Small sample HRRP radar target identification method, apparatus and device, and medium

Abstract

The invention relates to a small sample HRRP radar target identification method, device, equipment and medium, after normalization processing is carried out on a one-dimensional high-resolution range profile sequence of an unknown target with a small sample size, radar scattering physical prior is utilized to construct a static adjacency matrix, then a dynamic adjacency matrix is generated through a feature enhancement network, and the dynamic adjacency matrix is obtained. And after the static adjacent matrix and the dynamic adjacent matrix are fused, a hierarchical adaptive meta-learning graph neural network is adopted to carry out target identification training, so that high-precision identification of a position target is realized under data of fewer samples.

Description

Technical Field

[0001] The present application relates to the technical field of radar signal processing, and particularly to a small-sample HRRP radar target recognition method, device, equipment, and medium. Background Art

[0002] High-resolution range profile (HRRP) forms a one-dimensional feature signal by capturing the electromagnetic scattering center distribution of a target along the radar line of sight, and has advantages such as all-weather operation, strong anti-interference ability, and compact data structure, and is the core means of radar target recognition. Traditional methods such as template matching and Principal Component Analysis-Support Vector Machine (PCA-SVM) rely on artificial feature extraction and are difficult to handle the sensitivity of HRRP to target attitude angle, range migration, and noise.

[0003] However, although deep learning methods (such as convolutional neural networks and recurrent neural networks) have improved the feature expression ability, they face two major challenges in practical applications: one is that the dynamic migration of scattering centers caused by target attitude changes makes it difficult for static models to capture non-Euclidean space relationships; the other is that the labeled data of new target categories is scarce, and overfitting is likely to occur in conventional training. Summary of the Invention

[0004] Based on this, it is necessary to provide a small-sample HRRP radar target recognition method, device, equipment, and medium that can accurately recognize unknown targets under the premise of scarce samples for the above technical problems.

[0005] A small-sample HRRP radar target recognition method, the method includes: Obtain a small-sample HRRP data set about the same unknown target, where the small-sample HRRP data set includes a limited number of one-dimensional high-resolution range profile sequences, and perform normalization processing on each of the one-dimensional high-resolution range profile sequences to obtain a normalized sequence; For each of the normalized sequences, use the prior of radar scattering physics to construct a corresponding static adjacency matrix; Use a feature enhancement network to enhance the features of each of the normalized sequences to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix according to each of the enhanced feature matrices; After fusing the relevant dynamic adjacency matrix and static adjacency matrix, perform normalization processing to obtain a mixed adjacency matrix; Input each of the mixed adjacency matrices into a graph neural network in turn and perform forward propagation, and perform hierarchical adaptive meta-learning in each graph convolutional layer in the graph neural network to recognize the unknown target through the small-sample HRRP data set.

[0006] In one embodiment, when normalizing the one-dimensional high-resolution range image sequence, the following operations are performed on the samples at each range cell in the one-dimensional high-resolution range image sequence: ; In the above formula, represents the one-dimensional high-resolution range image sequence, where represents the number of range cells, represents the set of real numbers, , .

[0007] In one embodiment, the static adjacency matrix includes a plurality of elements, where the elements are represented as: ; In the above formula, the element represents the static correlation strength between each range cell and , represents the static mixing coefficient, represents the local neighborhood window size, represents the indicator function.

[0008] In one embodiment, the feature enhancement network includes three sequentially connected one-dimensional convolutional layers, and a batch normalization layer and a LeakyReLU activation function are connected after each one-dimensional convolutional layer; The first one-dimensional convolutional layer is provided with 16 filters, and the kernel size of each filter is 5, and the padding value is 2; The second one-dimensional convolutional layer is provided with 32 filters, and the kernel size of each filter is 3, and the padding value is 1; The third one-dimensional convolutional layer is provided with 64 filters, and the kernel size of each filter is 3, and the padding value is 1.

[0009] In one embodiment, when generating the dynamic adjacency matrix according to each enhanced feature matrix: Use the multi-head attention mechanism to calculate the query matrix, the key matrix, and the value matrix according to the enhanced feature matrix; Calculate the intra-head attention score matrix according to the query matrix and the key matrix, and then generate the dynamic adjacency matrix according to the intra-head attention score matrix.

[0010] In one embodiment, generating the dynamic adjacency matrix according to the intra-head attention score matrix is represented as: ; In the above formula, represents the intra-head attention score matrix, Represents the number of attention heads in the multi-head attention mechanism.

[0011] In one embodiment, the relevant dynamic adjacency matrix and static adjacency matrix are fused, expressed as: ; In the above formula, Represents the dynamic mixing coefficient, Represents the dynamic adjacency matrix, Represents the static adjacency matrix.

[0012] This application also provides a small-sample HRRP radar target recognition device, which includes: A small-sample dataset acquisition module, used to acquire a small-sample HRRP dataset about the same unknown target. The small-sample HRRP dataset includes a finite number of one-dimensional high-resolution range image sequences, and each of the one-dimensional high-resolution range image sequences is normalized to obtain a normalized sequence; A static adjacency matrix construction module, used to construct a corresponding static adjacency matrix for each of the normalized sequences by using the prior knowledge of radar scattering physics; A dynamic adjacency matrix generation module, used to perform feature enhancement on each of the normalized sequences by using a feature enhancement network to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix according to each of the enhanced feature matrices; A mixed adjacency matrix obtaining module, used to fuse the relevant dynamic adjacency matrix and static adjacency matrix and then perform normalization processing to obtain a mixed adjacency matrix; A small-sample unknown target recognition module, used to sequentially input each of the mixed adjacency matrices into a graph neural network and perform forward propagation, and perform hierarchical adaptive meta-learning in each graph convolutional layer in the graph neural network to recognize the unknown target through the small-sample HRRP dataset.

[0013] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned small-sample HRRP radar target recognition method are implemented.

[0014] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned small-sample HRRP radar target recognition method are implemented.

[0015] The above-mentioned small-sample HRRP radar target recognition method, device, equipment and medium first perform normalization processing on the one-dimensional high-resolution range image sequences of unknown targets with a small sample size, then construct a static adjacency matrix by using the prior knowledge of radar scattering physics, and then generate a dynamic adjacency matrix through a feature enhancement network. After fusing the static and dynamic adjacency matrices, a graph neural network with hierarchical adaptive meta-learning is used for target recognition training, so as to achieve high-precision recognition of position targets with less sample data. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic flowchart of the small-sample HRRP radar target recognition method in one embodiment; Figure 2 It is a schematic framework diagram of the small-sample HRRP radar target recognition method in one embodiment; Figure 3 It is a schematic diagram of the recognition performance comparison at different signal-to-noise ratios in a simulation experiment; Figure 4 It is a structural block diagram of the small-sample HRRP radar target recognition device in one embodiment; Figure 5 It is an internal structure diagram of a computer device in one embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0018] Aiming at the problems of insufficient dynamic scattering characteristic modeling, low meta-learning optimization efficiency, and poor noise robustness of the existing HRRP recognition method in the small-sample scenario, in the present application, as Figure 1 shown, a small-sample HRRP radar target recognition method is provided, including the following steps: Step S100: Obtain a small-sample HRRP data set about the same unknown target. The small-sample HRRP data set includes a limited number of one-dimensional high-resolution range image sequences, and perform normalization processing on each one-dimensional high-resolution range image sequence to obtain a normalized sequence.

[0019] Step S110: For each normalized sequence, construct a corresponding static adjacency matrix by using the prior knowledge of radar scattering physics.

[0020] Step S120: Use a feature enhancement network to enhance the features of each normalized sequence to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix according to each enhanced feature matrix.

[0021] Step S130: After fusing the relevant dynamic adjacency matrix and static adjacency matrix, perform normalization processing to obtain a hybrid adjacency matrix.

[0022] Step S140: Input each hybrid adjacency matrix into the graph neural network in sequence and perform forward propagation. Perform hierarchical adaptive meta-learning in each graph convolutional layer in the graph neural network to identify unknown targets through a small-sample HRRP dataset.

[0023] In step S100, obtain a small-sample HRRP dataset for the same unknown target. The small-sample HRRP dataset includes a finite number of one-dimensional high-resolution range image sequences. Here, 'finite number' or'small sample' specifically refers to the fact that compared with the large-scale datasets required for training traditional deep learning models, the number of available labeled samples for each target category is extremely limited. For example, in a single task of meta-learning, there may be only a single-digit or double-digit number of sample sequences, which is exactly the technical difficulty that this invention needs to solve.

[0024] In this embodiment, when performing normalization processing on the one-dimensional high-resolution range image sequences, the following operations are performed on the samples in each range cell of the one-dimensional high-resolution range image sequences: (1) In formula (1), represents the one-dimensional high-resolution range image sequence, where represents the number of range cells, represents the set of real numbers, , . The normalized signal is used as the input for subsequent steps.

[0025] In step S110, based on the prior knowledge of radar scattering physics, define the static adjacency matrix , the purpose of which is to incorporate the prior knowledge of radar scattering physics. The physical structure of the target determines that there are often inherent physical correlations among the range cells that are spatially adjacent. The static adjacency matrix aims to encode this relatively stable structural information based on physical distance and scattering mechanism, providing a basic graph structure with physical meaning for the model, which is particularly important in the small-sample scenario with sparse training data and can play a certain role in regularization and guidance. The static adjacency matrix includes multiple elements for representing the static correlation strength between the range cell and , expressed as: (2) In formula (2), the element represents each range cell and The static correlation strength between represents the static mixing coefficient, represents the local neighborhood window size, which is the maximum distance unit span of adjacent scattering centers. represents the indicator function, when outputs 1, otherwise 0.

[0026] In one embodiment, the static mixing coefficient is 0.5. The local neighborhood window size is 5.

[0027] Furthermore, in step S120, a corresponding dynamic adjacency matrix is generated based on the normalized sequence. First, a feature enhancement network is used to enhance the features of the normalized sequence, and then a multi-head attention mechanism is used to generate the dynamic adjacency matrix. The feature enhancement is to extract deeper, more robust, and more discriminative feature representations. The original one-dimensional HRRP sequence may contain noise and has relatively primitive features. Directly calculating the dynamic relationship based on the original sequence may not work well or be unstable enough. The enhanced features can better capture the internal patterns and connections between range cells, laying a foundation for accurately generating the dynamic adjacency matrix later. For the dynamic adjacency matrix, it should be emphasized that its purpose is to capture the dynamic scattering characteristics that change with observation conditions (such as target pose angle, pitch angle) in HRRP data and the complex relationships inherent in the data itself. The static matrix is based on fixed physical priors and is difficult to fully adapt to the dynamic changes such as the migration of target scattering centers with the viewing angle. By using the enhanced feature sequence and combining with the multi-head attention mechanism, the dynamic adjacency matrix can adaptively learn the correlation strength between range cells according to the features of each HRRP sample, dynamically adjust the connection relationship of the graph, and thus more flexibly and accurately model the dynamic characteristics of the target, making up for the deficiencies of the static graph in this regard.

[0028] In this embodiment, the feature enhancement network includes three consecutive one-dimensional convolutional layers (Conv1D), and a batch normalization layer (Batch Normalization, BN) and a LeakyReLU activation function (the negative slope can be set to 0.2) are connected after each one-dimensional convolutional layer. Among them, the first one-dimensional convolutional layer is set with 16 filters, and the kernel size of each filter is 5, and the padding value is 2; the second one-dimensional convolutional layer is set with 32 filters, and the kernel size of each filter is 3, and the padding value is 1; the third one-dimensional convolutional layer is set with 64 filters, and the kernel size of each filter is 3, and the padding value is 1. After being processed by the feature enhancement network, the output enhanced feature matrix , where, in is the hidden layer dimension (default value 64).

[0029] Further, when generating the dynamic adjacency matrix based on each enhanced feature matrix: the multi-head attention mechanism is used to calculate the query matrix, key matrix, and value matrix according to the enhanced feature matrix, calculate the intra-head attention score matrix according to the query matrix and the key matrix, and then generate the dynamic adjacency matrix according to the intra-head attention score matrix.

[0030] Specifically, using the multi-head attention mechanism, set the number of attention heads , the dimension of each head ; for each head , calculate the query matrix , key matrix , value matrix : (3) In formula (3), is a trainable parameter.

[0031] Further, calculate the intra-head attention score matrix , expressed as: (4) Then, generate the dynamic adjacency matrix according to the intra-head attention score matrix, expressed as: (5) In formula (5), represents the intra-head attention score matrix, represents the number of attention heads in the multi-head attention mechanism.

[0032] In step S130, fuse the dynamic adjacency matrix and the static adjacency matrix corresponding to the same one-dimensional high-resolution range image sequence, expressed as: (6) In the above formula, represents the dynamic mixing coefficient, represents the dynamic adjacency matrix, represents the static adjacency matrix. Among them, the dynamic mixing coefficient is set to 0.3.

[0033] Further, normalize the fused mixed graph to obtain the mixed adjacency matrix, expressed as: (7) In formula (7), represents the degree matrix, and the diagonal element .

[0034] In step S140, the hybrid adjacency matrix is input into the graph neural network, and hierarchical adaptive meta-learning is performed. Under the training of the hybrid adjacency matrix composed of a small number of one-dimensional high-resolution range image sequences, the final unknown target recognition result is obtained.

[0035] In this embodiment, several layers of graph convolutions are stacked in the graph neural network to update the node features in each layer , and the process is expressed as: (8) In formula (8), represents the LeakyReLU activation function, and its negative slope is set to 0.2. is a trainable weight. Dropout is a regularization technique that randomly sets the outputs of some neurons to zero with probability p during the training process to prevent the model from overfitting on small-sample data and enhance the generalization ability of the model. This probability p is usually set empirically, for example, it can be taken as 0.1, but it can also be adjusted according to specific applications and network structures.

[0036] Furthermore, the attention weights are calculated and the global features are aggregated , and the process is expressed as: (9) (10) (11) In formulas (9) to (11), both e and h represent the aggregated global features, where h represents each dimensional component, and the subscript i represents the i-th , , , are trainable parameters.

[0037] In this embodiment, in the graph neural network, hierarchical adaptive meta-learning is performed. During the meta-training stage, first, the internal loop task adaptation is carried out, and the support set i.e., the training set, is input, where is the number of categories, is the number of samples per category, represents a one-dimensional high-resolution range image sequence or its initial feature representation, is its corresponding target category label. The hybrid adjacency matrix is based on the input sequence Constructed through static and dynamic mechanisms. When identifying an unknown target (i.e., meta-testing), the model first loads the parameters obtained from meta-training, then quickly fine-tunes and adapts the model parameters using a small number of support samples provided by the current task, and finally classifies the query samples using the adapted model. Among them, the execution steps of gradient descent, and the step updates the parameters : (12) In formula (12), is the layer-specific learning rate, which is set according to the layer type. In this embodiment, the layer-specific learning rate of the graph convolutional layer is set to 0.004 - 0.006, and the layer-specific learning rate of the feature extraction layer is set to 0.002 - 0.004. represents the Hadamard product.

[0038] Furthermore, perform external loop meta-update, calculate the weighted multi-step loss on the query set , which is expressed as: (13) In formula (13), the weight . Use the Adam optimizer to update the meta-parameters, with the initial learning rate , and adopt the cosine annealing schedule. Finally, perform classification prediction testing, and calculate the class probability for the query sample : (14) In formula (14), , are the parameters of the classification layer.

[0039] After going through the meta-training stage, the trained graph neural network can be used to perform online hierarchical adaptive meta-learning based on the mixed adjacency matrix corresponding to the one-dimensional high-resolution range image sequence of small samples to identify the category of the unknown target.

[0040] In this embodiment, all the parameters of the feature enhancement network and the graph neural network part are jointly optimized and trained end-to-end through the above meta-learning framework.

[0041] In another embodiment, the schematic diagram of the process framework of the small sample unknown target recognition method is as Figure 3 shown. Among them, further divide this method into two stages. One stage is the generation stage of the mixed adjacency graph, and the second stage is meta-training and meta-testing, that is, the unknown target recognition stage.

[0042] To verify the performance of the method of this application, a simulation experiment is used for illustration. In this simulation experiment, the hyperparameter settings of the graph structure involved include , , . In terms of the network structure setting, graph convolutional layers are set, the dimension of the hidden layer is , rate is 0.1. In terms of the meta-learning strategy, the number of steps in the inner loop is , the number of task batches is , the initial learning rate is , and the cosine annealing schedule is adopted. The specific training and testing processes are as follows: 1) Meta-training stage: Sample 3 types of 5-sample tasks from the training set. Each task contains a support set and a query set, and the meta-parameters are updated after quickly adapting through the inner loop; 2) Meta-testing stage: Evaluate the model on unseen test categories, and perform classification after fine-tuning with only a small number of support samples. The dataset is based on an electromagnetic simulation software, and the X-band is selected for simulation. The radar signal bandwidth is set to , and the step size of the step is ; the azimuth angle range is , and the step size is ; the elevation angle range is , and the step size is . Six different aviation target models are selected for simulation to obtain an HRRP dataset with a size of , and it is randomly divided at a ratio of into categories to obtain a training set and a test set.

[0043] As Figure 3 shows, the results of the simulation experiment in the test under a signal-to-noise ratio (SNR) of 20 dB to -5 dB indicate that the present invention still maintains an accuracy of 59.4% at a very low SNR (-5 dB), which is 15.2% higher than the optimal comparison model. The comparison methods include MAML, MAML++, template matching, PCA-SVM, GCN, GAT, HRRPGraphNet, and CNN. The accuracy is the average accuracy on 600 query samples of 3-way 1-shot test tasks.

[0044] In the above small-sample HRRP radar target recognition method, first, the obtained HRRP data is preprocessed and normalized. Secondly, a static adjacency matrix is constructed. Thirdly, a dynamic adjacency matrix is generated. Then, a hybrid graph is generated and normalized. Furthermore, forward propagation is completed in the graph neural network. Finally, hierarchical adaptive meta-learning is performed. Among them, the generation and normalization of the static, dynamic, and hybrid graphs are the HRRP hybrid graph generation stage. The forward propagation of the graph neural network generates recognition results, and meta-learning helps the network obtain optimal initial parameters and optimize them on the support samples. This method synchronously models the physical association and data-driven relationship of range cells through the hybrid graph structure, improving the adaptability to pose changes and noise interference. At the same time, a hierarchical specific meta-learning optimization strategy is designed to accelerate the rapid adaptation of the model to new target categories, so that this method can still maintain high recognition accuracy in a harsh environment where the Signal-to-Noise Ratio (SNR) is lower than 0 dB.

[0045] It should be understood that although Figure 1 the steps in the flowchart of Figure 1 are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover,

[0046] In one embodiment, as Figure 4 shown, a small-sample HRRP radar target recognition device is provided, including: a small-sample data set acquisition module 200, a static adjacency matrix construction module 210, a dynamic adjacency matrix generation module 220, a hybrid adjacency matrix obtaining module 230, and a small-sample unknown target recognition module 240, where: The small-sample data set acquisition module 200 is used to acquire a small-sample HRRP data set about the same unknown target. The small-sample HRRP data set includes a finite number of one-dimensional high-resolution range profile sequences, and each of the one-dimensional high-resolution range profile sequences is normalized to obtain a normalized sequence; The static adjacency matrix construction module 210 is used to construct a corresponding static adjacency matrix for each of the normalized sequences by using the physical prior of radar scattering; A dynamic adjacency matrix generation module 220 is used to perform feature enhancement on each of the normalized sequences using a feature enhancement network to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix based on each of the enhanced feature matrices; A mixed adjacency matrix obtaining module 230 is used to fuse the relevant dynamic adjacency matrix and static adjacency matrix, and then perform normalization processing to obtain a mixed adjacency matrix; The small sample unknown target recognition module 240 is used to input each of the mixed adjacency matrices into the graph neural network in sequence, and forward propagate, and perform hierarchical adaptive meta-learning in each graph convolution layer in the graph neural network to identify the unknown target through the small sample HRRP dataset.

[0047] For the specific definition of the small sample HRRP radar target recognition device, please refer to the definition of the small sample HRRP radar target recognition method above, which will not be repeated here. Each module in the above-mentioned small sample HRRP radar target recognition device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0048] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a small sample HRRP radar target recognition method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0049] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0050] In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented: Obtain a small-sample HRRP data set about the same unknown target. The small-sample HRRP data set includes a finite number of one-dimensional high-resolution range image sequences. Perform normalization processing on each of the one-dimensional high-resolution range image sequences to obtain a normalized sequence; For each of the normalized sequences, use the prior knowledge of radar scattering physics to construct a corresponding static adjacency matrix; Use a feature enhancement network to enhance the features of each of the normalized sequences to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix according to each of the enhanced feature matrices; After fusing the relevant dynamic adjacency matrix and static adjacency matrix, perform normalization processing to obtain a mixed adjacency matrix; Input each of the mixed adjacency matrices into a graph neural network in sequence and perform forward propagation. Perform hierarchical adaptive meta-learning in each graph convolutional layer in the graph neural network to identify the unknown target through the small-sample HRRP data set.

[0051] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: Obtain a small-sample HRRP data set about the same unknown target. The small-sample HRRP data set includes a finite number of one-dimensional high-resolution range image sequences. Perform normalization processing on each of the one-dimensional high-resolution range image sequences to obtain a normalized sequence; For each of the normalized sequences, use the prior knowledge of radar scattering physics to construct a corresponding static adjacency matrix; Use a feature enhancement network to enhance the features of each of the normalized sequences to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix according to each of the enhanced feature matrices; After fusing the relevant dynamic adjacency matrix and static adjacency matrix, perform normalization processing to obtain a mixed adjacency matrix; Input each of the mixed adjacency matrices into a graph neural network in sequence and perform forward propagation. Perform hierarchical adaptive meta-learning in each graph convolutional layer in the graph neural network to identify the unknown target through the small-sample HRRP data set.

[0052] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0053] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0054] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A small sample HRRP radar target recognition method, characterized in that: The method comprises: Acquire a small sample HRRP data set about the same unknown target, wherein the small sample HRRP data set includes a limited number of one-dimensional high-resolution range image sequences, and perform normalization processing on each of the one-dimensional high-resolution range image sequences to obtain a normalized sequence; For each of the normalized sequences, using radar scattering physical priors, constructing a corresponding static adjacency matrix; Using a feature enhancement network to perform feature enhancement on each of the normalized sequences to obtain an enhanced feature matrix, and then generating a dynamic adjacency matrix based on each of the enhanced feature matrices; After fusing the relevant dynamic adjacency matrix and static adjacency matrix, normalizing them to obtain a mixed adjacency matrix; Each of the mixed adjacency matrices is sequentially input into the graph neural network and forward propagated, and hierarchical adaptive meta-learning is performed in each graph convolutional layer in the graph neural network to identify the unknown target through a small sample HRRP dataset.

2. The small sample HRRP radar target recognition method according to claim 1, characterized in that: When normalizing the one-dimensional high-resolution range image sequence, the following operations are performed on samples in each range unit in the one-dimensional high-resolution range image sequence: In the above formula, represents a one-dimensional high-resolution range image sequence, where Indicates the number of distance units, represents the set of real numbers, , .

3. The small sample HRRP radar target recognition method according to claim 2 is characterized in that: The static adjacency matrix includes multiple elements, where the elements are represented as: In the above formula, the element Represents each distance unit and The static correlation strength between represents the static mixing coefficient, represents the local neighborhood window size, Represents the indicator function.

4. The small sample HRRP radar target recognition method according to claim 3 is characterized in that: The feature enhancement network includes three one-dimensional convolutional layers connected in sequence, and each of the one-dimensional convolutional layers is connected with a batch normalization layer and a LeakyReLU activation function; The first one-dimensional convolutional layer has 16 filters, and the kernel size of each filter is 5 and the padding value is 2; The second one-dimensional convolutional layer is set with 32 filters, and the kernel size of each filter is 3 and the padding value is 1; The third one-dimensional convolutional layer is set with 64 filters, and the kernel size of each filter is 3 and the padding value is 1.

5. The small sample HRRP radar target recognition method according to claim 4, characterized in that: When generating a dynamic adjacency matrix according to each of the enhanced feature matrices: Calculate a query matrix, a key matrix, and a value matrix based on the enhanced feature matrix using a multi-head attention mechanism; The intra-head attention score matrix is ​​calculated according to the query matrix and the key matrix, and then the dynamic adjacency matrix is ​​generated according to the intra-head attention score matrix.

6. The small sample HRRP radar target recognition method according to claim 5, characterized in that: The dynamic adjacency matrix is ​​generated according to the intra-head attention score matrix, which is expressed as: In the above formula, represents the intra-head attention score matrix, Represents the number of attention heads in the multi-head attention mechanism.

7. The small sample HRRP radar target recognition method according to claim 6, characterized in that: The relevant dynamic adjacency matrix and static adjacency matrix are fused and expressed as: In the above formula, represents the dynamic mixing coefficient, represents the dynamic adjacency matrix, Represents the static adjacency matrix.

8. A small sample HRRP radar target recognition device, characterized in that: The device comprises: A small sample data set acquisition module is used to acquire a small sample HRRP data set about the same unknown target, wherein the small sample HRRP data set includes a limited number of one-dimensional high-resolution range image sequences, and each of the one-dimensional high-resolution range image sequences is normalized to obtain a normalized sequence; A static adjacency matrix construction module is used to construct a corresponding static adjacency matrix for each of the normalized sequences using radar scattering physical priors; A dynamic adjacency matrix generation module, used to perform feature enhancement on each of the normalized sequences using a feature enhancement network to obtain an enhanced feature matrix, and then generate a dynamic adjacency matrix based on each of the enhanced feature matrices; A mixed adjacency matrix obtaining module is used to fuse the relevant dynamic adjacency matrix and static adjacency matrix, and then perform normalization processing to obtain a mixed adjacency matrix; The small sample unknown target recognition module is used to input each of the mixed adjacency matrices into the graph neural network in sequence, and forward propagate, and perform hierarchical adaptive meta-learning in each graph convolution layer in the graph neural network to identify the unknown target through the small sample HRRP dataset.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images