Radar multi-dimensional data dynamic edge weight graph construction method

Abstract

The invention discloses a radar multi-dimensional data dynamic edge weight graph construction method, and relates to the technical field of radar signal processing and graph data modeling, and the method comprises the steps: obtaining a radar observation sequence; the radar observation sequence comprises a plurality of nodes; the nodes correspond to time units in a radar observation sequence; performing feature extraction and feature fusion on the node feature vector of each node to obtain a plurality of fusion feature vectors; based on a plurality of nodes, constructing a dynamic edge and a corresponding dynamic edge weight; the dynamic edge comprises a time sequence adjacent edge, a feature similar edge and a self-loop edge; performing fusion normalization on the dynamic edge weights to obtain a dynamic edge weight matrix after fusion normalization; and constructing a graph structure based on the plurality of fusion feature vectors, the dynamic edge and the dynamic edge weight matrix. According to the method, through time sequence attenuation, a feature similarity dynamic threshold and an edge weight fusion mechanism, the modeling capability of radar target multi-dimensional features is improved.

Description

Technical Field

[0001] This application relates to the technical fields of radar signal processing and graph data modeling, and particularly relates to a method for constructing a dynamic edge weight graph of radar multi-dimensional data. Background Art

[0002] Radar target echo data has multi-dimensionality (including Doppler spectrum, azimuth angle, elevation angle, distance, speed, etc.) and non-linear time-varying characteristics. Traditional graph construction methods usually adopt fixed edge weights or single connection methods, and it is difficult to effectively capture the complex correlations between target dynamic features. Specific problems include: First, insufficient modeling of temporal correlations: Traditional temporal graph construction methods adopt fixed neighborhoods or full connection strategies, which cannot adapt to the non-stationary dynamic characteristics of targets, resulting in noise interference and lack of feature continuity; Second, neglect of feature similarity: Existing methods do not fully utilize the intra-class similarity between multi-dimensional features, resulting in insufficient association of highly correlated nodes across time steps; Third, rigid edge weight assignment: The fixed edge weight mechanism cannot dynamically adjust the importance between nodes, weakening the sensitivity of the model to micro-motion features.

[0003] Therefore, how to solve the problems of fixed edge weights and insufficient dynamic association modeling in the processing of multi-dimensional radar data by traditional graph construction methods has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0004] The purpose of this application is to provide a method for constructing a dynamic edge weight graph of radar multi-dimensional data, which can solve the problems of fixed edge weights and insufficient dynamic association modeling in the processing of multi-dimensional radar data by traditional graph construction methods. Through the temporal attenuation, dynamic threshold of feature similarity, and edge weight fusion mechanism, the modeling ability of radar target multi-dimensional features is improved.

[0005] To achieve the above purpose, this application provides the following solutions.

[0006] This application provides a method for constructing a dynamic edge weight graph of radar multi-dimensional data, and the method for constructing a dynamic edge weight graph of radar multi-dimensional data includes the following steps.

[0007] S1: Obtain a radar observation sequence; the radar observation sequence includes a number of nodes; the nodes correspond to time units in the radar observation sequence; the node feature vector of the node is jointly composed of Doppler spectrum and physical motion parameters.

[0008] S2: Perform feature extraction and feature fusion on the node feature vectors of each node to obtain a number of fused feature vectors.

[0009] S3: Based on a number of nodes, construct dynamic edges and corresponding dynamic edge weights; the dynamic edges include: temporal adjacency edges, feature similarity edges, and self-loop edges.

[0010] S4: fusing and normalizing the dynamic edge weights to obtain a fused and normalized dynamic edge weight matrix.

[0011] S5: constructing a graph structure based on the plurality of fused feature vectors, the dynamic edges and the dynamic edge weight matrix.

[0012] According to the specific embodiments provided in this application, this application discloses the following technical effects.

[0013] The present application provides a method for constructing a dynamic edge weight graph of radar multi-dimensional data. By acquiring a radar observation sequence, the radar observation sequence includes a plurality of nodes, performing feature extraction and feature fusion on the node feature vector of each node, and obtaining a plurality of fused feature vectors, more representative and critical information can be mined, redundant or relatively minor parts can be removed, and features of different sources and different properties but interrelated can be integrated together to form a more compact and more expressive feature representation; by constructing dynamic edges and corresponding dynamic edge weights based on a plurality of nodes, the dynamic edges include: time-series adjacent edges, feature-similar edges and self-loop edges, the sequential logic between nodes can be clarified from a time-series perspective, the feature-similar edges can be used to mine the implicit connection between nodes close to each other at the feature level, and the self-loop edges also take into account factors such as the state continuation of the node itself, and provide diversified connection information and weight basis for constructing a more expressive graph structure; by fusing and normalizing the dynamic edge weights, a fused and normalized dynamic edge weight matrix can be obtained, the relative importance of different edges can be reflected more objectively and accurately, and the problem of some edge information being ignored due to excessive differences in edge weight values ​​can be avoided. In this way, when the graph structure is subsequently constructed based on the edge weight matrix, the relationship information represented by each dynamic edge can be better integrated, so that the graph structure can more reasonably reflect the real and comprehensive relationship between nodes, and improve the expression ability and accuracy of the entire graph structure for the dynamic characteristics of the target; finally, based on a number of the fused feature vectors, the dynamic edges and the dynamic edge weight matrix, the graph structure is constructed. This application solves the problems of fixed edge weights and insufficient dynamic association modeling in traditional graph construction methods in multi-dimensional radar data processing. Through multi-stream feature extraction, the fusion design of temporal adjacent attenuation edges, feature similarity edges and self-loop edges, combined with dynamic edge weight allocation and normalization mechanism, adaptive modeling and efficient fusion of multi-dimensional features of radar targets are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 This is an application environment diagram of a method for constructing a dynamic edge weight graph of radar multi-dimensional data in an embodiment of the present application.

[0016] Figure 2 This is a schematic flowchart of a method for constructing a dynamic edge weight graph of radar multi-dimensional data provided in an embodiment of the present application.

[0017] Figure 3 This is a schematic diagram of the overall process of the method for constructing a dynamic edge weight graph of radar multi-dimensional data provided in an embodiment of the present application.

[0018] Figure 4 This is a schematic diagram of the curve of the time-series attenuation weight changing with distance provided in an embodiment of the present application.

[0019] Figure 5 This is a schematic diagram of the dynamic threshold of feature similarity edges provided in an embodiment of the present application.

[0020] Figure 6 This is a schematic diagram of the dynamic edge weight fusion normalization matrix provided in an embodiment of the present application.

[0021] Figure 7 This is a schematic diagram of the complete graph structure provided in an embodiment of the present application. Detailed implementation manners

[0022] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0023] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0024] The method for constructing a dynamic edge weight graph of radar multi-dimensional data provided in the embodiments of the present application can be applied to, for example Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send a radar observation sequence to the server 104, wherein the radar observation sequence includes a plurality of nodes; the nodes correspond to the time units in the radar observation sequence; the node feature vector of the node is jointly composed of the Doppler spectrum and the physical motion parameter; after the server 104 receives the radar observation sequence, for the radar observation sequence, the server 104 performs feature extraction and feature fusion on the node feature vector of each node to obtain a plurality of fused feature vectors; based on a plurality of nodes, dynamic edges and corresponding dynamic edge weights are constructed; the dynamic edges include: time-series adjacent edges, feature-similar edges and self-loop edges; the dynamic edge weights are fused and normalized to obtain a fused and normalized dynamic edge weight matrix; based on a plurality of the fused feature vectors, the dynamic edges and the dynamic edge weight matrix, a graph structure is constructed. The server 104 can feed back the obtained graph structure to the terminal 102. In addition, in some embodiments, the method for constructing a dynamic edge weight graph of multi-dimensional data can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly construct a dynamic edge weight graph of radar multi-dimensional data for the radar observation sequence, or the server 104 can obtain the radar observation sequence from the data storage system and construct a dynamic edge weight graph of radar multi-dimensional data for the radar observation sequence.

[0025] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, and tablet computers. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0026] In an exemplary embodiment, Figure 2 As shown, a method for constructing a dynamic edge weight graph of radar multidimensional data is provided. The method is executed by a computer device, and specifically can be executed by a computer device such as a terminal or a server alone, or can be executed by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used for explanation, and the steps include the following steps S1 to S5.

[0027] S1: Acquire a radar observation sequence; the radar observation sequence includes a plurality of nodes; the nodes correspond to time units in the radar observation sequence; and the node feature vector of the node is jointly formed by a Doppler spectrum and a physical motion parameter.

[0028] S2: Perform feature extraction and feature fusion on the node feature vector of each node to obtain several fused feature vectors.

[0029] S3: Based on a number of nodes, construct dynamic edges and corresponding dynamic edge weights; the dynamic edges include: temporal adjacent edges, feature similarity edges and self-loop edges.

[0030] S4: fusing and normalizing the dynamic edge weights to obtain a fused and normalized dynamic edge weight matrix.

[0031] S5: constructing a graph structure based on the plurality of fused feature vectors, the dynamic edges and the dynamic edge weight matrix.

[0032] By implementing the above steps S1 to S5, through feature extraction and fusion, the fusion design of temporal adjacent attenuation edges, feature similarity edges and self-loop edges, combined with dynamic edge weight allocation and normalization mechanism, adaptive modeling and efficient fusion of multi-dimensional features of radar targets are achieved. Compared with traditional methods, the graph data constructed by the dynamic edge weight mechanism of this application can significantly improve the target classification accuracy and noise resistance performance as the input of the graph convolutional network.

[0033] As an optional implementation, in step S2, feature extraction and feature fusion are performed on the node feature vector of each node to obtain a plurality of fused feature vectors, which specifically includes the following steps.

[0034] S21: Use a one-dimensional convolutional layer to extract features from the Doppler spectrum of each node to obtain local time-frequency features.

[0035] S22: Use a one-dimensional convolutional layer to extract features of the physical motion parameters of each node to obtain physical motion features.

[0036] S23: Concatenate the local time-frequency features and physical motion features corresponding to each node to obtain several fused feature vectors.

[0037] Specifically, the expression of the fused feature vector is as follows.

[0038] ; ; ; in, To extract the local time-frequency features of the Doppler spectrum through a one-dimensional convolutional layer; It is a one-dimensional convolutional layer with a kernel size of 3 and an output channel of 64; For the Doppler spectrum characteristics of each time unit; To map physical motion parameters through a one-dimensional convolutional layer; It is the physical motion characteristic; is the fused feature vector.

[0039] As an optional implementation, in step S3, constructing temporal adjacent edges and edge weights includes the following steps.

[0040] A1: local connections between nodes are established based on a time sliding window to obtain temporal adjacent edges; the time sliding window is adaptively adjusted according to the template acceleration change rate.

[0041] A2: Based on the temporal adjacent edges, a mixed attenuation function is used to determine the edge weight of the temporal adjacent edges; the mixed attenuation function is a function obtained by combining an exponential attenuation function with a Gaussian kernel function.

[0042] Specifically, the expression of the local connection between the nodes is as follows.

[0043] ; ; in, For the node; For the node; is the local connection between nodes; is the total number of time steps, which is equal to the total number of nodes; For the Window radius; is the base window radius; is the acceleration change rate coefficient; is the minimum value of the window radius; is the maximum value of the window radius; For the The acceleration of time.

[0044] The calculation formula of the temporal adjacent edge weight is as follows.

[0045] ; in, is the temporal adjacent edge weight; is the exponential decay weight coefficient; is the exponential decay rate; is the Gaussian kernel width; and is a constant.

[0046] As an optional implementation, in step S3, constructing feature similarity edges and edge weights includes the following steps.

[0047] B1: Based on the fused feature vector, calculate the cosine similarity between nodes.

[0048] B2: Calculate the cosine similarity of all pairs of similar nodes.

[0049] B3: The quantile of the cosine similarity of all similar node pairs is used as a dynamic threshold.

[0050] B4: The edges whose cosine similarity between the nodes is greater than the dynamic threshold are regarded as feature similarity edges.

[0051] B5: Based on the feature similarity edges, Z-score normalization is used to obtain the feature similarity edge weights.

[0052] Specifically, the calculation formula of the feature similarity edge weight is as follows.

[0053] ; in, The edge weights are similar in characteristics; is the mean similarity of nodes of the same type; is the standard deviation of similarity of nodes of the same type; is the weight distribution coefficient; For Node and The cosine similarity of .

[0054] As an optional implementation, in step S3, the construction of self-loop edges and edge weights includes the following steps.

[0055] C1: Self-connect several nodes to obtain self-loop edges.

[0056] C2: Based on the self-loop edge, add a fixed weight to each node to obtain the self-loop edge weight.

[0057] As an optional implementation, the expression of the fusion normalization is as follows.

[0058] ; ; in, is the edge weight matrix after fusion; is the temporal adjacent edge weight; The edge weights are similar in characteristics; is the secondary weight coefficient; is the dynamic edge weight matrix after fusion and normalization; is the sum index variable, which represents the sum of nodes Traverse all neighbor edge weights of .

[0059] In an exemplary embodiment, a method for constructing a dynamic edge weight graph of radar multi-dimensional data is provided. The overall flow chart is as follows: Figure 3 As shown, the following steps are included.

[0060] Step 1: Initialize the graph structure.

[0061] Modeling radar observation sequences as graph structures ,in is a node set, is the edge set, is the edge weight matrix. Each node Corresponding to the time unit in the radar observation sequence, its feature vector is composed of the Doppler spectrum and motion parameters (including azimuth, pitch angle, distance, speed, etc.).

[0062] ; in, For the Doppler spectrum characteristics of time units ; Include azimuth , Pitch angle ,distance ,speed , acceleration And other 5-dimensional physical motion characteristics.

[0063] Step 2: Feature extraction and fusion.

[0064] The Doppler flow and physical flow features are extracted separately through independent convolutional layers. A one-dimensional convolutional layer is used to process the Doppler spectrum and output local time-frequency features. A convolutional layer with the same structure is used to process physical motion parameters and output physical motion features. The dual-stream features are spliced ​​into fusion features to retain the complementarity of multimodal information.

[0065] The local time-frequency features of the Doppler spectrum are extracted through a one-dimensional convolutional layer.

[0066] ; in, To extract the local time-frequency features of the Doppler spectrum through a one-dimensional convolutional layer; It is a one-dimensional convolutional layer with a kernel size of 3 and an output channel of 64; For the Doppler spectrum characteristics of time units .

[0067] The physical motion parameters are mapped through a one-dimensional convolutional layer.

[0068] ; in, To map physical motion parameters through a one-dimensional convolutional layer; Characteristics of physical motion.

[0069] Concatenate the two-stream features into a fused feature vector.

[0070] ; in, is the fused feature vector.

[0071] Step 3: Dynamic edge weight construction.

[0072] Three types of edges are generated based on the target motion characteristics and feature similarity: (1) Temporal adjacent edges: The time window is dynamically adjusted according to the acceleration change rate, and weights are assigned to neighboring time nodes through a mixed attenuation function to balance local correlation and noise resistance. (2) Feature similarity edges: Node similarity is calculated based on fused features, and highly correlated nodes are screened by dynamic thresholds. Edge weights are weighted and assigned after standardization. (3) Self-loop edges: Each node adds a fixed weight self-connection to ensure that its own features are retained during information transmission.

[0073] (1) Construction of temporal adjacent edges and edge weights.

[0074] Establish local connections based on time sliding windows.

[0075] ; in, For the node; For the node; is the local connection between nodes; is the total number of time steps, which is equal to the total number of nodes; For the The window radius is adaptively adjusted according to the target acceleration change rate. The formula is as follows.

[0076] ; in, is the base window radius; is the acceleration change rate coefficient; For the The acceleration of the moment; It is the minimum value of the window radius and can be set according to actual conditions; It is the maximum value of the window radius and can be set according to actual conditions.

[0077] In order to balance local sensitivity and noise resistance, the temporal edge weight decays with the interval distance. The mixed decay function combines exponential decay and Gaussian kernel function.

[0078] ; in, is the temporal adjacent edge weight; is the exponential decay weight coefficient; is the exponential decay rate; is the Gaussian kernel width, which can be set according to the actual situation; and are constants. The curve of the temporal decay weight changing with distance is as shown in Figure 4 .

[0079] (2) Construction of feature-similar edges and edge weights.

[0080] To overcome the insufficient sensitivity of pure temporal connections to the subtle feature differences of the target, feature-similar edges based on the dynamic threshold of the intra-class similarity distribution are introduced, and the dynamic threshold strategy is adopted to determine the effective connections. Calculate the cosine similarity of all pairs of nodes in the same class, and take the 25th percentile as the dynamic threshold , and retain the edges where is greater than .

[0081] First, calculate the cosine similarity of nodes and and based on the fused feature vector

[0082] ; Calculate the cosine similarity of all pairs of nodes in the same class.

[0083] ; Take the percentile as the dynamic threshold , and the schematic diagram of the dynamic threshold of the feature-similar edge is as shown in Figure 5 .

[0084] ; Screen the effective connections through the dynamic threshold, that is, retain the edges where is greater than through the dynamic threshold.

[0085] ; Map the similarity to the weight through Z-score normalization.

[0086] ; Among them, is the mean of the similarities of nodes in the same class; is the standard deviation of the similarities of nodes in the same class; is the weight distribution coefficient, which can be set according to the actual situation; represents node and Nodes belonging to the same category of radar targets.

[0087] (3)Self-loop edge and edge weight construction.

[0088] Add self-loop edges with a fixed weight of 1.0 to each node to retain the characteristics of the node itself.

[0089] ; Step 4: Dynamic edge weight fusion normalization.

[0090] Perform weighted fusion on the temporal edges and feature edges, giving priority to retaining the information of the high-weight edges, and the secondary edge weights are superimposed proportionally. The fused edge weight matrix is processed by normalization to avoid numerical instability problems and ensure the reliability of the input data for subsequent tasks.

[0091] If there are both temporal edges and feature edges between nodes, a weighted summation strategy is adopted.

[0092] ; Among them, is the fused edge weight matrix; is the secondary weight coefficient, which is set according to the actual situation. The self-loop edges are retained independently and do not conflict with the other edges.

[0093] Perform L2 normalization on the fused edge weight matrix by row to ensure numerical stability, and the weight matrix distribution is shown in Figure 6.

[0094] ; Among them, is the dynamic edge weight matrix after fusion normalization; is the summation index variable, indicating that all neighbor edge weights of node are traversed (including temporal edges, feature similarity edges, and self-loop edges); The value of is not a parameter to be set, but all nodes connected to node are traversed (i.e., nodes that satisfy ). For example, if node has 10 neighbors (including self-loop edges), then

[0095] takes values from 1 to 10. By normalizing the edge weights by row (node dimension), it is ensured that the sum of the out-edge weights of each node is 1, avoiding numerical instability problems (such as gradient explosion).

[0096] Generate a graph structure containing node features and the dynamic edge weight matrix (such as Figure 7As shown in the figure), it serves as the input data for downstream tasks (such as classification and tracking). This graph structure provides a highly robust foundation for radar target analysis by adaptively modeling spatio-temporal correlations and multi-modal fusion.

[0097] The beneficial effects of a method for constructing a dynamic edge-weighted graph of radar target multi-dimensional data proposed in this application are as follows.

[0098] (1) Dynamic edge-weight mechanism: By mixing attenuation functions and dynamic threshold strategies, the temporal and feature edge weights are adaptively adjusted to enhance the ability to model the target's motion state.

[0099] (2) Multi-edge type fusion: Temporal edges capture local continuity, feature edges enhance cross-time-step correlations, and self-loop edges retain the characteristics of the nodes themselves.

[0100] (3) Feature extraction and fusion: Through the strategy of first extracting features and then splicing by a one-dimensional convolutional layer, the complementarity of Doppler spectra and physical parameters is fully exploited.

[0101] (4) Anti-noise performance optimization: The joint dynamic threshold screening and normalization strategies suppress noise interference and improve the robustness of the graph data.

[0102] This application also provides an application scenario that applies the above method for constructing a dynamic edge-weighted graph of radar multi-dimensional data. Specifically: The method for constructing a dynamic edge-weighted graph of radar multi-dimensional data provided in this embodiment can be applied in radar signal processing and graph data modeling scenarios. The radar signal processing and graph data modeling scenarios include: a data acquisition link, a feature extraction and feature fusion link, a dynamic edge and edge weight construction link, a fusion normalization link, and a graph structure construction link; First, a radar observation sequence is obtained; the radar observation sequence includes a number of nodes; the nodes correspond to time units in the radar observation sequence; the node feature vectors of the nodes are jointly composed of Doppler spectra and physical motion parameters; feature extraction and feature fusion are performed on the node feature vectors of each node to obtain a number of fused feature vectors; Secondly, based on a number of nodes, dynamic edges and corresponding dynamic edge weights are constructed; the dynamic edges include: temporal adjacent edges, feature similarity edges, and self-loop edges; Then, the dynamic edge weights are fused and normalized to obtain a fused and normalized dynamic edge weight matrix; Finally, based on a number of the fused feature vectors, the dynamic edges, and the dynamic edge weight matrix, a graph structure can be constructed.

[0103] 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 to be within the scope described in this specification.

[0104] In this article, specific examples are used to elaborate on the principles and implementation manners of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation on this application.

Claims

1. A method for constructing a dynamic edge weight graph of radar multidimensional data, characterized in that: The radar multi-dimensional data dynamic edge weight graph construction method comprises: Acquire a radar observation sequence; the radar observation sequence includes a plurality of nodes; the nodes correspond to time units in the radar observation sequence; a node feature vector of the node is jointly formed by a Doppler spectrum and a physical motion parameter; Perform feature extraction and feature fusion on the node feature vector of each node to obtain several fused feature vectors; Based on a number of nodes, dynamic edges and corresponding dynamic edge weights are constructed; the dynamic edges include: temporal adjacency edges, feature similarity edges and self-loop edges; The dynamic edge weights are fused and normalized to obtain a fused and normalized dynamic edge weight matrix; A graph structure is constructed based on a plurality of the fused feature vectors, the dynamic edges and the dynamic edge weight matrix.

2. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 1, characterized in that: Perform feature extraction and feature fusion on the node feature vector of each node to obtain several fused feature vectors, including: A one-dimensional convolutional layer is used to extract the features of the Doppler spectrum of each node to obtain local time-frequency features; A one-dimensional convolutional layer is used to extract the features of the physical motion parameters of each node to obtain the physical motion features; The local time-frequency features and physical motion features corresponding to each node are concatenated to obtain several fused feature vectors.

3. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 2, characterized in that: The expression of the fused feature vector is: ； ； ； in, To extract the local time-frequency features of the Doppler spectrum through a one-dimensional convolutional layer; It is a one-dimensional convolutional layer with a kernel size of 3 and an output channel of 64; For the Doppler spectrum characteristics of each time unit; To map physical motion parameters through a one-dimensional convolutional layer; It is a physical motion feature; is the fused feature vector.

4. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 1, characterized in that: Based on several nodes, dynamic edges and corresponding dynamic edge weights are constructed, including: Establishing local connections between nodes based on a time sliding window to obtain time-series adjacent edges; the time sliding window is adaptively adjusted according to the template acceleration change rate; Based on the time-series adjacent edges, a mixed attenuation function is used to determine the edge weights of the time-series adjacent edges; the mixed attenuation function is a function obtained by combining an exponential attenuation function with a Gaussian kernel function.

5. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 4, characterized in that: The expression of the local connection between the nodes is: ； ； in, For the node; For the node; is the local connection between nodes; is the total number of time steps, which is equal to the total number of nodes; For the Window radius; is the base window radius; is the acceleration change rate coefficient; is the minimum value of the window radius; is the maximum value of the window radius; For the The acceleration of time.

6. The method for constructing a dynamic edge weight graph of radar multidimensional data according to claim 4, characterized in that: The calculation formula of the temporal adjacent edge weight is: ； in, is the temporal adjacent edge weight; is the exponential decay weight coefficient; is the exponential decay rate; is the Gaussian kernel width; and is a constant.

7. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 1, characterized in that: Based on several nodes, dynamic edges and corresponding dynamic edge weights are constructed, including: Based on the fused feature vector, the cosine similarity between nodes is calculated; Calculate the cosine similarity of all similar node pairs; The quantile of cosine similarity of all similar node pairs is used as a dynamic threshold; The edges whose cosine similarity between the nodes is greater than the dynamic threshold are regarded as feature similarity edges; Based on the feature similarity edges, Z-score normalization is adopted to obtain the feature similarity edge weights.

8. The method for constructing a dynamic edge weight graph of radar multidimensional data according to claim 7, characterized in that: The calculation formula of the feature similarity edge weight is: ； in, The edge rights are similar in characteristics; is the mean similarity of nodes of the same type; is the standard deviation of similarity of nodes of the same type; is the weight distribution coefficient; For Node and The cosine similarity of .

9. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 1, characterized in that: Based on several nodes, dynamic edges and corresponding dynamic edge weights are constructed, including: Connect several nodes to obtain self-loop edges; Based on the self-loop edge, a fixed weight is added to each node to obtain the self-loop edge weight.

10. The radar multi-dimensional data dynamic edge weight graph construction method according to claim 1, characterized in that: The expression of the fusion normalization is: ； ； in, is the edge weight matrix after fusion; is the temporal adjacent edge weight; The edge rights are similar in characteristics; is the secondary weight coefficient; is the dynamic edge weight matrix after fusion and normalization; is the sum index variable, which represents the sum of nodes Traverse all neighbor edge weights of .

Images