Microwave device large-range modeling method and system based on adjoint neural network

Abstract

The invention provides a microwave device large-range modeling method and system based on an adjoint neural network, and the method comprises the steps: obtaining a range of a plurality of device parameters of a microwave device, dividing the range of the device parameters into a plurality of sub-intervals with overlapping intervals, building a sub-model based on the plurality of sub-intervals of the plurality of device parameters, and carrying out the modeling of the plurality of device parameters. Each sub-model is correspondingly provided with an adjoint neural network; pre-training each sub-model based on the training data, and performing auxiliary training through the adjoint neural network corresponding to each sub-model; constructing a sigmoid function corresponding to each overlapping interval for the upper boundary and the lower boundary of each overlapping interval; determining a correction function of each sub-model based on the sub-interval of each device parameter corresponding to each sub-model; and constructing all the sub-models provided with the correction functions into a large-range model, outputting electromagnetic frequency domain characteristic values corresponding to the to-be-calculated device parameter group based on the large-range model, and outputting derivative information corresponding to the device parameter group through an adjoint neural network.

Description

Technical Field

[0001] The present invention relates to the technical field of microwave device design, and in particular to a large-scale modeling method and system for microwave devices based on adjoint neural networks. Background Art

[0002] With the continuous development of technologies such as the new generation of communication networks and the Internet of Things, various new demands and new applications continue to emerge. It is crucial to develop RF microwave devices and circuits with high precision, low design cost and stable performance to meet these new demands and new applications. In order to achieve the above design requirements, RF microwave designers are increasingly relying on computer-aided design (CAD) to simulate and verify circuits. RF microwave modeling and simulation have become two indispensable steps in the design of RF microwave devices and systems. However, when using existing 3D simulation software for electromagnetic (EM) design, the calculation cost is very expensive and time-consuming because it is necessary to repeatedly adjust the geometric parameter values ​​for calculation to find the appropriate values. Establishing an accurate and reliable RF microwave device model can effectively speed up the design of high-frequency, high-performance RF microwave circuits, shorten the design cycle, and meet the current development needs of RF microwave technology.

[0003] Among the existing technologies for quickly acquiring accurate microwave device replacement models, artificial neural networks (ANN) are recognized as a powerful technology for parametric modeling and design optimization of microwave devices. ANN can effectively learn the input-output relationship of electromagnetic devices, and the trained neural network model can accurately and quickly predict the electromagnetic response of microwave devices with geometric parameters as variables. Various technologies have been introduced to develop and improve the parametric models of microwave devices. For example, the adjoint neural network model (SAANN) uses sensitivity analysis to reduce the number of training data required for development and predict the derivatives of the output response to each geometric parameter, and the model has high generalization. For the design of microwave devices with geometric variables varying over a large range, existing technologies cannot solve it. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a large-scale modeling method for microwave devices based on an adjoint neural network to eliminate or improve one or more defects existing in the prior art.

[0005] One aspect of the present invention provides a large-scale modeling method for microwave devices based on adjoint neural network, the method comprising the steps of: Acquire the range of multiple device parameters of the microwave device, divide the range of the device parameters into multiple sub-intervals, there are overlapping intervals between adjacent sub-intervals, and construct sub-models based on the multiple sub-intervals of the multiple device parameters, each sub-model is correspondingly provided with an accompanying neural network; Acquire training data based on the subinterval of the device parameter corresponding to each sub-model, pre-train each sub-model based on the training data, and perform auxiliary training through the accompanying neural network corresponding to each sub-model during the training process; For the upper and lower boundaries of each overlapping interval, construct a sigmoid function corresponding to each overlapping interval; Determine a correction sub-function of each sub-model for each device parameter based on a sub-interval of each device parameter corresponding to each sub-model, and construct a correction function based on all correction sub-functions of the sub-model; All sub-models provided with correction functions are constructed into a large-scale model, and electromagnetic frequency domain characteristic values ​​corresponding to the device parameter group to be calculated are output based on the large-scale model.

[0006] The above scheme is adopted. In the initial training process, this scheme reduces the training time of each sub-model by dividing the intervals, and preliminarily ensures the degree of bridging between the models by overlapping the intervals. In the subsequent processing process, the correction function of each sub-model is constructed by the sigmoid function. The sigmoid function only works on the boundary part of the sub-model, while the internal part remains unchanged. This scheme uses the sigmoid function to modify the trained sub-model. For each segmented area at each geometric input parameter, a sigmoid function is used to perform the modification process, so that the internal output value of the sub-model remains unchanged, the value at the boundary is modified, and the value outside the sub-model range is 0. This will not affect the output values ​​of other models. While ensuring the output accuracy, it can be suitable for the design of microwave devices with geometric variables varying in a large range.

[0007] In some embodiments of the present invention, in the step of constructing a sigmoid function corresponding to each overlapping interval for the upper and lower boundaries of each overlapping interval, the sigmoid function is calculated based on the upper boundary of the overlapping interval. Parameters and The parameter value of the parameter constructs the sigmoid function corresponding to the overlapping interval.

[0008] In some embodiments of the present invention, the sigmoid function is calculated based on the upper boundary of the overlapping interval. Parameters and In the step of calculating the parameter value of the parameter, the sigmoid function is calculated based on the following formula Parameters and parameter: Among them, r is the preset calculation parameter, is the value of the upper boundary of the overlapping interval, is the value of the lower boundary of the overlapping interval.

[0009] In some embodiments of the present invention, the calculation parameter r is calculated according to the following formula: in, The preset adjustment width value.

[0010] In some embodiments of the present invention, in the step of determining a correction sub-function of each sub-model for each device parameter based on the sub-interval of each device parameter corresponding to each sub-model: Determine whether there are adjacent sub-intervals in a shrinking direction and an increasing direction in the sub-interval of the device parameter corresponding to the sub-model; If the sub-model does not have adjacent sub-intervals in the shrinking direction, but has adjacent sub-intervals in the increasing direction, then constructing a modified sub-function based on the first adjustment method; If the sub-model has adjacent sub-intervals in the shrinking direction but does not have adjacent sub-intervals in the increasing direction, constructing a modified sub-function based on the second adjustment method; If the sub-model has adjacent sub-intervals in the shrinking direction and adjacent sub-intervals in the increasing direction, a modified sub-function is constructed based on the third adjustment method.

[0011] In some embodiments of the present invention, the first adjustment method is to use the sigmoid function S corresponding to the overlapping interval of the sub-interval as a correcting sub-function; the second adjustment method is to calculate 1-S for the sigmoid function S corresponding to the overlapping interval of the sub-interval as a correcting sub-function; the third adjustment method is to use two sigmoid functions corresponding to the overlapping interval of the sub-interval, and for the sigmoid function S1 corresponding to the overlapping interval on the left and the sigmoid function S2 corresponding to the overlapping interval on the right of the two overlapping intervals, calculate S1 (1-S2) as the correcting sub-function.

[0012] In some embodiments of the present invention, in the step of constructing a correction function based on all correction sub-functions of the sub-model, the correction sub-functions corresponding to each device parameter of the sub-model are multiplied to obtain a correction function corresponding to each sub-model.

[0013] In some embodiments of the present invention, in the step of outputting the electromagnetic frequency domain characteristic value corresponding to the device parameter group to be calculated based on the large-scale model: Inputting a device parameter set to be calculated into the large-scale model; Correcting the output of each sub-model in the large-scale model by using a corresponding correction function to obtain a value of the correction function; The values ​​of the correction functions of all sub-models are superimposed to obtain the final electromagnetic frequency domain characteristic value.

[0014] In some embodiments of the present invention, in the step of constructing a sub-model based on multiple sub-intervals of multiple device parameters, the sub-intervals for each device parameter are sequentially numbered based on the positions of the sub-intervals, and each sub-model corresponds to a numbering combination of the sub-intervals of each device parameter.

[0015] The second aspect of the present invention also provides a large-scale modeling system for microwave devices based on an adjoint neural network, the system comprising a computer device, the computer device comprising a processor and a memory, the memory storing computer instructions, the processor being used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the system implements the steps implemented by the method described above.

[0016] The third aspect of the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps implemented by the aforementioned large-scale modeling method of microwave devices based on an accompanying neural network.

[0017] Additional advantages, purposes, and features of the present invention will be described in part in the following description, and will become apparent to those skilled in the art after studying the following, or may be learned from the practice of the present invention. The purposes and other advantages of the present invention can be specifically pointed out and obtained in the specification and the accompanying drawings.

[0018] Those skilled in the art will appreciate that the objectives and advantages that can be achieved with the present invention are not limited to the above specific description, and the above and other objectives that can be achieved by the present invention will be more clearly understood from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of this application, and do not constitute a limitation of the present invention.

[0020] Figure 1 A schematic diagram of an implementation of a large-scale modeling method for microwave devices based on an adjoint neural network according to the present invention; Figure 2 It is a schematic diagram of the architecture of the large-scale modeling method of microwave devices based on adjoint neural network of the present invention; Figure 3 This is the output diagram of this scheme when no correction function is set; Figure 4 This is a schematic diagram of the output of the sigmoid function; Figure 5Schematic diagram of the processing principle of the correction function. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0022] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, only structures and / or processing steps closely related to the solutions according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0023] like Figure 1 and 2 As shown, the present invention proposes a large-scale modeling method for microwave devices based on an adjoint neural network, the steps of the method comprising: Step S100, obtaining ranges of multiple device parameters of a microwave device, dividing the ranges of the device parameters into multiple sub-intervals, where there are overlapping intervals between adjacent sub-intervals, and constructing sub-models based on the multiple sub-intervals of the multiple device parameters, each sub-model being correspondingly provided with an accompanying neural network; In some embodiments of the present invention, the microwave device is a microwave radio frequency device, and the device parameters include the height, length, width of the device and the material properties of the device itself, such as dielectric constant and magnetic permeability.

[0024] Step S200, acquiring training data based on the sub-interval of the device parameter corresponding to each sub-model, pre-training each sub-model based on the training data, and performing auxiliary training through the accompanying neural network corresponding to each sub-model during the training process; In the specific implementation process, the sub-model adopts the SAANN neural network structure. SAANN (Sparse Adversarial Adjoint Neural Network) is a neural network structure with strong generalization under the assistance of sensitivity information; The SAANN model consists of two parts: one is the original neural network that trains the electromagnetic response relationship of microwave devices, and the other is the adjoint neural network that trains the derivative information of the electromagnetic response with respect to the sensitivity variable. Both share a set of internal neuron connection parameters, but their activation functions are different. The adjoint neural network has the same input as the original neural network, and the output of the entire SAANN model includes the output of the original network and the output of the adjoint network. During the training process, both the original network and the adjoint network exist, but the final model only includes the structure of the original network.

[0025] In some embodiments of the present invention, during the pre-training process, the sub-model can simultaneously learn the input and output behavior of the filter and the derivative of the output response to the sensitivity parameter, and the accompanying neural network model is derived from the original neural network model, and shares the same weight coefficients with the original neural network. After the model training is completed, the final simple neural network model structure is directly obtained. The input variables of the model are usually the design variables and frequency parameters of the device.

[0026] With the above scheme, after the sub-model adds the sensitivity derivative, the derivative can provide a guide for the change direction of the accompanying neural network fitting function, so that less data can be used for modeling, which speeds up the modeling process in terms of data volume and simulation, and due to the addition of the derivative, the generalization of the model is better. In order to speed up the modeling speed, the training of each sub-model is carried out in parallel. When training the sub-model, in order to avoid over-training or under-training of the neural network model, the number of neurons is adjusted according to the training error and the test error during the training process until a neural network replacement model that meets the user-defined accuracy requirements is established. The established proxy model can provide a fast and accurate prediction of the electromagnetic characteristics and its derivatives of the microwave device within a certain range within its training range.

[0027] Step S300, constructing a sigmoid function corresponding to each overlapping interval for the upper boundary and the lower boundary of each overlapping interval; Step S400, determining a correction sub-function of each sub-model for each device parameter based on the sub-interval of each device parameter corresponding to each sub-model, and constructing a correction function based on all the correction sub-functions of the sub-model; Step S500: construct all sub-models provided with correction functions into a large-scale model, and output electromagnetic frequency domain characteristic values ​​corresponding to the device parameter group to be calculated based on the large-scale model.

[0028] In the specific implementation process, the derivative information of the corresponding device parameter group is output through the accompanying neural network.

[0029] In a specific implementation process, the electromagnetic frequency domain characteristic value may be an S parameter value, a gain or a quality factor, etc.

[0030] The above scheme is adopted. In the initial training process, this scheme reduces the training time of each sub-model by dividing the intervals, and preliminarily ensures the degree of bridging between the models by overlapping the intervals. In the subsequent processing process, the correction function of each sub-model is constructed by the sigmoid function. The sigmoid function only works on the boundary part of the sub-model, while the internal part remains unchanged. This scheme uses the sigmoid function to modify the trained sub-model. For each segmented area at each geometric input parameter, a sigmoid function is used to perform the modification process, so that the internal output value of the sub-model remains unchanged, the value at the boundary is modified, and the value outside the sub-model range is 0, which will not affect the output values ​​of other models. This scheme corrects the boundary of each sub-model through the correction function, while ensuring the output accuracy, it can be applied to the design of microwave devices with geometric variables varying in a large range.

[0031] In some embodiments of the present invention, in the step of constructing a sigmoid function corresponding to each overlapping interval for the upper and lower boundaries of each overlapping interval, the sigmoid function is calculated based on the upper boundary of the overlapping interval. Parameters and The parameter value of the parameter constructs the sigmoid function corresponding to the overlapping interval.

[0032] In some embodiments of the present invention, the sigmoid function is calculated based on the upper boundary of the overlapping interval. Parameters and In the step of calculating the parameter value of the parameter, the sigmoid function is calculated based on the following formula Parameters and parameter: Among them, r is the preset calculation parameter, is the value of the upper boundary of the overlapping interval, is the value of the lower boundary of the overlapping interval.

[0033] In some embodiments of the present invention, the calculation parameter r is calculated according to the following formula: in, It is the preset adjustment width value, which can be set to 0.999, 0.997 or 0.998, etc.

[0034] In some embodiments of the present invention, in the step of determining a correction sub-function of each sub-model for each device parameter based on the sub-interval of each device parameter corresponding to each sub-model: Determine whether there are adjacent sub-intervals in a shrinking direction and an increasing direction in the sub-interval of the device parameter corresponding to the sub-model; If the sub-model does not have adjacent sub-intervals in the shrinking direction, but has adjacent sub-intervals in the increasing direction, then constructing a modified sub-function based on the first adjustment method; If the sub-model has adjacent sub-intervals in the shrinking direction but does not have adjacent sub-intervals in the increasing direction, constructing a modified sub-function based on the second adjustment method; If the sub-model has adjacent sub-intervals in the shrinking direction and adjacent sub-intervals in the increasing direction, a modified sub-function is constructed based on the third adjustment method.

[0035] In some embodiments of the present invention, the first adjustment method is to use the sigmoid function S corresponding to the overlapping interval of the sub-interval as a correcting sub-function; the second adjustment method is to calculate 1-S for the sigmoid function S corresponding to the overlapping interval of the sub-interval as a correcting sub-function; the third adjustment method is to use two sigmoid functions corresponding to the overlapping interval of the sub-interval, and for the sigmoid function S1 corresponding to the overlapping interval on the left and the sigmoid function S2 corresponding to the overlapping interval on the right of the two overlapping intervals, calculate S1 (1-S2) as the correcting sub-function.

[0036] like Figure 4 and 5 As shown, Figure 5 Where S(x) is the original sigmoid function, h(x)=0.01 In the specific implementation process, when the adjacent sub-model of the j-th model is on its right side, in order to keep the internal value output of the j-th sub-model unchanged and the value of the adjacent sub-model does not affect the output of the j-th sub-model, multiply (1-S). On the contrary, when the adjacent sub-model of the j-th model is on its left side, in order to keep the internal value output of the j-th sub-model unchanged and the value of the adjacent sub-model does not affect the output of the j-th sub-model, multiply S. When the j-th sub-model has adjacent sub-models on both the left and right sides, multiply S*(1-S).

[0037] Using the above solution, the characteristic of the sigmoid function is that the value on the left side of the processing range remains unchanged, and the value on the right side of the processing range becomes 0. The 1-sigmoid function is the opposite.

[0038] In the specific implementation process, this scheme can process the boundary part of the sub-model function while keeping the internal part unchanged, and the sigmoid function has such characteristics. Figure 4 As shown, it can be seen that by constraining the device parameters and , the output function fD that does not connect multiple sub-models only works at the boundary part of the sub-model function, while the internal part remains unchanged. This scheme uses the sigmoid function to modify the trained sub-model function. For each segmentation area at each geometric input parameter, a sigmoid function is required to perform the modification process so that the output value inside the sub-model remains unchanged, the value at the boundary is modified, and the value outside the sub-model range is 0, which will not affect the output values ​​of other models.

[0039] In some embodiments of the present invention, in the step of constructing a correction function based on all correction sub-functions of the sub-model, the correction sub-functions corresponding to each device parameter of the sub-model are multiplied to obtain a correction function corresponding to each sub-model.

[0040] In some embodiments of the present invention, in the step of outputting the electromagnetic frequency domain characteristic value corresponding to the device parameter group to be calculated based on the large-scale model: Inputting a device parameter set to be calculated into the large-scale model; Correcting the output of each sub-model in the large-scale model by using a corresponding correction function to obtain a value of the correction function; The values ​​of the correction functions of all sub-models are superimposed to obtain the final electromagnetic frequency domain characteristic value.

[0041] With the above scheme, after the models of each sub-region are obtained through pre-training, the sub-models are trained independently, and discontinuity problems will occur at the boundaries of adjacent models. In order to reduce the difference between the outputs of adjacent sub-models at the boundaries, this scheme proposes to set an overlapping area for adjacent sub-models along each dimension as mentioned above, and the adjacent sub-models will be trained to this part of the data. After the sub-model training process, the adjacent sub-models have very similar output responses at the boundaries, and then the mathematical characteristics of the sigmoid function are used to correct each dimension of each sub-model to solve the multi-dimensional discontinuity problem; the sigmoid function can keep the data within the sub-model range unchanged, and make the data outside the sub-model range 0, and reasonably modify the data on the boundary. After the sigmoid function is used to perform special processing on the boundary of the overall model, a continuous large model will be obtained. Finally, the output responses of each sub-model are superimposed to obtain the output response of the total model.

[0042] In some embodiments of the present invention, in the step of constructing a sub-model based on multiple sub-intervals of multiple device parameters, the sub-intervals for each device parameter are sequentially numbered based on the positions of the sub-intervals, and each sub-model corresponds to a numbering combination of the sub-intervals of each device parameter.

[0043] For example, if the device parameters include A and B, the interval of each device parameter is (0, 1), and each device parameter is divided into two sub-intervals (0, 0.6) and (0.4, 1), then (0, 0.6) is numbered 1 and (0.4, 1) is numbered 2. The corresponding sub-models include sub-model No. 1 corresponding to the sub-interval of A parameter number 1 and the sub-interval of B parameter number 1; sub-model No. 2 corresponding to the sub-interval of A parameter number 1 and the sub-interval of B parameter number 2; sub-model No. 3 corresponding to the sub-interval of A parameter number 2 and the sub-interval of B parameter number 1; sub-model No. 4 corresponding to the sub-interval of A parameter number 2 and the sub-interval of B parameter number 2.

[0044] In summary, in the current process of RF microwave device modeling, both designers and users hope that the model can work in a larger range, that is, when the design variables change in a larger range, the model always remains accurate. Establishing an accurate large-scale parameterized model for the electromagnetic response of RF microwave devices in an electromagnetic environment can increase the scope of use of the model and provide accurate and fast output response in a large range; at the same time, it can reduce the number of iterations during the optimization design of RF microwave devices and avoid falling into a local optimal solution. However, as the modeling scope, that is, the range of design variable changes, increases, the complexity and nonlinearity of the electromagnetic response increase rapidly.

[0045] In order to establish an accurate large-scale model, the most commonly used methods in the prior art are the knowledge-based neural network modeling method and the spatial mapping method combined with the neural network. These two methods increase the working range of the model by introducing the equivalent circuit of the device or the prior physical knowledge, mathematical model and other information of the device. However, both methods rely on empirical models or prior knowledge, and therefore have certain application limitations. When a new type of RF microwave device does not have an applicable empirical model or prior knowledge, these two methods cannot be used to model the device. Moreover, although the above two methods can expand the working range of the model, they do not really solve the problem of large-scale modeling. In the process of using the existing modeling method based on a single neural network for large-scale parametric modeling, there are problems such as large data volume requirements, high model complexity, high computational cost, and low model accuracy.

[0046] Therefore, this scheme proposes a large-scale parameterized model of microwave devices with adjoint neural networks based on decomposition technology to quickly establish an accurate parameterized modeling for microwave devices with geometric variables varying over a large range. This scheme is suitable for the design of microwave RF filters, such as cavity filters, microstrip line filters, and interdigital filters. Compared with the existing large-scale parameterized modeling of microwave filters, the model proposed in this scheme has better generalization, requires less training data, speeds up modeling, and can predict the derivatives of output responses with respect to various geometric parameters, which greatly solves the difficulty of large-scale modeling of microwave filters and speeds up the design and optimization of microwave RF filters.

[0047] An embodiment of the present invention also provides a large-scale modeling system for microwave devices based on an adjoint neural network, the system comprising a computer device, the computer device comprising a processor and a memory, the memory storing computer instructions, the processor being used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the system implements the steps implemented by the method described above.

[0048] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps implemented by the aforementioned large-scale modeling method of microwave devices based on accompanying neural networks are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0049] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0050] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0051] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0052] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A large-scale modeling method for microwave devices based on adjoint neural network, characterized in that: The steps of the method include: Acquire the range of multiple device parameters of the microwave device, divide the range of the device parameters into multiple sub-intervals, there are overlapping intervals between adjacent sub-intervals, and construct sub-models based on the multiple sub-intervals of the multiple device parameters, each sub-model is correspondingly provided with an accompanying neural network; Acquire training data based on the subinterval of the device parameter corresponding to each sub-model, pre-train each sub-model based on the training data, and perform auxiliary training through the accompanying neural network corresponding to each sub-model during the training process; For the upper and lower boundaries of each overlapping interval, construct a sigmoid function corresponding to each overlapping interval; Determine a correction sub-function of each sub-model for each device parameter based on a sub-interval of each device parameter corresponding to each sub-model, and construct a correction function based on all correction sub-functions of the sub-model; All sub-models provided with correction functions are constructed into a large-scale model, and electromagnetic frequency domain characteristic values ​​corresponding to the device parameter group to be calculated are output based on the large-scale model.

2. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 1 is characterized in that: In the step of constructing a sigmoid function corresponding to each overlapping interval for the upper and lower boundaries of each overlapping interval, the sigmoid function is calculated based on the upper boundary of the overlapping interval. Parameters and The parameter value of the parameter constructs the sigmoid function corresponding to the overlapping interval.

3. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 2 is characterized in that: Calculate the sigmoid function based on the upper boundary of the overlapping interval Parameters and In the step of calculating the parameter value of the parameter, the sigmoid function is calculated based on the following formula Parameters and parameter: Among them, r is the preset calculation parameter, is the value of the upper boundary of the overlapping interval, is the value of the lower boundary of the overlapping interval.

4. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 3 is characterized in that: The calculation parameter r is calculated according to the following formula: in, The preset adjustment width value.

5. The large-scale modeling method of microwave devices based on adjoint neural network according to any one of claims 1 to 4, characterized in that: In the step of determining a correction sub-function of each sub-model for each device parameter based on the sub-interval of each device parameter corresponding to each sub-model: Determine whether there are adjacent sub-intervals in a shrinking direction and an increasing direction in the sub-interval of the device parameter corresponding to the sub-model; If the sub-model does not have adjacent sub-intervals in the shrinking direction, but has adjacent sub-intervals in the increasing direction, then constructing a modified sub-function based on the first adjustment method; If the sub-model has adjacent sub-intervals in the shrinking direction but does not have adjacent sub-intervals in the increasing direction, constructing a modified sub-function based on the second adjustment method; If the sub-model has adjacent sub-intervals in the shrinking direction and adjacent sub-intervals in the increasing direction, a modified sub-function is constructed based on the third adjustment method.

6. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 5, characterized in that: The first adjustment method is to use the sigmoid function S corresponding to the overlapping interval of the sub-interval as a correction sub-function; the second adjustment method is to calculate 1-S for the sigmoid function S corresponding to the overlapping interval of the sub-interval as a correction sub-function; the third adjustment method is to correspond the overlapping interval of the sub-interval to two sigmoid functions, and for the sigmoid function S1 corresponding to the overlapping interval on the left and the sigmoid function S2 corresponding to the overlapping interval on the right of the two overlapping intervals, calculate S1 (1-S2) as the correction sub-function.

7. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 1 or 6, characterized in that: In the step of constructing a correction function based on all correction sub-functions of the sub-model, the correction sub-functions corresponding to each device parameter of the sub-model are multiplied together to obtain a correction function corresponding to each sub-model.

8. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 1 is characterized in that: In the step of outputting electromagnetic frequency domain characteristic values ​​corresponding to the device parameter group to be calculated based on the large-scale model: Inputting a device parameter set to be calculated into the large-scale model; Correcting the output of each sub-model in the large-scale model by using a corresponding correction function to obtain a value of the correction function; The values ​​of the correction functions of all sub-models are superimposed to obtain the final electromagnetic frequency domain characteristic value.

9. The large-scale modeling method of microwave devices based on adjoint neural network according to claim 1, characterized in that: In the step of constructing a sub-model based on multiple sub-intervals of multiple device parameters, the sub-intervals of each device parameter are sequentially numbered based on the positions of the sub-intervals, and each sub-model corresponds to a number combination of the sub-intervals of each device parameter.

10. A microwave device large-scale modeling system based on adjoint neural network, characterized in that: The system includes a computer device, which includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps implemented by the method as described in any one of claims 1 to 9.

Images