Phased-array antenna structure high-precision reconstruction and electrical performance analysis method based on small sample data fusion

Abstract

The invention provides a phased-array antenna structure high-precision reconstruction and electrical performance analysis method based on small sample data fusion, and relates to the field of structure deformation reconstruction and analysis. The method comprises the following steps: firstly, constructing a finite element model according to a theoretical geometric model of a phased-array antenna array surface structure, and obtaining low-fidelity strain-deformation data by setting different constraint conditions and loads; and meanwhile, a strain sensor is arranged on an actual array surface structure to obtain high-fidelity strain-deformation data. After the two types of data are decomposed and recombined, a multi-channel deep convolution generation adversarial neural network is used for fitting a mapping relation, a generator and discriminator model is trained, and accurate reconstruction of array surface structure deformation is achieved. And based on a reconstruction result, calculating an array element position offset, a phase difference and a directional diagram function, regarding array plane deformation as a random field variable, and analyzing electrical performance. The method solves the problems that a traditional method is complex in modeling and poor in applicability, and a sensing basis is provided for active control over array surface structure deformation.

Description

Technical Field

[0001] The present invention relates to the field of structural deformation reconstruction and analysis, and in particular to a method for high-precision reconstruction and electrical performance analysis of phased array antenna structures based on small sample data fusion. Background Art

[0002] In the actual service environment of active phased array radar, antenna surface deformation caused by wind load, overload, random road vibration, impact vibration, etc. will also cause the antenna array element position to shift, thereby affecting the electromagnetic performance of the active phased array antenna. The accuracy of the array surface structure reconstruction will affect the subsequent electromagnetic performance analysis of the phased array antenna. Through the electromagnetic performance analysis of the phased array antenna, the overall performance of the phased array antenna can be directly reflected. The influence of the antenna surface structure deformation under the action of service load on the electrical performance can be studied, and the coupling relationship between the antenna structure displacement field and the electromagnetic field can be determined, providing a basis for the subsequent online perception and multi-level compensation of the phased array antenna surface deformation.

[0003] There are two problems with the array structure deformation reconstruction method based on the theoretical model:

[0004] On the one hand, the modal method, inverse finite element method, and curvature method all require the establishment of accurate theoretical models. However, the antenna array structure is actually composed of many parts with different material properties and complex structures assembled through a variety of connection processes. Establishing an accurate model of the structure is difficult and time-consuming.

[0005] On the other hand, the improved traditional method relies on calibration data (based on experimentally measured stress-deformation data) to improve model accuracy. However, high-precision calibration data has problems such as small sample size and difficulty in obtaining. Although low-precision prediction data has a large sample size, the quality of the sample data is low, which affects the accuracy of the phased array antenna surface deformation reconstruction and leads to limited improvement in model accuracy.

[0006] Compared to the limitations of traditional deformation reconstruction methods, structural reconstruction methods based on neural network algorithms do not require a precise structural model, have a simple reconstruction process, and are widely applicable. They are an effective way to monitor and reconstruct fiber Bragg grating (FBG) structures in phased array antenna arrays. Currently, related research has used CNN convolutional neural networks for phased array antenna array deformation reconstruction. However, CNN convolutional neural networks cannot be trained under supervision, and the introduction of fully connected layers in the model results in relatively many parameters and is prone to overfitting. The generative adversarial neural network algorithm discards the fully connected layers, avoiding potential problems with CNN convolutional neural networks. Furthermore, the generator can continuously generate new data, which is then judged by the discriminator for authenticity. The generative adversarial process introduces a supervised loss, and the generator and discriminator parameters are updated in real time, effectively improving the accuracy of phased array antenna array reconstruction.

[0007] However, structural deformation reconstruction methods based on generative adversarial neural networks (GANs) are essentially adversarial training problems. Reconstruction accuracy depends on the training performance of the generative model and the diversity of the data. While not requiring a precise structural model, these methods also face the challenges of limited and time-consuming calibration data samples. Classic GAN-based structural deformation reconstruction methods either rely on a large number of randomly generated low-fidelity samples (generating low-quality sample data) or directly train on calibrated high-fidelity sample data to establish a mapping between structural displacement and normalized probability values. Both approaches have limitations. The former, in which the GAN model is trained on low-fidelity samples, suffers from low sample precision, resulting in significant discrepancies between the generated samples and the actual situation. This requires continuous generation of adversarial data, is time-consuming, and results in low reconstruction accuracy. The latter, while trained on high-fidelity samples, suffers from a smaller sample size. Under the complex loads experienced during service, the model's generalization ability is weak and overfitting can occur. Summary of the Invention

[0008] Purpose of the invention: To propose a high-precision reconstruction and electrical performance analysis method for phased array antenna structures based on small sample data fusion. By fusing low-fidelity and low-cost simulation data with high-fidelity and high-cost calibration data, accurate prediction of array surface structure deformation under complex service loads can be achieved, and the electromagnetic performance of phased array antennas can be analyzed. This solves the problems of complex modeling and poor applicability of traditional methods, and provides a perception basis for active deformation control of array surface structures.

[0009] The present invention proposes a method for high-precision reconstruction and electrical performance analysis of phased array antenna structures based on small sample data fusion, comprising the following steps:

[0010] S1. Construct the corresponding finite element model based on the theoretical geometric model of the phased array antenna surface structure;

[0011] S2. Setting different constraints and loads of the finite element model to simulate different service conditions of the phased array antenna surface structure and obtain low-fidelity strain-deformation data of the phased array antenna surface structure. ;

[0012] S3. Arrange strain sensors on the actual phased array antenna array structure, establish the coordinate system of the phased array antenna array structure, apply different loads to the phased array antenna array structure, and obtain high-fidelity strain-deformation data of the phased array antenna array structure. ;

[0013] S4. Low-fidelity strain-deformation data and high-fidelity strain-deformation data Decompose according to the orthogonal relationship of the spatial position in the coordinate system to obtain the strain data set under the XY component;

[0014] S5. Reorganize the strain data set under the XY component into a low-fidelity strain image under the XY component according to the spatial position through a predetermined mapping relationship 、 and high-fidelity strain images 、 ;

[0015] S6. Low-fidelity strain-deformation data and high-fidelity strain-deformation data The displacement data in the image is decomposed into low-fidelity displacement component data sets in the XYZ directions according to the coordinate system. 、 、 and high-fidelity displacement component datasets 、 、 ;

[0016] S7, based on multi-channel deep convolution to generate adversarial neural network, respectively fitting low-fidelity strain images 、 With low-fidelity displacement component dataset 、 、 The mapping relationship between them is used to obtain the generator model 、 、 ;

[0017] S8, high-fidelity strain image 、 Input the generator model 、 、 , continuously generate confrontation and obtain the reconstruction result of the phased array antenna surface structure;

[0018] S9. Analyze the electrical performance of the phased array antenna based on the reconstruction results of the array surface structure, including calculating the element position offset according to the deformation of the array surface structure, deriving the phase difference between the elements, and then obtaining the radiation pattern function of the phased array antenna, and analyzing the characteristics of the radiation pattern function.

[0019] In a further embodiment, in step S1, the finite element model is an Abaqus finite element model. In step S3, the strain sensor is a fiber Bragg grating strain sensor; and a coordinate system for the phased array antenna array structure is established using a binocular vision sensor. In step S3, the load applied to the phased array antenna array structure includes a thermal load and a concentrated force load.

[0020] In a further embodiment, in step S4, the low-fidelity strain-deformation data obtained in step S2 is and the high-fidelity strain-deformation data obtained in step S3 Decompose according to the orthogonal relationship of the spatial position in the coordinate system to obtain the strain data set under the XY component 、 、 、 ;in, , ; Represents the XY components of low-fidelity strain data A real matrix with M rows and N columns; Represents the XY components of high-fidelity strain data 、 A real matrix with P rows and Q columns.

[0021] In a further embodiment, the method for high-precision reconstruction and electrical performance analysis of a phased array antenna structure further includes:

[0022] Training the discriminator: Using a mini-batch strategy, we can train the discriminator from high-fidelity strain-deformation data. Extract real samples from the generator and input them into the discriminator to calculate the loss, and calculate the gradient through the back propagation algorithm; use the current generator to generate fake samples and input them into the discriminator to calculate the loss and back propagate again to calculate the gradient; accumulate the gradients of the two batches of samples and update the discriminator parameters;

[0023] Training the generator: The generator is trained by maximizing the loss function to update the generator parameters so that the discriminator can identify the samples generated by the generator as real samples.

[0024] In a further embodiment, the loss functions of the discriminator and the generator are as follows:

[0025]

[0026] Where, is the predicted output of the generator model, ranging from (0,1); is the true label, the true sample is marked as 1, and the false sample is marked as 0; is the loss of a single sample; L is the set of losses for all samples.

[0027] In a further embodiment, in step S8, confrontation is continuously generated, and finally the deformation result of the phased array antenna surface structure is reconstructed to determine the number and distribution characteristics of the elements of the phased array antenna. The array elements are equally spaced rectangular grid plates, and the element spacing along the x-axis and y-axis directions is and .

[0028] In a further embodiment, step S9 comprises:

[0029] Assume that any point in space ,point Relative to the coordinate system It can be expressed as direction cosines ,point The angle between the coordinate axis and the coordinate axis is expressed as:

[0030]

[0031] Set up the first Array Elements The position offset is , leading to the Array element relative to the The spatial phase difference of the unit is:

[0032]

[0033] Assuming that the deformation of the antenna surface only affects the electric field phase of the array unit and does not change its amplitude, we can finally get the first Unit relative to the The directional pattern function of the unit is:

[0034]

[0035] Where, For the The array pattern of the antenna elements; is the unit excitation current, The phase difference within the array of the unit is ; j is an imaginary number; is the propagation constant of free space, and the position deformations are 、 、 .

[0036] In a further embodiment, step S9 further includes:

[0037] Treat the deformation of the array as a random field variable and the deformation of the array as a random variable , for which antenna units, the deformation of the array is expressed as , the mean and variance are and ;

[0038] The mean variance of the unit phase difference is derived based on the random error of the array unit deformation, and the mean variance of the pattern function of the entire array antenna at each observation point is calculated. ; get the first Array element relative to the The spatial phase difference of the unit is:

[0039]

[0040] make

[0041] Then the pattern function of the phased array antenna is expressed as:

[0042]

[0043] Directional pattern function The value of changes with the unit position, by analyzing the unit deformation The mean and variance of the calculation unit position direction pattern function.

[0044] In addition, the present invention also discloses an electronic device, which includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the above-mentioned high-precision reconstruction and electrical performance analysis method of phased array antenna structure based on small sample data fusion.

[0045] In addition, the present invention also discloses a computer-readable storage medium, which stores at least one executable instruction. When the executable instruction is run on an electronic device, the electronic device executes the above-mentioned high-precision reconstruction and electrical performance analysis method of phased array antenna structure based on small sample data fusion.

[0046] Compared with the prior art, the present invention has at least the following beneficial effects:

[0047] (1) High-precision reconstruction of array structure deformation is achieved, and the steps of array deformation reconstruction are simplified. Compared with traditional methods (curvature method, inverse finite element method, curvature method, etc.), there is no need to establish complex models, obtain calibration data, and have limited sample data.

[0048] (2) A structural strain-deformation parameter model is established by using a large amount of low-cost, low-fidelity data (obtained through simulation experiments), and then a limited amount of high-cost, high-fidelity calibration strain data is input into the strain-deformation parameter model to generate predicted deformation data. Finally, the mapping relationship between the measured deformation data, predicted deformation data and normalized probability values ​​is fitted. Compared with the existing machine learning methods, this method effectively improves the accuracy and generalization ability of the array structure deformation reconstruction, and can be used to predict structural deformation under complex loads.

[0049] (3) The establishment of the structural strain-deformation mapping relationship is realized in the form of an image, and the relationship between strain, deformation and normalized probability value is established by using a generative adversarial neural network model. Compared with the existing deep learning method, it can effectively reduce the number of model parameters, realize the multi-level feature extraction of strain-deformation-normalized probability value, improve the training speed of the model, shorten the training time of the model, and enhance the robustness of the model.

[0050] (4) The phased array antenna array surface structure deformation reconstruction based on the generative adversarial neural network framework can predict the electrical performance of the reconstructed array surface. The uncertainty deformation of the antenna array surface under the service load can be used to infer the mean variance of the directional pattern function of the entire array antenna at each observation point, simplifying the electrical performance analysis of the model. It is suitable for the analysis and prediction of the electrical performance of the antenna array surface under complex service loads. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is an overall flow chart of the method of the present invention.

[0052] Figure 2 3. It is a structural deformation reconstruction framework diagram of multi-fidelity data fusion in an embodiment.

[0053] Figure 3 2 is a schematic diagram of the field intensity pattern function of the phased array antenna in the embodiment.

[0054] Figure 4 is a mean value diagram of the phased array antenna pattern function in the embodiment.

[0055] Figure 5 is a variance diagram of the phased array antenna pattern function in the embodiment.

[0056] Figure 6 Schematic diagram of the Abaqus finite element simulation model of the array structure in the embodiment.

[0057] Figure 7 2 is a displacement diagram of the array structure measuring points in the embodiment.

[0058] Figure 8 1 is a reconstruction error graph of measured data trained using a generative adversarial neural network in an embodiment.

[0059] Figure 9 3 is a graph showing the error in the deformation reconstruction by fusion of measured data and simulation data in the embodiment. DETAILED DESCRIPTION

[0060] In the following description, numerous specific details are provided to provide a more thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without one or more of these details. In other instances, certain technical features well known in the art have not been described to avoid confusion with the present invention.

[0061] To address the issues of high reliance on the number of real samples, weak generalization of the reconstruction model, and low reconstruction accuracy, this paper proposes a high-precision reconstruction method and electrical performance analysis for phased array antenna structures based on small-sample data fusion. This method differs from existing array surface structure prediction methods in that it uses limited, high-cost, high-fidelity calibration data to improve a large amount of low-cost, low-fidelity data. The low-fidelity data comes from structural deformation simulations using Abaqus, while the high-fidelity data comes from actual calibration tests. Furthermore, the proposed data fusion method uses images to represent the strain monitoring values ​​of fiber Bragg gratings (FBGs) spatially positioned on the structure. A deep convolutional generative adversarial network (DCGAN) model is then used to establish mappings between strain and deformation, and between deformation and normalized probability values, across the different data. Compared to existing methods, this method reduces the number of model parameters, enables multi-layered feature extraction of strain, deformation, and normalized probability values, and enhances the robustness of the model. Thirdly, the electrical performance analysis method proposed in the present invention studies the array surface after high-precision reconstruction of the phased array antenna structure, analyzes the directional pattern characteristics of the reconstructed array surface, and calculates the mean and variance of the phased array directional pattern function after reconstruction and deformation.

[0062] This embodiment discloses a method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion. The flow chart is shown in Figure 1 , and combined with Figure 2 The present invention is further illustrated by the structural deformation reconstruction framework diagram of multi-fidelity data fusion. The specific steps are as follows:

[0063] 1) Based on the theoretical geometric model of the phased array antenna array structure, an Abaqus finite element model of the array structure is constructed.

[0064] 2) Use Python scripts to set different constraints and loads on the finite element model, simulate different service conditions of the array structure, and obtain a large amount of low-fidelity strain-deformation data of the phased array antenna array structure. .

[0065] 3) Fiber Bragg grating strain sensors are placed on the actual array structure of the phased array antenna, and a binocular vision sensor is used to establish the coordinate system of the array structure. Different thermal loads and concentrated force loads are applied to the physical structure of the phased array antenna array to obtain a small number of high-fidelity data samples of the array structure strain-deformation. .

[0066] 4) 、 The strain data is decomposed into data sets according to the orthogonal relationship of spatial position 、 、 、 , , .

[0067] 5) Reorganize the decomposed strain data set into an image according to the spatial position through a certain mapping relationship 、 、 、 .

[0068] 6) 、 The displacement data is decomposed into displacement component data sets in the x, y, and z directions according to the workpiece coordinate system 、 、 、 、 、 .

[0069] 7) Generate adversarial neural networks based on multi-channel deep convolution to fit 、 and 、 、 The mapping relationship of the output features in the generator model is obtained 、 、 .

[0070] 8) Reconstructing high-fidelity data into images 、 Enter the DCGAN generator model 、 、 , we get the output prediction 、 、 .

[0071] 9) Generate adversarial neural network fitting input based on multi-channel deep convolution 、 、 and normalized probability value The output probability value approaches 0, indicating that the input data is fake data generated by simulation.

[0072] 10) Generate adversarial neural network fitting input based on multi-channel deep convolution 、 、 and normalized probability value The output probability value approaches 1, indicating that the input data is true data.

[0073] 11) Complete the parameter settings for the generator and discriminator, specifying how they learn through the loss function and optimizer. Use the binary cross entropy loss (BCELoss) function defined in PyTorch:

[0074]

[0075] in is the predicted output of the model, ranging from (0,1); is the true label, the true sample is marked as 1, and the false sample is marked as 0; is the loss of a single sample; L is the set of losses for all samples.

[0076] 12) Discriminator training: Using a mini-batch strategy, a batch of real samples are extracted from a small number of high-fidelity data samples of the strain and deformation of the array structure, and these samples are input into the discriminator to calculate the loss. , the gradient is calculated by the back-propagation algorithm.

[0077] 13) Use the current generator to generate a batch of fake samples, and input these fake samples into the discriminator to calculate the loss , calculate the gradient again through back propagation; accumulate the gradients of these two batches of samples and then update the discriminator parameters.

[0078] 14) Generator training: The goal of the generator is to minimize the loss function Generate more realistic samples. Since this method may not provide enough gradients in the early stage of training, the learning effect of the generator is poor. Therefore, the maximum Training the generator and updating the generator parameters is equivalent to making the discriminator believe that the samples generated by the generator are real.

[0079] 15) By continuously generating confrontation through the above steps, we can eventually reconstruct the array deformation and determine the number and distribution characteristics of the phased array antenna elements. Assume The array elements are equally spaced rectangular grid plates, and the element spacing along the x-axis and y-axis directions is and .

[0080] 16) Assume that any point in space ,point Relative to the coordinate system Direction cosines can be expressed as ,point The angle between the coordinate axis and the coordinate axis is expressed as follows:

[0081]

[0082] 17) Assume that Array Elements The position offset is , leading to the Array element relative to the The spatial phase difference of the unit is shown in formula (2):

[0083]

[0084] 18) Assuming that the deformation of the antenna surface only affects the phase of the electric field of the array element and does not change its amplitude, we can finally get the first Unit relative to the The directivity pattern function of the unit is shown in the following equations (3) and Figure 3 shown.

[0085]

[0086] in, For the The array pattern of the antenna elements, is the unit excitation current, unit The phase difference within the array is , j is an imaginary number, is the propagation constant (phase shift constant) of free space, and the position deformations are , , .

[0087] 19) When an active phased array antenna is in service, the array surface will be subjected to loads such as its own weight, wind load, and temperature, and will undergo slight deformation, which will increase the difficulty of subsequent deformation perception and multi-level compensation. The array surface deformation will affect the electrical performance of the antenna. In order to predict the electrical performance of the reconstructed array surface, the array surface deformation is regarded as a random field variable. , for which antenna units, the deformation of the array is expressed as , the mean and variance are and .

[0088] 20) Based on the random error of array unit deformation, the mean variance of the unit phase difference can be derived, and the mean variance of the pattern function of the entire array antenna at each observation point can be calculated, where ; Substituting into formula (2) we can get Array element relative to the The spatial phase difference of the unit is:

[0089]

[0090] make

[0091] Then the pattern function of the phased array antenna can be expressed as:

[0092]

[0093] According to the above expression, the pattern function The value of changes with the unit position, by analyzing the unit deformation The mean and variance of the unit position pattern function can be calculated by Figure 4 and Figure 5 .

[0094] The present invention uses a phased array antenna array structure array model test fixture, Figure 6 To establish the Abaqus finite element simulation model of the array structure, Figure 7 In order to obtain the displacement data of the array structure measurement points during the simulation process, Figure 8 In order to directly generate the structural deformation displacement error obtained by adversarial neural network training through the measured deformation data of the array structure, the maximum error of the measurement point displacement prediction is 0.235mm, and the average reconstruction error is 0.182mm. Figure 9 This is a schematic diagram of the phased array antenna structure deformation reconstruction error by fusing low-fidelity simulation data and high-fidelity calibration data in the present invention. The maximum measurement point displacement reconstruction error is 0.082mm, and the average measurement point displacement error is 0.071mm. The reconstruction accuracy of the present invention is higher.

[0095] The present invention can be applied to the deformation perception and active control of the array surface structure of a phased array antenna. By arranging fiber grating sensors in the array surface structure, the structural deformation perception of the array surface structure under complex service loads can be achieved, and the reconstructed array surface pattern function can be calculated and the electrical performance of the reconstructed array surface can be analyzed.

[0096] The technical process of the method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion disclosed in the above embodiment can be implemented in whole or in part through software, hardware, firmware or any other combination.

[0097] When implemented using hardware, the aforementioned embodiments can be run on an electronic device by compiling all or part of the operating logic and computing processes into software. The electronic device includes a processor, a memory, a communication interface, and a communication bus. The processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction, which enables the processor to execute the technical process disclosed in the aforementioned embodiments.

[0098] When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. If the above methods are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application, or the part that contributes to the relevant technology, can be embodied in the form of a software product. The software product is stored in a storage medium and includes several instructions for enabling an electronic device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present application. The aforementioned storage media include various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), magnetic disks, or optical disks. Thus, the embodiments of the present application are not limited to any specific hardware, software, or firmware, or any combination of hardware, software, and firmware.

[0099] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A method for high-precision reconstruction and electrical performance analysis of phased array antenna structures based on small sample data fusion, characterized in that: The steps include: S1. Construct the corresponding finite element model based on the theoretical geometric model of the phased array antenna surface structure; S2. Setting different constraints and loads of the finite element model to simulate different service conditions of the phased array antenna surface structure and obtain low-fidelity strain-deformation data of the phased array antenna surface structure. ; S3. Arrange strain sensors on the actual phased array antenna array structure, establish the coordinate system of the phased array antenna array structure, apply different loads to the phased array antenna array structure, and obtain high-fidelity strain-deformation data of the phased array antenna array structure. ; S4. Low-fidelity strain-deformation data and high-fidelity strain-deformation data Decompose according to the orthogonal relationship of the spatial position in the coordinate system to obtain the strain data set under the XY component; S5. Reorganize the strain data set under the XY component into a low-fidelity strain image under the XY component according to the spatial position through a predetermined mapping relationship 、 and high-fidelity strain images 、 ; S6. Low-fidelity strain-deformation data and high-fidelity strain-deformation data The displacement data in the image is decomposed into low-fidelity displacement component data sets in the XYZ directions according to the coordinate system. 、 、 and high-fidelity displacement component datasets 、 、 ; S7, based on multi-channel deep convolution to generate adversarial neural network, respectively fitting low-fidelity strain images 、 With low-fidelity displacement component dataset 、 、 The mapping relationship between them is used to obtain the generator model 、 、 ; S8, high-fidelity strain image 、 Input the generator model 、 、 , continuously generate confrontation and obtain the reconstruction result of the phased array antenna surface structure; S9. Analyze the electrical performance of the phased array antenna based on the reconstruction results of the array surface structure, including calculating the element position offset according to the deformation of the array surface structure, deriving the phase difference between the elements, and then obtaining the radiation pattern function of the phased array antenna, and analyzing the characteristics of the radiation pattern function.

2. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 1 is characterized in that: In step S1, the finite element model is an Abaqus finite element model; In step S3, the strain sensor is a fiber Bragg grating strain sensor; a coordinate system of the phased array antenna array structure is established by a binocular vision sensor; In step S3, the load applied to the phased array antenna surface structure includes a thermal load and a concentrated force load.

3. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 1 is characterized in that: In step S4, the low-fidelity strain-deformation data obtained in step S2 is and the high-fidelity strain-deformation data obtained in step S3 Decompose according to the orthogonal relationship of the spatial position in the coordinate system to obtain the strain data set under the XY component 、 、 、 ;in, , , Represents the XY components of low-fidelity strain data A real matrix with M rows and N columns; Represents the XY components of high-fidelity strain data 、 A real matrix with P rows and Q columns.

4. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 1 is characterized in that: Also includes: Training the discriminator: Using a mini-batch strategy, we can train the discriminator from high-fidelity strain-deformation data. Extract real samples from the generator and input them into the discriminator to calculate the loss, and calculate the gradient through the back propagation algorithm; use the current generator to generate fake samples and input them into the discriminator to calculate the loss and back propagate again to calculate the gradient; accumulate the gradients of the two batches of samples and update the discriminator parameters; Training the generator: The generator is trained by maximizing the loss function to update the generator parameters so that the discriminator can identify the samples generated by the generator as real samples.

5. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 4 is characterized in that: The loss functions of the discriminator and generator are as follows: ； Where, is the predicted output of the generator model, ranging from (0,1); is the true label, the true sample is marked as 1, and the false sample is marked as 0; is the loss of a single sample; L is the set of losses for all samples.

6. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 1 is characterized in that: In step S8, the confrontation is continuously generated, and the deformation result of the phased array antenna surface structure is finally reconstructed to determine the number and distribution characteristics of the elements of the phased array antenna. The array elements are equally spaced rectangular grid plates, and the element spacing along the x-axis and y-axis directions is and .

7. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 6, characterized in that: Step S9 includes: Assume that any point in space ,point Relative to the coordinate system It can be expressed as direction cosines ,point The angle between the coordinate axis and the coordinate axis is expressed as: ； Set up the first Array Elements The position offset is , leading to the Array element relative to the The spatial phase difference of the unit is: ； Assuming that the deformation of the antenna surface only affects the electric field phase of the array unit and does not change its amplitude, we can finally get the first Unit relative to the The directional pattern function of the unit is: ； Where, For the The array pattern of the antenna elements; is the unit excitation current, The phase difference within the array of the unit is ; j is an imaginary number; is the propagation constant of free space, and the position deformations are 、 、 .

8. The method for high-precision reconstruction and electrical performance analysis of phased array antenna structure based on small sample data fusion according to claim 1 or 7, characterized in that: Step S9 further includes: Treat the deformation of the array as a random field variable and the deformation of the array as a random variable , for which antenna units, the deformation of the array is expressed as , the mean and variance are and ; The mean variance of the unit phase difference is derived based on the random error of the array unit deformation, and the mean variance of the pattern function of the entire array antenna at each observation point is calculated. ; get the first Array element relative to the The spatial phase difference of the unit is: ； make ; Then the pattern function of the phased array antenna is expressed as: ； Directional pattern function The value of changes with the unit position, by analyzing the unit deformation The mean and variance of the calculation unit position direction pattern function.

9. An electronic device, characterized in that: The electronic device includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the high-precision reconstruction and electrical performance analysis method of phased array antenna structure based on small sample data fusion as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that The storage medium stores at least one executable instruction, and when the executable instruction is executed on the electronic device, the electronic device executes the high-precision reconstruction and electrical performance analysis method of the phased array antenna structure based on small sample data fusion according to any one of claims 1 to 8.

Images