Radar interference type identification and parameter estimation method, electronic equipment and storage medium

Abstract

The invention provides a radar interference type identification and parameter estimation method, electronic equipment and a storage medium, and the method comprises the steps: converting an obtained radar signal into a time-frequency spectrogram, extracting the amplitude information of the time-frequency spectrogram, carrying out the normalization processing of the amplitude information, carrying out the bicubic linear interpolation, adjusting the image proportion, and generating a pseudo-color image in a preset format. Thirdly, identifying the pseudo-color image and positioning an interference mode by using a YOLO target detection network, and extracting a corresponding image area as an input sample; then, the input samples are input into a deentanglement variational auto-encoder (DVAE) to optimize network parameters so as to maximize mutual information between different dimensions in hidden variables and minimize reconstruction errors, and target hidden variables are obtained. And finally, inputting the target hidden variable into a multi-layer perceptron (MLP) for decoding to obtain a specific interference parameter estimation value. The accuracy of interference type identification is improved, the accuracy of interference parameter estimation is enhanced, and the anti-interference capability of a radar system is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of radar interference type identification, and in particular to a radar interference type identification and parameter estimation method, electronic equipment and storage medium. Background Art

[0002] Radar interference parameter estimation is an important research direction in the field of radar signal processing, which aims to identify the interference type and further identify the interference parameters by analyzing the characteristic parameters of the interference signal.

[0003] Jiaxiang Zhang et al. proposed a composite interference recognition method based on deep learning object detection network to identify, detect individual interference elements, and estimate their key parameters. First, the time-frequency distribution (TFD) is used to characterize the characteristics of interference in multiple dimensions (time, frequency, and energy). Then, YOLOv5 is selected as the backbone network of the entire process because of its strong flexibility, fast processing speed, and high mean average precision (mAP). Then, the object detection network is applied to identify and locate interference in the time-frequency domain. Finally, according to the type of interference parameters, the corresponding parameter estimation method is given.

[0004] Based on the above method, the present invention further adjusts the module order of data preprocessing, improves the classification and positioning accuracy, adds feature disentanglement coding to facilitate downstream tasks, adds feature decoding by parameter estimation, and the parameters are given by the decoder instead of manual calculation, which improves the real-time performance of the estimation. Summary of the invention

[0005] In a first aspect of the present disclosure, a radar interference type identification and parameter estimation method is provided, comprising the following steps: S101: Obtaining an interference signal received by a radar and processing it, specifically including: performing a short-time Fourier transform on a one-dimensional complex signal of the interference signal to obtain a time-frequency spectrum, extracting amplitude information from the time-frequency spectrum, performing maximum and minimum normalization processing on the amplitude information to generate normalized amplitude features, using bicubic linear interpolation to adjust the image ratio of the normalized amplitude features to form an image in a preset format, and converting the image into a pseudo-color image; S102: Using a pre-trained YOLO target detection network to identify the pseudo-color image and locate the interference pattern, and output information including interference category, location, size and probability; S103: extracting the corresponding image of the information in the pseudo-color image as an input sample, encoding and decoding the input sample through a disentangled variational autoencoder, optimizing network parameters, and obtaining a target latent variable; S104: Input the target latent variable into a multi-layer perceptron and decode it to obtain a specific interference parameter estimation value.

[0006] In combination with the first aspect, the one-dimensional complex signal of the interference signal is subjected to short-time Fourier transform to obtain a time-frequency spectrum diagram calculated by the following formula: ， in, and Represent the time window and frequency index respectively, is the window function, is the sampling interval, represents discrete time points, is the complex exponential part of the Fourier transform, where is an imaginary unit, is the frequency index, is the sampling frequency, is a time variable.

[0007] In combination with the first aspect, the amplitude information extracted therefrom is calculated by the following formula: .

[0008] In combination with the first aspect, the maximum and minimum normalization processing of the amplitude information is calculated by the following formula: .

[0009] In combination with the first aspect, the image ratio is adjusted by using bicubic linear interpolation to adjust the normalized amplitude feature to form an image in a preset format, which is calculated by the following formula: ， in, and is relative to The offset of the point, , , are the cubic polynomial coefficients obtained by least squares fitting, and Represent the order of the polynomial respectively; The image is converted into a pseudo-color image by the following formula: , is an image in a preset format, where and Represent the row and column indices of the image, respectively. Represents the gray value of the corresponding pixel. is the output pseudo-color image matrix, represents the color channel, .

[0010] In combination with the first aspect, the using a pre-trained YOLO target detection network to identify the pseudo-color image and locate the interference pattern includes: The pseudo-color image is identified by the following formula: , in, is the total number of prediction boxes, is the true label, for each predicted box , if the box contains the target, then ,otherwise , For the The probability distribution of the prediction box output indicates the probability that the box belongs to a certain category; The interference pattern of the pseudo-color image is located by the following formula: , in, It is the ratio of the overlapping area between the predicted box and the true box to the total area of ​​the two boxes. is the Euclidean distance between the center point of the predicted box and the true box, is the diagonal distance between the closed areas of the two rectangular frames, is the weight coefficient, which is used to balance the importance of different items. Measures the consistency of the relative proportions of two rectangular boxes. , , , , Represent the width and height of the label and predicted anchor box respectively.

[0011] In combination with the first aspect, the disentangled variational autoencoder includes an encoder and a decoder. The encoder includes a first basic block, a second basic block, a third basic block and a fourth basic block, wherein the input and output channels of each basic block are increased in the order of the first basic block, the second basic block, the third basic block and the fourth basic block, and the input sample is input from the first basic block and output to the image convolution layer through the fourth basic block; The decoder includes a first negative basic block, a second negative basic block, a third negative basic block and a fourth negative basic block, wherein the first negative basic block has the least input channels and the most output channels, and the output channels of each basic block decrease in the order of the first basic block, the second basic block, the third basic block and the fourth basic block.

[0012] In combination with the first aspect, encoding and decoding the input sample by the disentangled variational autoencoder, optimizing the network parameters, and obtaining the target latent variable includes: S1031: According to the position information output by the YOLO target detection network, extract the corresponding area in the pseudo-color image as an input sample; S1032: Input the input sample to the encoder, reduce the spatial dimension of the input sample, increase the number of channels, and convert it into a target latent variable.

[0013] According to a second aspect of the present disclosure, there is provided an electronic device, comprising: one or more processors; A storage unit is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors can implement the radar interference type identification and parameter estimation method.

[0014] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, the radar interference type identification and parameter estimation method can be implemented.

[0015] Beneficial effect: The present invention provides a radar interference type identification and parameter estimation method, which converts the acquired one-dimensional complex radar signal into a time-frequency spectrum by short-time Fourier transform, extracts its amplitude information, and then performs maximum and minimum normalization processing and bicubic linear interpolation to adjust the image ratio to generate a pseudo-color image in a preset format. Then, the pre-trained YOLO target detection network is used to identify and locate the interference pattern of the pseudo-color image, and the corresponding image area is extracted as an input sample. Then, these input samples are input into the disentangled variational autoencoder (DVAE), and the target hidden variables with good disentanglement effect are obtained by optimizing the network parameters to maximize the mutual information between different dimensions within the hidden variables and minimize the reconstruction error. Finally, these target hidden variables are input into a multi-layer perceptron (MLP) for decoding to obtain specific interference parameter estimates. This method not only improves the accuracy of interference type identification, but also enhances the accuracy of interference parameter estimation, thereby effectively improving the anti-interference capability of the radar system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a flow chart of a radar interference type identification and parameter estimation method according to an embodiment of the present disclosure; Figure 2 An electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0017] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the embodiments of the present disclosure.

[0018] The terms used in the disclosed embodiments are only for the purpose of describing specific embodiments and are not intended to limit the disclosed embodiments. The singular forms of "a", "said" and "the" used in the disclosed embodiments and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0019] It should be understood that although the terms first, second, third, etc. may be used to describe various information in the disclosed embodiments, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the disclosed embodiments, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0020] like Figure 1 FIG. 1 is a flow chart of a radar interference type identification and parameter estimation method according to an embodiment of the present disclosure, including: S101: Obtaining the interference signal received by the radar and processing it, specifically including: performing short-time Fourier transform on the one-dimensional complex signal of the interference signal to obtain a time-frequency spectrum, extracting amplitude information from it, performing maximum and minimum normalization processing on the amplitude information to generate normalized amplitude features, using bicubic linear interpolation to adjust the image ratio of the normalized amplitude features to form an image in a preset format, and converting the image into a pseudo-color image.

[0021] Specifically, since radar data fluctuates greatly, and neural networks have poor fitting, generalization, and abstraction capabilities for data of larger magnitude, radar data needs to be compressed and converted, including: The short-time Fourier transform is performed on the one-dimensional complex signal of the interference signal to obtain a time-frequency spectrum diagram by the following formula: ， in, and Represent the time window and frequency index respectively. The complex signal received by the radar is J(t). is the window function, is the sampling interval, represents discrete time points, is the complex exponential part of the Fourier transform, where is an imaginary unit, is the frequency index, is the sampling frequency, is a time variable.

[0022] Then, the amplitude information is extracted from it using the following formula: .

[0023] This step is from Extract amplitude information from the result. For each time-frequency point, calculate the modulus (i.e., absolute value) of the corresponding complex value to obtain the amplitude of the point. Recorded as , which represents the energy of the signal at different times and frequencies.

[0024] Furthermore, the fluctuation of the interference amplitude information is large, and maximum and minimum normalization processing is required: , Beneficial effects: Normalization can help the model converge faster and improve its performance and stability. It also makes the amplitude information of radar signals more suitable for input into the deep learning model for further analysis.

[0025] Then, in order to balance the image aspect ratio so that the network can capture features, bicubic linear interpolation is used to resize the image: The image in the preset format is calculated by the following formula: ， in, and is relative to The offset of the point, , , are the cubic polynomial coefficients obtained by least squares fitting, and Represent the order of the polynomial respectively; The image is converted into a pseudo-color image by the following formula: , is an image in a preset format, where and Represent the row and column indices of the image, respectively. Represents the gray value of the corresponding pixel. is the output pseudo-color image matrix, represents the color channel, .

[0026] The images in the preset format include two-dimensional grayscale images.

[0027] Beneficial effects: The main purpose of bicubic interpolation is to provide smoother results than nearest neighbor interpolation or bilinear interpolation when resizing or scaling an image. It can better preserve image details and smoothness by considering more information from neighboring pixels and using a higher-order polynomial interpolation.

[0028] Furthermore, it is necessary to convert the above two-dimensional grayscale image into a pseudo-color image that can be processed by the deep learning model (YOLO target detection network).

[0029] The image is converted into a pseudo-color image by the following formula: , is an image in a preset format, where and Represent the row and column indices of the image, respectively. Represents the gray value of the corresponding pixel. is the output pseudo-color image matrix, represents the color channel, .

[0030] S102: Using a pre-trained YOLO target detection network to identify the pseudo-color image and locate the interference pattern, and output information including interference category, location, size and probability.

[0031] Specifically, the pseudo-color image is identified by the following formula: , in, is the total number of prediction boxes, is the true label, for each predicted box , if the box contains the target, then ,otherwise , For the The probability distribution of the prediction box output indicates the probability that the box belongs to a certain category; The above calculations are used to predict the category of the interference. For each predicted box, the model outputs a probability distribution indicating the likelihood that the interference belongs to each category.

[0032] The interference pattern of the pseudo-color image is located by the following formula: , in, It is the ratio of the overlapping area between the predicted box and the true box to the total area of ​​the two boxes. is the Euclidean distance between the center point of the predicted box and the true box, is the diagonal distance between the closed areas of the two rectangular frames, is the weight coefficient, which is used to balance the importance of different items. Measures the consistency of the relative proportions of two rectangular boxes. , , , , Represent the width and height of the label and predicted anchor box respectively.

[0033] The above calculations are used to optimize the position and size of the prediction box, which involves the regression problem of the center point coordinates, width, and height of the prediction box.

[0034] Finally, we get the interference category, horizontal and vertical coordinates, width, length, size information and corresponding probability marked by YOLO. .

[0035] S103: extracting the corresponding image of the information in the pseudo-color image as an input sample, encoding and decoding the input sample through a disentangled variational autoencoder, optimizing network parameters, and obtaining a target latent variable.

[0036] Specifically, the disentangled variational autoencoder includes an encoder and a decoder. The encoder includes a first basic block, a second basic block, a third basic block and a fourth basic block, wherein the input and output channels of each basic block are increased in the order of the first basic block, the second basic block, the third basic block and the fourth basic block, and the input sample is input from the first basic block and output to the image convolution layer through the fourth basic block. The decoder includes a first negative basic block, a second negative basic block, a third negative basic block and a fourth negative basic block, wherein the first negative basic block has the least input channels and the most output channels, and the output channels of each basic block decrease in the order of the first basic block, the second basic block, the third basic block and the fourth basic block.

[0037] Specifically, the encoder consists of several Basic_blocks with different parameters.

[0038] Exemplarily, the encoder includes a first basic block, a second basic block, a third basic block, and a fourth basic block, and the corresponding input channel numbers and output channel numbers are: (3, 32), (32, 64), (64, 128), (128, 256). Each Basic_block contains a convolution layer, a batch normalization layer, and a ReLU activation function. These layers gradually reduce the spatial dimension of the input sample while increasing the number of channels, and finally convert the input sample into a low-dimensional latent variable representation.

[0039] The decoder consists of several Inverse_Basic_blocks with different parameters.

[0040] Exemplarily, the decoder includes a first inverse basic block, a second inverse basic block, a third inverse basic block, and a fourth inverse basic block, and the corresponding input channel numbers and output channel numbers are: (10, 256), (256, 128), (128, 64), (64, 32). Each Inverse_Basic_block contains a transposed convolution layer, a batch normalization layer, and a ReLU activation function. These layers gradually restore the spatial dimensions of the input samples and finally reconstruct the input samples.

[0041] Exemplary, sample An example of ,The following introduces the derivation of the disentanglement and reconstruction loss of this network, A common variational lower bound is: , in, is a hidden variable, is the parameterized posterior distribution, The parameterized reconstruction distribution of for The prior for is usually a standard normal distribution, There are two distribution functions: and The KL (Kullback-Leibler) divergence of a continuous random variable is used to measure the asymmetric difference between two distribution functions. When the two distribution functions are the same, the KL divergence reaches its minimum value. . It is an artificially introduced distribution function, and different forms of distribution functions can be constructed according to different needs.

[0042] Above It is used to measure the difference between the approximate posterior probability distribution function and the prior distribution. This item can prevent overfitting; It is the log-likelihood estimate, which is equivalent to the reconstruction loss function in the autoencoder and is used to reconstruct the original sample data. can be seen as an encoding process, and It can be seen as a decoding process. Relative to a given sample The output encoding, that is, the data re-expression , This is equivalent to the reconstructed sample of the decoded output .

[0043] In order to further enhance the disentanglement effect, Let's break it down further. In network optimization, the KL divergence is usually calculated as the mean of the batch: , , , ， in Indicates The distribution of samples, Indicates The joint distribution of samples and their latent variables, for No. The distribution of components.

[0044] In the above formula, is the index encoding mutual information (MI), which is related to the mutual information between the data variables and the latent variables.

[0045] A higher MI indicates a better disentanglement effect.

[0046] is the total correlation (TC), which represents the degree of interdependence between variables in the latent variable space and is a measure of redundancy. Total correlation as a penalty forces the model to find statistical independence factors in the distribution. A heavier penalty for TC means stronger statistical independence of semantics in the posterior probability distribution, and enhanced disentanglement capabilities. When they are all independent, this item is 0, which is the most ideal de-entanglement effect.

[0047] The dimension representing each latent variable cannot deviate too much from the dimension of the prior latent variable. The weight of can further enhance the disentanglement effect.

[0048] but It is still impossible to estimate, so the total amount of data is set ,sampling Data, using techniques: , in express It is distributed according to the condition The distribution obtained by sampling.

[0049] In this way, the final loss of the network is obtained : , Finally, a hidden variable (target hidden variable) with better disentanglement effect is obtained through the above loss optimization network.

[0050] S104: Input the target latent variable into a multi-layer perceptron and decode it to obtain a specific interference parameter estimation value.

[0051] Specifically, according to the hidden variable obtained in step S103 , we directly input it into the multi-layer perceptron (MLP) and decode it to get the interference parameters.

[0052] Among them, Linear of Multi-Layer Perceptron (MLP) is a fully connected layer, in_features is the number of input features, out_features is the number of output features, ReLU is a linear activation unit, and Softplus is a soft plus function, which aims to adjust the output to a non-negative real number, corresponding to the non-negativity of the estimated parameters. n_features is the number of parameters to be estimated, which corresponds to the number of estimated parameters in interference modeling. Network design loss is the mean square error (MSE): , in and For the The actual interference parameters and the interference parameters estimated by the network, is the sample size.

[0053] The electronic device 200 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 200 may include, but is not limited to, a processor 201 and a memory 202. Those skilled in the art will appreciate that Figure 2 It is only an example of the electronic device 200 and does not constitute a limitation of the electronic device 200. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device may also include input and output devices, network access devices, buses, etc.

[0054] The processor 201 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0055] The memory 202 may be an internal storage unit of the electronic device 200, for example, a hard disk or memory of the electronic device 200. The memory 202 may also be an external storage device of the electronic device 200, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 200. Further, the memory 202 may also include both an internal storage unit of the electronic device 200 and an external storage device. The memory 202 is used to store the computer program 203 and other programs and data required by the electronic device. The memory 202 may also be used to temporarily store data that has been output or is to be output.

[0056] In the embodiments provided in the present disclosure, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms.

[0057] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, and the computer program code may be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.

[0058] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be included in the protection scope of the present disclosure.

Claims

1. A radar interference type identification and parameter estimation method, characterized in that: The following steps are involved: S101: Obtaining an interference signal received by a radar and processing it, specifically including: performing a short-time Fourier transform on a one-dimensional complex signal of the interference signal to obtain a time-frequency spectrum, extracting amplitude information from the time-frequency spectrum, performing maximum and minimum normalization processing on the amplitude information to generate normalized amplitude features, using bicubic linear interpolation to adjust the image ratio of the normalized amplitude features to form an image in a preset format, and converting the image into a pseudo-color image; S102: Using a pre-trained YOLO target detection network to identify the pseudo-color image and locate the interference pattern, and output information including interference category, location, size and probability; S103: extracting the corresponding image of the information in the pseudo-color image as an input sample, encoding and decoding the input sample through a disentangled variational autoencoder, optimizing network parameters, and obtaining a target latent variable; S104: Input the target latent variable into a multi-layer perceptron and decode it to obtain a specific interference parameter estimation value.

2. The method according to claim 1, characterized in that The short-time Fourier transform is performed on the one-dimensional complex signal of the interference signal to obtain a time-frequency spectrum diagram by calculating the following formula: ， in, and Represent the time window and frequency index respectively. The complex signal received by the radar is J(t). is the window function, is the sampling interval, represents discrete time points, is the complex exponential part of the Fourier transform, where is an imaginary unit, is the frequency index, is the sampling frequency, is a time variable.

3. The method according to claim 2, characterized in that The amplitude information extracted therefrom is calculated by the following formula: 。 4. The method according to claim 1, characterized in that: The maximum and minimum normalization processing of the amplitude information is calculated by the following formula: 。 5. The method according to claim 1, characterized in that The bicubic linear interpolation is used to adjust the image ratio of the normalized amplitude feature to form an image in a preset format, which is calculated by the following formula: ， in, and is relative to The offset of the point, , , are the cubic polynomial coefficients obtained by least squares fitting, and Represent the order of the polynomial respectively; The image is converted into a pseudo-color image by the following formula: ， is an image in a preset format, where and Represent the row and column indices of the image, respectively. Represents the gray value of the corresponding pixel. is the output pseudo-color image matrix, represents the color channel, .

6. The method according to claim 1, characterized in that The using of a pre-trained YOLO target detection network to identify the pseudo-color image and locate the interference pattern includes: The pseudo-color image is identified by the following formula: ， in, is the total number of prediction boxes, is the true label, for each predicted box , if the box contains the target, then ,otherwise , For the The probability distribution of the prediction box output indicates the probability that the box belongs to a certain category; The interference pattern of the pseudo-color image is located by the following formula: ， in, It is the ratio of the overlapping area between the predicted box and the true box to the total area of ​​the two boxes. is the Euclidean distance between the center point of the predicted box and the true box, is the diagonal distance between the two rectangular boxes. is the weight coefficient, which is used to balance the importance of different items. Measures the consistency of the relative proportions of two rectangular boxes. , , , , Represent the width and height of the label and predicted anchor box respectively.

7. The method according to claim 1, characterized in that The disentangled variational autoencoder includes an encoder and a decoder. The encoder includes a first basic block, a second basic block, a third basic block and a fourth basic block, wherein the input and output channels of each basic block are increased in the order of the first basic block, the second basic block, the third basic block and the fourth basic block. The input sample is input from the first basic block and output to the image convolution layer through the fourth basic block. The decoder includes a first negative basic block, a second negative basic block, a third negative basic block and a fourth negative basic block, wherein the first negative basic block has the least input channels and the most output channels, and the output channels of each basic block decrease in the order of the first basic block, the second basic block, the third basic block and the fourth basic block.

8. The method according to claim 7, characterized in that The encoding and decoding of the input sample by the disentangled variational autoencoder, optimizing the network parameters, and obtaining the target latent variable include: S1031: According to the position information output by the YOLO target detection network, extract the corresponding area in the pseudo-color image as an input sample; S1032: Input the input sample to the encoder, reduce the spatial dimension of the input sample, increase the number of channels, and convert it into a target latent variable.

9. An electronic device, characterized in that: include: one or more processors; A storage unit, used to store one or more programs, which, when executed by the one or more processors, enable the one or more processors to implement the radar interference type identification and parameter estimation method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it can implement the radar interference type identification and parameter estimation method according to any one of claims 1 to 8.

Images