub-interface, an automatic phase extraction sub-interface, an automatic expression generation sub-interface, and a proxy model sub-interface. Different sub-interfaces of the visual interactive interface have different functions.

[0077] For example, for the algorithm setting sub-interface, the platform provides a variety of optional and switchable parametric modeling algorithms and optimization algorithms, which can be implemented based on the parameter settings in the algorithm setting sub-interface. Figure 8 and Figure 9The algorithm settings sub-interface includes a Select page for selecting an algorithm and a Search page for setting proxy model parameters and search algorithm parameters. The platform integrates VF algorithms for multi-port devices such as filters, VF algorithms, FFT algorithms, and VF-FFT algorithms for antennas and array antennas, as well as general algorithms such as NSGA-II-I, NSGA-II-D, NSDE-I, NSDE-D, OCDE-I, and OCDE-D for all electromagnetic devices. It also integrates a variety of Kriging proxy model algorithms, allowing the construction of Kriging proxy models based on various regression and correlation models. It also includes search algorithms for these proxy models, including GA / NSGA-II, NSDE, and OCDE. Furthermore, it includes classification and ranking algorithms for design variable combinations, allowing users to switch between different classification and ranking algorithms based on actual conditions during the platform optimization process.

[0078] In this embodiment, the platform performs modeling and optimization based on the performance curves of electromagnetic devices, which can be achieved by selecting a VF algorithm applicable to S-parameter curves, an FFT algorithm applicable to directional pattern curves, a VF-FFT algorithm applicable to S-parameter and directional pattern curves, and a general algorithm "-D" type algorithm (such as NSGA-II-D, NSDE-D, and OCDE-D algorithms).

[0079] In this embodiment, the platform performs modeling and optimization based on the performance indicators of electromagnetic devices, which can be achieved by selecting a "-I" type algorithm of a general algorithm (such as NSGA-II-I, NSDE-I, and OCDE-I algorithms).

[0080] For example, for the proxy model sub-interface, refer to Figure 10 The platform performs parametric modeling and can provide a visual interactive interface for modeling results through the proxy model sub-interface. It can dynamically generate tables based on user parameter settings and allow users to edit them, such as changing the value of the design variable by changing the position of the slider or directly changing the value in the table. By clicking the setting button on the left and selecting the object, the platform will automatically draw the parametric modeling results, such as the performance curve of the electromagnetic device (S parameters, directional diagram), and predict the corresponding target loss function value and feedback in the table.

[0081] In this embodiment, when the control layer performs the interactive parameter setting and automatic error detection functions, it may specifically perform the following steps:

[0082] A1. Display the parameter setting interface through a visual interactive interface. The parameter setting interface is used for interactive input of various operating parameters. Users can use the parameter setting interface to input operating parameters such as algorithm parameters, path parameters, design variable parameters, electromagnetic simulation software parameters, custom observation variables, target parameters, and debugging parameters.

[0083] A2. Perform error detection on information entered by the user through the visual interactive interface; specifically, error detection includes string validity detection, data type detection, and value range detection;

[0084] A3. Capture error messages generated during the operation of the view layer, control layer, and model layer.

[0085] The control layer can provide users with customized operation parameter settings by performing interactive parameter settings and automatic error detection functions, proactively discover errors in user input information, and error information generated by the platform during operation, thereby improving the success rate of modeling and optimization.

[0086] In this embodiment, when executing the modeling algorithm and the optimization algorithm, the model layer performs the parameterized modeling process, which specifically includes the following steps:

[0087] B1. Construct a Kriging proxy model based on the user-defined regression model and correlation model. The proxy model outputs the performance curve or performance index of the electromagnetic device.

[0088] B2. Predict the observed variables and target loss function values set by the user based on the surrogate model;

[0089] B3. Displaying the observed variable and the target loss function value in a visual interactive interface of the proxy model sub-interface; wherein the observed variable corresponds to the performance curve, and the target loss function value corresponds to the performance index;

[0090] B4. Update the parametric modeling interface of the design variables and update the output of the proxy model corresponding to the design variables in real time.

[0091] In this embodiment, when the model layer performs the parameterized modeling process, the Kriging proxy model can be constructed in the following manner:

[0092] The output of the Kriging model is divided into a deterministic trend term and an autocorrelated error term, which are calculated by the regression model and the correlation model respectively. Therefore, the output of the surrogate model has the following form:

[0093] y m (x)=F(θ;m;x)+α(θ;m;x)

[0094] Where F(θ; m; x) is the deterministic trend term of the mth output value, and the estimation of θ can be obtained using the regression model Then, predict the regression values of other unknown design variables x. Figure 9 The platform provides three regression models: constant, linear, and cubic. The default is the linear model. In addition, α(θ; m; x) is the autocorrelation error term of the m-th output value. The correlation model can be used to obtain the estimate of θ. Then, predict the autocorrelated random error values of other unknown design variables x. Figure 9 The platform provides five related models: absolute_exponential, squared_exponential, generalized_exponential, cubic, and linear. The default model is squared_exponential. In this embodiment, the Kriging model is implemented based on the scikit-learn library.

[0095] In this embodiment, based on the above-mentioned Kriging proxy model, the output of the platform parameterized model has the following form:

[0096] Y=m(X)

[0097] Where m is the mapping function of the Kriging proxy model, the row vectors of matrix X correspond to a set of design variables x, and the row vectors of matrix Y are the predicted output vectors y corresponding to the row vectors of matrix X. Specifically, in "-D" algorithms, y corresponds to the numerical value of the discrete points of the performance curve; in "-I" algorithms, y corresponds to the numerical value of the performance index; in VF algorithms, y corresponds to the numerical value of the coupling matrix coefficients of the S parameters and other discrete points of the performance curve that do not use the VF algorithm; in FFT algorithms, y corresponds to the FFT coefficients of the directional pattern and other discrete points of the performance curve that do not use the FFT algorithm; in VF-FFT algorithms, y corresponds to the numerical value of the coupling matrix coefficients of the S parameters, the FFT coefficients of the directional pattern, and other discrete points of the performance curve that do not use the VF-FFT algorithm.

[0098] In this embodiment, the platform executes the modeling algorithm and the optimization algorithm. By clicking the "Run Program" button on the platform interface, the parameterized modeling process and the optimization process are executed to achieve the optimization of the electromagnetic device. Figure 11 , the optimization process includes steps S1-S8, further:

[0099] In step S1, after the user completes the platform parameter settings, the loop processing steps S2-S7 are executed, and each loop processing step generates a generation of data files;

[0100] In step S2, when the executed loop processing process is the first round of loop processing, a design variable combination is generated according to the initialization algorithm selected by the user; when the executed loop processing process is not the first round of loop processing, the parametric modeling process is automatically executed based on the existing simulation results and the algorithm selected by the user, and the search algorithm selected by the user searches around the search center on the proxy model to generate a contemporary design variable combination;

[0101] In step S3, a vbs file with relevant settings is automatically generated and the electromagnetic simulation software HFSS is called for simulation;

[0102] In step S4, after the simulation is completed, the simulation results are automatically read to obtain the values of the observation variables set by the user;

[0103] In step S5, the loss function value of the target set by the user is automatically calculated;

[0104] In step S6, the design variable combinations are automatically classified and sorted;

[0105] In step S7, the visual interactive interface of the platform is automatically updated;

[0106] After executing steps S2-S7 in the i-th round of loop processing, a check is performed to determine whether the loop termination condition is met. In this embodiment, the loop termination condition may specifically be "the optimization result of the last round of loop processing meets the user-set target" or "the cumulative number of loop processing rounds reaches the user-set total number of generations." If it is detected that "the optimization result of the last round of loop processing does not meet the user-set target" and "the cumulative number of loop processing rounds does not reach the user-set total number of generations," then the counter is incremented to i=i+1, and the next loop processing round is executed.

[0107] In step S8, the loop processing process ends, completing the optimization process. Specifically, the optimization results can be displayed in the platform's visual interactive interface. The user can optimize the electromagnetic device based on the design variables returned in step S8. Because the design variables returned in step S8 are those optimized through simulation, the user does not need to perform additional optimization on the design variables returned in step S8, thereby eliminating the need for manual optimization work and improving the efficiency of electromagnetic device optimization.

[0108] Specifically, in step S1, the user needs to complete the platform parameter settings, including algorithm parameter settings, path parameter settings, design variable parameter settings, electromagnetic simulation software parameter settings, custom observation variables and target parameter settings, debugging parameter settings, etc.

[0109] Specifically, refer to Figure 6 and Figure 8The algorithms involved in steps S2 and S6 can be implemented through parameter settings in the algorithm setting sub-interface.

[0110] Specifically, in step S3, to automatically generate a vbs file, the platform has compiled common operations of the electromagnetic simulation software into corresponding API functions. For example, calling the Analysis, Optimetrics, Report, and Radiation items of the electromagnetic simulation software HFSS for parameter setting can be implemented using the self-written API functions HfssInsertSetup, HfssInsertOptim, HfssInsertReport, and HfssInsertRadiation, respectively.

[0111] Specifically, refer to Figure 12 and Figure 13 ,Setting the observation variables and targets in steps S4 and S5 can be ,achieved through parameter settings in the target setting sub-interface.

[0112] In this embodiment, refer to Figure 14 , the platform imports existing data files, which can be achieved by importing the global data file of the project through the File-Import shortcut key in the menu bar above the platform interface, or by importing the data file generated during any round of the cycle processing in any project. Figure 15 In the proxy model sub-interface, you can use the File-Import shortcut key in the menu bar above the sub-interface to import the data files generated during any round of the loop processing in any project, so that users can view the parametric modeling results.

[0113] In this embodiment, the platform updates the existing data files, or continues to run based on the optimization results of any round of the cyclic processing process in any project, which can be achieved through the parameter settings in the debugging setting sub-interface. Figure 16 After importing the corresponding data file, select the "First Run", "Continue Running", "Data Migration", "Update Target Expression", "Update Target Value" and other functions provided by the platform to update the data file or continue running the program.

[0114] In this embodiment, the platform is applied to the modeling and optimization of a stacked filter antenna, whose structure is as follows: Figure 17 As shown. Among them, the dielectric substrate parameters are: relative dielectric constant ε r =3.66, damage angle tanδ = 0.004. The initial optimization parameters are x = [H1, H2, L1, W1, L2, W2] = [0.101, 0.762, 3.8, 3.4, 3.8, 2.5], all in mm. The simulation curve is as follows Figure 18As shown. According to the optimization requirements, the design index is defined as:

[0115] F1:|S 11 |≤-10dB,18.3GHz≤f≤19.7GHz

[0116] F2:max(RealizedGain)≥6.5dBi, f0=19GHz, Phi=0deg

[0117] F3:max(RealizedGain)≥6.5dBi, f0=19GHz, Phi=90deg

[0118] It can be seen that the device needs to optimize the S parameters and the actual gain of the E and H planes at the same time, so the VF-FFT algorithm in the platform can be used, that is, the VF algorithm is used to optimize the S parameters of the antenna and the FFT algorithm is used to optimize the actual gain of the antenna. Some algorithm parameter settings in the platform are as follows Figure 8 、 Figure 9 shown.

[0119] To automatically read the required simulation data, refer to Figure 12 , you can customize three observation variables S11, RGT0 and RGT90, which represent the |S in 18.3-19.7GHz respectively. 11 Parameters, actual gain of xoz surface, actual gain of yoz surface. Referring to FIG13 , three target expressions S11<=-10, max(RGT0)>=6.5, and max(RGT90)>=6.5 can be defined based on the design indicators.

[0120] After completing the other parameter settings in the platform, set the run time to 15 generations in the main interface, click the "Run Program" button in the main interface, and the platform will automatically execute the parameterized modeling and optimization process. It is a very small amount, so in the convergence curve of the platform feedback, the vertical axis uses a logarithmic coordinate axis, and the conversion relationship is To avoid infinite terms, the minimum value of the loss function is cut off at 1e-5. After the 12th round of the loop processing process, the platform has completed the set optimization target, and the error function values of all targets have dropped to within 1e-5. Select the "Continue Run" function in the debug settings sub-interface and set the total number of generations to 15. You can continue to run to 15 generations. The screenshot of the interface at the end of the final run is as follows Figure 2 As shown, some console information is as follows Figure 5As shown in the figure, after 15 rounds of the cyclic processing, the platform found 6 optimal solutions, namely the second solution of the 12th generation, the second and third solutions of the 13th generation, the third and fourth solutions of the 14th generation, and the second solution of the 15th generation.

[0121] Furthermore, you can view the simulation curve in the visual interactive interface within the platform interface. Figure 3 Select the database GBest, the data source Epoch15 / 2, and the report names dB and RGTotal. Click "Confirm" to generate data objects such as Epoch15 / 2-dB-1 and Epoch15 / 2-RGTotal-1. Select the first two data objects again and click "Confirm" below them to view the simulation curve in the image display. As you can see, the simulation curve meets the three set objectives.

[0122] Furthermore, you can view the values of custom observation variables and targets in the visual interactive interface within the platform interface. Figure 4 After selecting the data source, click the "Confirm" button, and the platform will automatically generate the corresponding data table. As you can see, the data size of the three observed variables is automatically listed in the table, and they are distinguished by the text in the table header. Double-clicking a cell can expand the secondary data to display it, that is, to view the specific value. Figure 19 By checking the x of the 15th generation GBest, you can obtain the values of the global optimal design variables and distinguish the design variable names through the text in the table header.

[0123] Furthermore, during the program running, you can click the "Proxy Model" button on the main interface to start the proxy model sub-interface to view the modeling results. Figure 10 , select the data objects in the VF-Report and FFT-Report to view the prediction curve of the Kriging proxy model in the image display box. By changing the slider position of each design variable in the dynamic table or directly modifying the value of the design variable in the table, the prediction curve of the proxy model can be updated in real time, and the loss function value of the prediction target is fed back to the table in the lower left corner.

[0124] In this embodiment, the interactive modeling and optimization method of electromagnetic devices can be executed by running the interactive modeling and optimization platform of electromagnetic devices. The interactive modeling and optimization method of electromagnetic devices includes the following steps:

[0125] P1 displays a visual interactive interface, the visual interactive interface includes at least one sub-interface;

[0126] P2. Stores data such as performance curves and performance indicators corresponding to various electromagnetic devices;

[0127] P3. Perform interactive parameter setting and automatic error detection functions;

[0128] P4. Display the corresponding options of modeling algorithms and optimization algorithms through a visual interactive interface;

[0129] P5. Obtain the modeling and optimization algorithms selected by the user through a visual interactive interface, and receive the observation variables and objectives set by the user;

[0130] P6. Execute the parametric modeling and optimization processes based on the modeling and optimization algorithms selected by the user and the set observation variables and objectives;

[0131] P7. Display the execution results of the parametric modeling process and the optimization process in a visual interactive interface.

[0132] Step P1 of the interactive modeling and optimization method for electromagnetic devices can be performed by the view layer of the interactive modeling and optimization platform for electromagnetic devices, steps P2-P5 and P7 can be performed by the control layer, and step P6 can be performed by the model layer. By executing steps P1-P7, the same technical effects as those of the interactive modeling and optimization platform for electromagnetic devices in this embodiment can be achieved.

[0133] A computer program that executes the interactive modeling and optimization method of the electromagnetic device in this embodiment can be written and written into a computer device or storage medium. When the computer program is read out and run, the interactive modeling and optimization method of the electromagnetic device in this embodiment is executed, thereby achieving the same technical effect as the interactive modeling and optimization method of the electromagnetic device in the embodiment.

[0134] It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly fixed or connected to the other feature, or it may be indirectly fixed or connected to the other feature. In addition, the descriptions of up, down, left, right, etc. used in this disclosure are only relative to the relative positional relationship of the components of the present disclosure in the accompanying drawings. The singular forms of "a", "said" and "the" used in this disclosure are also intended to include the plural forms, unless the context clearly indicates otherwise. In addition, unless otherwise defined, all technical and scientific terms used in this embodiment have the same meaning as those generally understood by those skilled in the art. The terms used in the description of this embodiment are only for describing specific embodiments and are not intended to limit the present invention. The term "and / or" used in this embodiment includes any combination of one or more related listed items.

[0135] It should be understood that, although the present disclosure may adopt the term first, second, third etc. to describe various elements, these elements should not be limited to these terms.These terms are only used to distinguish the elements of the same type from each other.For example, without departing from the scope of the present disclosure, the first element may also be referred to as the second element, and similarly, the second element may also be referred to as the first element.The use of any and all examples or exemplary language ("for example", "such as" etc.) provided by the present embodiment is only intended to better illustrate embodiments of the present invention, and unless otherwise required, the scope of the present invention will not be limited.

[0136] It should be appreciated that embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory. The methods described can be implemented in a computer program using standard programming techniques—including a non-transitory computer-readable storage medium configured with a computer program, wherein the storage medium so configured causes a computer to operate in a specific and predefined manner—according to the methods described in the specific embodiments and the accompanying drawings. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if desired, the program can be implemented in assembly or machine language. In any case, the language can be compiled or interpreted. Furthermore, the program can be run on an application-specific integrated circuit programmed for this purpose.

[0137] In addition, the operations of the processes described in this embodiment may be performed in any suitable order, unless otherwise indicated in this embodiment or otherwise clearly contradicted by the context. The processes described in this embodiment (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that is executed collectively on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions that can be executed by one or more processors.

[0138] Furthermore, the methods can be implemented in any type of computing platform that is operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc. Various aspects of the present invention can be implemented as machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, RAM, ROM, etc., so that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein. In addition, the machine-readable code, or portions thereof, can be transmitted over a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the invention described in this embodiment includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention also includes the computer itself.

[0139] The computer program can be applied to input data to perform the functions described in the present embodiment, thereby converting the input data to generate output data that is stored in a non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the converted data represents a physical and tangible object, including a specific visual depiction of the physical and tangible object produced on the display.

[0140] The above description is merely a preferred embodiment of the present invention. The present invention is not limited to the aforementioned embodiments. As long as the technical effects of the present invention are achieved by the same means, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention. Within the scope of protection of the present invention, various modifications and variations of the technical solutions and / or implementation methods are possible.

Claims

1. An interactive modeling and optimization platform for electromagnetic devices, characterized by: The interactive modeling and optimization platform for electromagnetic devices includes: View layer; the view layer is used to display a visual interactive interface, the visual interactive interface includes at least one sub-interface; Control layer; the control layer is used to store the performance curves and performance indicators corresponding to various electromagnetic devices, perform interactive parameter settings and automatic error detection functions, display options corresponding to modeling algorithms and optimization algorithms through the visual interactive interface, obtain the modeling algorithms and optimization algorithms selected by the user through the visual interactive interface, receive observation variables and targets set by the user, display the execution results of the parameterized modeling process and the optimization process in the visual interactive interface, and import and update existing data files; Model layer; the model layer is used to execute the modeling algorithm and optimization algorithm, and execute the parameterized modeling process and optimization process according to the modeling algorithm and optimization algorithm selected by the user and the set observation variables and objectives; The visual interactive interface is used as a numerical interface for displaying numerical values, a picture interface for displaying simulation curves, a console interface for displaying optimization progress, or a parametric modeling interface; The parametric modeling process includes: Based on electromagnetic simulation data and a user-selected algorithm, a Kriging proxy model is constructed to predict the observed variables and target loss function values set by the user, and the observed variables and target loss function values are displayed in a visual interactive interface of the proxy model sub-interface; wherein the observed variables correspond to the performance curve, and the target loss function value corresponds to the performance index; Update the parametric modeling interface of design variables and update the output of the proxy model corresponding to the design variables in real time; The optimization process includes: After the user completes the platform parameter settings, a cyclic processing process is executed, and each cyclic processing process generates a generation of data files; The cyclic processing process includes: When the executed loop processing process is the first round of loop processing, an initial design variable combination is generated according to the algorithm selected by the user; when the executed loop processing process is not the first round of loop processing, the parametric modeling process is automatically executed according to the existing simulation results and the algorithm selected by the user, and the search algorithm selected by the user searches around the search center on the proxy model to generate a current design variable combination; Automatically generate vbs files with relevant settings and call electromagnetic simulation software HFSS for simulation; After the simulation is finished, the simulation results are automatically read to obtain the values of the observation variables set by the user; Automatically calculate the loss function value of the user-set target; Automatically classify and rank design variable combinations; Automatically update the platform's visual interactive interface; When the loop end condition is met, the execution of all the loop processing processes is ended.

2. The interactive modeling and optimization platform for electromagnetic devices according to claim 1, characterized in that: The at least one sub-interface includes an algorithm setting sub-interface, a path setting sub-interface, a parameter setting sub-interface, a simulation setting sub-interface, a target setting sub-interface, a debugging setting sub-interface, an automatic phase extraction sub-interface, an automatic expression generation sub-interface, and a proxy model sub-interface.

3. The interactive modeling and optimization platform for electromagnetic devices according to claim 1, characterized in that: The interactive parameter setting and automatic error detection functions include: A parameter setting interactive interface is displayed through the visual interactive interface, wherein the parameter setting interactive interface is used for interactive input of various operating parameters, wherein the operating parameters include algorithm parameters, path parameters, design variable parameters, electromagnetic simulation software parameters, custom observation variables and target parameters, and debugging parameters; Performing error detection on information input by the user through the visual interactive interface; the error detection includes character string validity detection, data type detection and value range detection; Capture error information generated during the operation of the view layer, the control layer, and the model layer.

4. The interactive modeling and optimization platform for electromagnetic devices according to claim 1, characterized in that: The electromagnetic devices include filters, power dividers, duplexers, couplers, antenna units and array antennas; The performance curves include an S-parameter curve, a directivity pattern curve, an amplitude curve and a phase curve; The performance indicators include return loss, insertion loss, antenna gain and sidelobe level.

5. The interactive modeling and optimization platform for electromagnetic devices according to claim 1, characterized in that: The modeling algorithm and optimization algorithm include: A Kriging proxy model is constructed based on different regression models and correlation models, and the output of the proxy model is a performance curve or performance index of the electromagnetic device; A dedicated optimization algorithm or a universal optimization algorithm is run on the Kriging agent model; the dedicated optimization algorithm includes a VF algorithm applicable to filters, power splitters, duplexers, and couplers, as well as a VF algorithm, FFT algorithm, and VF-FFT algorithm applicable to antenna units and array antennas; the universal optimization algorithm includes an NSGA-II-I algorithm, an NSGA-II-D algorithm, an NSDE-I algorithm, an NSDE-D algorithm, an OCDE-I algorithm, and an OCDE-D algorithm applicable to all electromagnetic devices; The optimization algorithm includes a search algorithm running on the Kriging agent model, and the search algorithm includes a GA / NSGA-II algorithm, an NSDE algorithm, and an OCDE algorithm; The optimization algorithm includes classification and sorting algorithms for design variable combinations. During the operation of the platform optimization program, different classification and sorting algorithms are allowed to be switched according to actual conditions.

6. The interactive modeling and optimization platform for electromagnetic devices according to claim 1, characterized in that: The loop end conditions include: The optimization result of the last round of cyclic processing meets the goal set by the user; or The cumulative number of rounds of the cyclic processing process reaches the total number of generations set by the user.

7. The interactive modeling and optimization platform for electromagnetic devices according to any one of claims 1 to 5, characterized in that: The functions of importing and updating existing data files include: Import global data files of different projects; Import data files from any cycle of processing in any project; Based on the imported data files and combined with the debugging setting function provided by the platform, the data files can be updated; Continue running based on the optimization results of any round of the cyclic processing process in any project.

8. An interactive modeling and optimization method for electromagnetic devices, characterized in that: The interactive modeling and optimization method of the electromagnetic device includes: Displaying a visual interactive interface, wherein the visual interactive interface includes at least one sub-interface; Store the performance curves and performance indicators corresponding to various electromagnetic devices; Perform interactive parameter setting and automatic error detection functions; Displaying options corresponding to the modeling algorithm and the optimization algorithm through the visual interactive interface; Obtaining the modeling algorithm and optimization algorithm selected by the user through the visual interactive interface, and receiving the observation variables and targets set by the user; Executing the modeling algorithm and the optimization algorithm, and performing the parameterized modeling process and the optimization process according to the modeling algorithm and the optimization algorithm selected by the user and the set observation variables and objectives; Displaying the execution results of the parametric modeling process and the optimization process in the visual interactive interface; The visual interactive interface is used as a numerical interface for displaying numerical values, a picture interface for displaying simulation curves, a console interface for displaying optimization progress, or a parametric modeling interface; The parametric modeling process includes: Based on electromagnetic simulation data and a user-selected algorithm, a Kriging proxy model is constructed to predict the observed variables and target loss function values set by the user, and the observed variables and target loss function values are displayed in a visual interactive interface of the proxy model sub-interface; wherein the observed variables correspond to the performance curve, and the target loss function value corresponds to the performance index; Update the parametric modeling interface of design variables and update the output of the proxy model corresponding to the design variables in real time; The optimization process includes: After the user completes the platform parameter settings, a cyclic processing process is executed, and each cyclic processing process generates a generation of data files; The cyclic processing process includes: When the executed loop processing process is the first round of loop processing, an initial design variable combination is generated according to the algorithm selected by the user; when the executed loop processing process is not the first round of loop processing, the parametric modeling process is automatically executed according to the existing simulation results and the algorithm selected by the user, and the search algorithm selected by the user searches around the search center on the proxy model to generate a current design variable combination; Automatically generate vbs files with relevant settings and call electromagnetic simulation software HFSS for simulation; After the simulation is finished, the simulation results are automatically read to obtain the values of the observation variables set by the user; Automatically calculate the loss function value of the user-set target; Automatically classify and rank design variable combinations; Automatically update the platform's visual interactive interface; When the loop end condition is met, the execution of all the loop processing processes is ended.

Images