A multi-source information mapping method for radar signal processing system

Abstract

The present invention proposes a multi-source information mapping method for a radar signal processing system, which relates to the technical field of radar signal processing. The method comprises the following steps: acquiring radar echo data, performing feature extraction on the radar echo data to obtain radar feature data; constructing a radar echo feature matrix at each moment using the radar feature data; creating a radar point cloud image, mapping nodes of the radar point cloud image to elements of the radar echo feature matrix through a quantized mapping matrix; establishing edge connections between nodes in the radar point cloud image, performing target position continuity constraints, improving the accuracy of target tracking and positioning, enabling the system to more stably track the target's motion trajectory, and more accurately estimating the target's state.

Description

Technical Field

[0001] The present invention relates to the technical field of radar signal processing, and in particular to a multi-source information mapping method for a radar signal processing system. Background Art

[0002] With the advancement of technology, radar has been widely used in many fields, including target detection, tracking and identification, air traffic control, weather monitoring, and autonomous driving. As application scenarios continue to expand and the requirements for radar performance continue to increase, more advanced signal processing methods are needed to improve the performance of radar systems.

[0003] Modern radar systems often acquire data from multiple types and sources. Effectively fusing and processing this multi-source information is an important research topic. Mapping radar echo data into point cloud images and incorporating relevant constraints is a specific implementation of multi-source information fusion in radar signal processing. This helps leverage the strengths of different data types and improve the radar system's target perception capabilities. Image analysis and processing technology has rapidly advanced in the field of computer vision. For example, deep learning algorithms have achieved remarkable results in image recognition and target detection. Converting radar data into image form can leverage these mature image analysis techniques to provide new insights and methods for radar signal processing, thereby improving the intelligence and performance of radar systems.

[0004] However, existing technologies often contain a large amount of redundant information, making direct processing inefficient and potentially hindering the accurate extraction and analysis of target features. Furthermore, traditional radar data representation methods hinder the use of mature techniques such as image analysis. In particular, data discontinuity or a lack of positional continuity constraints can lead to tracking instability, false detection, and misidentification. A multi-source information mapping method for radar signal processing systems can effectively address these issues and improve the stability and accuracy of data recognition. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention proposes a multi-source information mapping method for a radar signal processing system, comprising the following steps:

[0006] S1, obtaining radar echo data, performing feature extraction on the radar echo data to obtain radar feature data;

[0007] S2. constructing a radar echo feature matrix at each moment using radar feature data;

[0008] S3. Create a radar point cloud image, and map the nodes of the radar point cloud image to the elements of the radar echo feature matrix through a quantized mapping matrix;

[0009] S4. Perform target position continuity constraints and establish temporal edge connections between corresponding nodes at two moments in the radar point cloud image.

[0010] In a preferred embodiment, in step S2, each element R of the radar echo feature matrix R IJ for:

[0011] R IJ =A IJ +jP IJ ;

[0012] Among them, I and J represent the row number and column number of the matrix respectively, A IJ Indicates element status data, P IJ Represents radar phase data, and j is an imaginary unit.

[0013] In a preferred embodiment, the total number of rows N and the total number of columns M of the radar echo feature matrix R are determined according to the resolution and range of the radar data:

[0014] ;

[0015] ;

[0016] Among them, D range and D azimuth Respectively represent the maximum detection range and azimuth coverage of the radar, ΔD range and ΔD azimuth They represent the range resolution and azimuth resolution of radar data respectively.

[0017] In a preferred embodiment, the element status data A IJ is the three-dimensional coordinate and velocity vector The concatenated six-dimensional vector is expressed as: ;

[0018] Corresponding to the velocities in the x, y, and z directions respectively.

[0019] In a preferred embodiment, the elements S(a, b) of the quantization mapping matrix S are:

[0020] Among them, a and b are the row and column indices used to index the elements of the radar echo feature matrix, represents the azimuth of the radar echo data point, Indicates the quantization step size of the azimuth angle, Indicates rounding down;

[0021] For each radar echo feature matrix element state data A IJThe location of the node in the corresponding radar point cloud image Calculated by the following formula:

[0022] ;

[0023] ;

[0024] Where: M' and N' are the total number of rows and columns of nodes in the radar point cloud image.

[0025] In a preferred embodiment, the radar echo feature matrix covers an area of ​​L radians, and the image covers an area of ​​L′ radians. The radar data is scaled to the image coverage and then quantized. The quantization mapping matrix element S(a, b) becomes :

[0026] .

[0027] In a preferred embodiment, in step S4, the position continuity constraint of the target is established in combination with the speed information:

[0028] Let the time interval be , in the x direction, by the velocity and displacement Relationship:

[0029] ;

[0030] In the y direction, the velocity and displacement Relationship:

[0031] ;

[0032] In the z direction, the velocity and displacement Relationship:

[0033] ;

[0034] Establish target position continuity constraints:

[0035] Define an error threshold , the target position is considered to be continuous when the following conditions are met:

[0036] ;

[0037] If this condition is met, a temporal edge connection is established between the corresponding nodes at two moments in the radar point cloud image.

[0038] Compared with the prior art, the present invention has the following beneficial technical effects:

[0039] 1. Obtain radar echo data and perform feature extraction: By extracting features from radar echo data to obtain radar feature data, it is possible to extract useful information for tasks such as target detection and recognition from complex echo signals, remove redundant information, reduce data dimensions, and improve the efficiency and accuracy of subsequent processing.

[0040] 2. Construct a radar echo feature matrix: Use radar feature data to construct a radar echo feature matrix at each moment. Organizing the feature data in a matrix form facilitates mathematical operations and processing, provides a structured data representation for subsequent analysis and mapping, and helps to mine the spatiotemporal features and relationships in the data.

[0041] 3. Create radar point cloud images and perform quantitative mapping: This creates a radar point cloud image and maps nodes to radar echo feature matrix elements using a quantitative mapping matrix. This converts radar data from abstract numerical form to intuitive image form, facilitating the use of image analysis and processing techniques for target detection and recognition. The quantitative mapping process discretizes continuous radar data to meet the requirements of image representation and processing while preserving the data's key characteristics.

[0042] 4. Establishing edge connections between nodes and enforcing target position continuity constraints: Establishing edge connections between nodes can reflect the spatial relationships or correlations between different nodes in the radar point cloud image, helping to better understand and analyze the structure and distribution of targets. Enforcing target position continuity constraints can leverage the target's temporal and spatial continuity information to reduce false detections and misidentifications, improve target tracking and positioning accuracy, enable the system to more stably track the target's trajectory, and more accurately estimate the target's state. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 The flowchart of the multi-source information mapping method for a radar signal processing system of the present invention is shown in FIG.

[0045] Figure 2 This is a matching diagram of the points after correct mapping.

[0046] Figure 3 Schematic diagram of the structure of the multi-source information mapping system for radar signal processing system of the present invention.

[0047] Figure 4 This is a display diagram of radar monitoring data of the present invention. DETAILED DESCRIPTION

[0048] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0049] In the drawings of the specific embodiments of the present invention, in order to better and more clearly describe the working principles of the various components in the system, the connection relationship of the various parts in the device is shown, which only clearly distinguishes the relative position relationship between the various components, and does not constitute a limitation on the signal transmission direction, connection sequence and size, dimension and shape of the components or structures.

[0050] Example 1

[0051] like Figure 1 FIG. 1 is a flow chart of a multi-source information mapping method for a radar signal processing system according to the present invention. The multi-source information mapping method includes the following steps:

[0052] S1. Acquire radar echo data and perform feature extraction on the radar echo data to obtain radar feature data.

[0053] De-noising radar echo data acquired from multiple sensors removes noise interference to improve data quality. Data calibration is performed to ensure consistency and comparability between different sensors, such as calibration time and spatial coordinates.

[0054] The radar echo data is subjected to feature extraction to obtain radar feature data, including radar amplitude data, radar phase data, and target point speed.

[0055] S2. Use radar feature data to construct the radar echo feature matrix at each moment.

[0056] Each element R of the radar echo feature matrix R IJ for:

[0057] R IJ =A IJ +jP IJ Among them, I and J represent the row number and column number of the matrix respectively, A IJ Indicates element status data, P IJ Represents radar phase data, and j is an imaginary unit.

[0058] P IJ It is obtained through the raw measurement data of the radar system and comes directly from the output of the phase data channel of the radar receiver.

[0059] The total number of rows N and columns M of the radar echo feature matrix R is determined according to the resolution and range of the radar data:

[0060] ;

[0061] ;

[0062] Among them, D range Indicates the maximum detection range of the radar, that is, the maximum range of the target that the radar can detect, ΔD range It is the range resolution of radar data, which refers to the minimum distance interval at which the radar can distinguish two targets that are close in distance. The number of rows in the matrix obtained by dividing the two reflects the quantization in the distance dimension.

[0063] D azimuth is the azimuth coverage, that is, the angular range that the radar can detect in the azimuth direction; ΔD azimuth is the azimuth resolution, that is, the minimum angular separation at which the radar can distinguish two targets that are close in azimuth; the division result determines the number of columns in the matrix, reflecting the degree of quantization of the azimuth dimension.

[0064] The radar echo feature matrix constructed in the above manner can organize the feature data received by the radar in an orderly manner in the form of a matrix, which facilitates subsequent analysis, processing, and target identification of the radar data.

[0065] Define element status data A according to the state requirements of the target at different times IJ , element status data A IJ Including: the three-dimensional coordinates of the target are (x, y, z), the velocity vector .

[0066] A IJ The information contained is intended to describe the state of the target at a specific moment, and will be defined and adjusted accordingly as the target state changes to meet the need for accurate representation of the target state.

[0067] Element status data A IJ Assign a unique identifier. When processing status data of multiple targets or the same target at different times, the unique identifier can easily distinguish these data to avoid confusion. It also makes it easier to accurately locate and use the corresponding data in subsequent data processing, analysis, and data connection between different modules.

[0068] Element status data A IJ Is a vector that contains the target's three-dimensional coordinates (x, y, z) and velocity vector The three-dimensional coordinates (x, y, z) of the target clearly define the position of the target in space, and determine the specific direction of the target through three dimensions; the velocity vector It describes the target's movement speed in three-dimensional space. The three components correspond to the speed in the x, y, and z directions, respectively, which can reflect the speed and direction of the target's movement.

[0069] These data are spliced ​​into a six-dimensional vector, whose mathematical expression is:

[0070] ;

[0071] This six-dimensional vector representation integrates target position and motion state information into a single data structure, making it easy to comprehensively analyze and process the target's state during radar data processing. In target tracking algorithms, this six-dimensional vector can be used to predict and update the target's motion trajectory. Furthermore, this representation ensures that each element of the state data contains the target's position and velocity at that moment, and each element of the state data can be uniquely identified through an index. Ultimately, the matrix element state data can be represented by a six-dimensional vector containing the target's three-dimensional coordinates and velocity vector.

[0072] S3. Create a radar point cloud image, and map the nodes of the radar point cloud image to the elements of the radar echo feature matrix through a quantized mapping matrix.

[0073] For each radar echo data obtained by the sensor, it is converted into radar point cloud form to create a radar point cloud image I P .

[0074] Preferably, use the plotting tools provided by Matlab to create a 3D scatter plot using functions to visualize the point cloud. Map the x, y, and z coordinates in the point cloud data to positions in three-dimensional space to create a point cloud image. Use the axis equal command to adjust the axis scale for a more intuitive image display.

[0075] The nodes of the radar point cloud image correspond to the elements of the radar echo feature matrix, and each node of the radar point cloud image is associated with the state node of the corresponding matrix element.

[0076] Preferably, radar point cloud image nodes are associated with matrix elements. First, ensure that each node in the radar point cloud image has a unique identifier that corresponds to the position of the radar echo feature matrix element. Matrix elements can be numbered sequentially by row, starting from the top left corner, according to the matrix's row and column order. The nodes in the radar point cloud image are also numbered using the same order and rule.

[0077] According to the generation principle of radar point cloud images, the spatial distribution of its nodes is determined based on target distance, angle, and other information in the radar echo data. The rows and columns of the radar echo feature matrix correspond to dimensions such as distance and azimuth, respectively. Therefore, nodes in the point cloud image that reflect the same distance and azimuth information are associated with elements in the corresponding rows and columns of the matrix. Preferably, nodes at a specific angle and distance in the point cloud image can be associated with elements in the matrix at positions corresponding to azimuth and distance resolution.

[0078] Radar echo feature matrix elements contain state data (target 3D coordinates, velocity vector). During correlation, point cloud image nodes are connected to the state information of the matrix elements. This means that nodes in the point cloud image not only correspond to matrix elements in position but also carry the target state information represented by the matrix elements, allowing the point cloud image to more comprehensively reflect the target's characteristics.

[0079] Radar point cloud image I P The size of is mapped to the radar echo feature matrix through the quantization mapping matrix S. The quantization mapping matrix S is used to establish the radar point cloud image I P The mapping relationship between the size of and the radar echo feature matrix enables the radar echo data to be reasonably reflected in the radar point cloud image.

[0080] The element S(a,b) of the quantization mapping matrix S is:

[0081] Among them, a and b are the row and column indices used to index the elements of the radar echo feature matrix, The azimuth angle of the radar echo data point, that is, the angular position information of the data point in space, is used to determine the attributes of the data point in the azimuth dimension; The quantization step size of the azimuth is the interval size when the azimuth is quantized. The azimuth information is quantized by dividing the azimuth by the quantization step size and rounding down to obtain the element value of the corresponding position in the quantization mapping matrix. Indicates rounding down, that is, taking the largest integer that is not greater than the internal value.

[0082] For each radar echo feature matrix element state data A IJ The location of the node in the corresponding radar point cloud image Calculated by the following formula:

[0083] ;

[0084] ;

[0085] Where: M' and N' are the total number of rows and columns of nodes in the radar point cloud image, Represents a rounding operation. This formula determines the row position of the node in the point cloud image by dividing the quantized mapping matrix element S(a,b) by the number of node rows in the radar point cloud image and rounding it down. ; Determine the column position of the node in the point cloud image by dividing the quantization mapping matrix element S(a,b) by the number of radar point cloud image node columns and rounding down .

[0086] By constructing the above-mentioned quantitative mapping matrix S and calculating the node positions, the information in the radar echo feature matrix can be accurately mapped to the radar point cloud image, realizing the association and conversion between the two, which facilitates the subsequent analysis and processing of the radar data.

[0087] Preferably, if the resolutions of the radar echo feature matrix and the radar point cloud image are different, they can be matched by adjusting the quantization step size. If the physical sizes of the radar echo feature matrix and the radar point cloud image are also different, the area covered by the radar echo feature matrix is ​​L radians, while the area covered by the image is radians, the radar data can be scaled to the coverage of the image before quantization.

[0088] The calculation formula of the quantization mapping matrix element S(a,b) becomes:

[0089] By adjusting the quantization step size and performing necessary scaling operations, the correct mapping relationship can be established even if the radar echo feature matrix and the radar point cloud image do not match in resolution and physical size, such as Figure 2 The figure shows the matching diagram of the points after correct mapping, which ensures that the radar data can be accurately and effectively reflected in the point cloud image.

[0090] Table 1. Mapping error statistics under different algorithms.

[0091] algorithm e1 / pixel e2 / pixel em / pixel er / pixel Zernike algorithm 12.62 1.73 3.24 0.62 iLM-FFD algorithm 2.87 0.46 1.26 0.20 This embodiment 1.62 0.28 0.96 0.12 e1 / pixel: represents the maximum absolute mapping error. It measures the maximum absolute value of the difference between the actual position of a single point and the ideal mapped position during the mapping process, reflecting the maximum possible degree of mapping deviation.

[0092] e² / pixel: Mean absolute mapping error. This is calculated by averaging the absolute differences between the actual and ideal positions of all mapped points. It is used to assess the average level of overall mapping deviation and reflects the overall average deviation of the mapping error.

[0093] em / pixel: This is the root mean square (RMS) mapping error. This is calculated by averaging the squares of the differences between the actual and ideal positions of all mapped points and taking the square root. This comprehensively considers the error size of each point and is more sensitive to larger errors, providing a more comprehensive picture of the overall mapping error.

[0094] er / pixel: This is the relative mapping error. It is calculated by dividing the absolute mapping error by a reference value (such as image size or target feature size). It measures the proportional relationship between the mapping error and the reference, facilitating error comparisons across mapping tasks of different sizes or scales.

[0095] Table 1 shows that different algorithms perform differently in terms of various mapping error metrics. Compared to the Zernike algorithm, the iLM-FFD algorithm significantly reduces all mapping errors. The proposed algorithm also achieves relatively lower mapping error metrics, suggesting that it may offer better performance in controlling mapping errors.

[0096] S4. Perform target position continuity constraints and establish temporal edge connections between corresponding nodes at two moments in the radar point cloud image.

[0097] At a certain time t, the radar detects a target at a specific location, then the corresponding radar point cloud node is connected to the matrix state node A IJ associated.

[0098] Time series edge connection: for the element state data A(t) at adjacent times t and t+1 IJ and A(t+1) IJ , A(t) IJ and A(t+1) IJ The element state data of the radar echo feature matrix corresponding to two moments (t and t+1) respectively.

[0099] Establish the time sequence edge connection I between the nodes of the element state data of the radar echo feature matrix corresponding to two moments (t and t+1) in the radar point cloud image P (i',j',t)-I P (i',j',t+1), establish the position continuity constraint of the target.

[0100] In a preferred embodiment, the position continuity constraint of the target is established in combination with the velocity information:

[0101] According to the velocity vector information, ideally, the position change of the target at adjacent moments should conform to the law of motion. Assuming the time interval is , in the x direction, by the velocity and displacement The relationship can be obtained:

[0102] ;

[0103] Similarly, in the y direction, the velocity and displacement The relationship can be obtained:

[0104] ;

[0105] In the z direction, the velocity and displacement The relationship can be obtained:

[0106] ;

[0107] Establish target position continuity constraints:

[0108] Preferably define an error threshold , the target position is considered to be continuous when the following conditions are met:

[0109] If this condition is met, a temporal edge connection is established between the corresponding nodes at two moments in the radar point cloud image.

[0110] Example 2

[0111] This embodiment proposes a multi-source information mapping system for implementing the multi-source information mapping method for a radar signal processing system of embodiment 1, such as Figure 3 As shown, the multi-source information mapping system includes: a data acquisition and feature extraction module, a feature matrix construction module, a point cloud image creation and mapping module, and a target position constraint module.

[0112] The data acquisition and feature extraction module is used to acquire radar echo data, perform feature extraction on the radar echo data, and obtain radar feature data.

[0113] The data acquisition and feature extraction module includes: a radar echo data acquisition unit and a feature extraction processor.

[0114] The radar echo data acquisition unit is responsible for acquiring radar echo data. The radar echo data acquisition unit is directly connected to the radar equipment and can receive and transmit radar echo signals in real time. It is the data source of the entire multi-source information mapping system.

[0115] The feature extraction processor is used to extract features from the collected radar echo data. It preferably adopts various signal processing algorithms and machine learning algorithms, such as Fourier transform, wavelet transform, deep learning network (such as convolutional neural network), etc., to extract representative features from the original radar echo data to obtain radar feature data.

[0116] The feature matrix construction module is used to construct the radar echo feature matrix at each moment using radar feature data.

[0117] The feature matrix building module includes: a time series processing unit and a matrix builder.

[0118] The time series processing unit processes the radar characteristic data at each moment according to the time sequence of the radar characteristic data, and can accurately identify and divide the data at different moments, preparing for constructing the radar echo characteristic matrix at each moment.

[0119] Based on the time series processing unit, the matrix builder constructs the radar signature data at each moment into a radar echo signature matrix. The matrix construction method can be designed based on the dimension and type of the radar signature data to ensure that the matrix accurately reflects the characteristic information of the radar echo.

[0120] The point cloud image creation and mapping module is used to create radar point cloud images. The nodes of the radar point cloud image are mapped to the elements of the radar echo feature matrix through the quantization mapping matrix.

[0121] The point cloud image creation and mapping module includes: a point cloud image generator, a quantized mapping matrix generator and a mapping processor.

[0122] The point cloud image generator creates radar point cloud images and represents the spatial information of radar echoes in the form of point clouds. The point cloud image generator can generate point cloud images with appropriate density and distribution based on the radar's scanning range, resolution and other parameters.

[0123] The Quantization Mapping Matrix Generator is used to generate a quantization mapping matrix that maps the nodes of the radar point cloud image to the elements of the radar echo feature matrix. The generation of the quantization mapping matrix needs to consider the dimensions and structure of the point cloud image and the radar echo feature matrix to ensure the accuracy and consistency of the mapping.

[0124] The mapping processor is used to realize the mapping relationship between the nodes of the radar point cloud image and the elements of the radar echo feature matrix according to the quantized mapping matrix. It can match and associate the node information in the point cloud image with the element information of the radar echo feature matrix, providing a basis for subsequent analysis and processing.

[0125] The target position constraint module is used to establish edge connections between nodes in the radar point cloud image and perform target position continuity constraints.

[0126] The target position constraint module includes an edge connection builder and a continuity constraint processor. The edge connection builder is used to establish edge connections between nodes in the radar point cloud image. Edge connections can be determined based on factors such as spatial distance and similarity between nodes. By establishing edge connections, the nodes in the point cloud image can be organized into a graph structure, facilitating target position analysis and constraints. The continuity constraint processor applies target position continuity constraints to the established edge connections, optimizing and adjusting them based on target position continuity requirements to ensure target position accuracy and continuity.

[0127] In a preferred embodiment, the multi-source information mapping system also includes a system control and management module, which is responsible for the operation control and coordination of the entire system, including the start, stop, parameter setting and other operations of each module. It can reasonably allocate resources according to the system requirements and operating status to ensure the efficient operation of the system. It stores and manages the data generated during the operation of the system, including radar echo data, radar feature data, radar echo feature matrix, radar point cloud image, quantitative mapping matrix, etc. This unit can ensure the security and integrity of the data and facilitate subsequent data query and analysis, such as Figure 4 The figure shows a schematic diagram of data display.

[0128] The above modules work together to achieve the mapping of multi-source information and the constraint of target location. In actual application, the system structure can be adjusted and optimized according to specific needs and scenarios.

[0129] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted via the computer-readable storage medium. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrations. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0130] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A multi-source information mapping method for a radar signal processing system, characterized in that: The steps include: S1, obtaining radar echo data, performing feature extraction on the radar echo data to obtain radar feature data; S2. constructing a radar echo feature matrix at each moment using radar feature data; S3. Create a radar point cloud image, and map the nodes of the radar point cloud image to the elements of the radar echo feature matrix through a quantized mapping matrix; The element S(a,b) of the quantization mapping matrix S is: Among them, a and b are the row and column indices used to index the elements of the radar echo feature matrix, represents the azimuth of the radar echo data point, Indicates the quantization step size of the azimuth angle, Indicates rounding down; For each radar echo feature matrix element state data A IJ The location of the node in the corresponding radar point cloud image Calculated by the following formula: ； ； Where: M' and N' are the total number of rows and columns of nodes in the radar point cloud image; S4. Perform target position continuity constraints and establish temporal edge connections between corresponding nodes at two moments in the radar point cloud image.

2. The multi-source information mapping method for a radar signal processing system according to claim 1, characterized in that: In step S2, each element R of the radar echo feature matrix R IJ for: R IJ =A IJ +jP IJ ; Among them, I and J represent the row number and column number of the matrix respectively, A IJ Indicates element status data, P IJ Represents radar phase data, and j is an imaginary unit.

3. The multi-source information mapping method for a radar signal processing system according to claim 2, characterized in that: The total number of rows N and columns M of the radar echo feature matrix R is determined according to the resolution and range of the radar data: ； ； Among them, D range and D azimuth Respectively represent the maximum detection range and azimuth coverage of the radar, ΔD range and ΔD azimuth They represent the range resolution and azimuth resolution of radar data respectively.

4. The multi-source information mapping method for a radar signal processing system according to claim 3, characterized in that: Element status data A IJ is the three-dimensional coordinate and velocity vector The concatenated six-dimensional vector is expressed as: ; Corresponding to the velocities in the x, y, and z directions respectively.

5. The multi-source information mapping method for a radar signal processing system according to claim 1, characterized in that: Assume that the area covered by the radar echo feature matrix is ​​L radians, and the area covered by the image is L′ radians. Scale the radar data to the coverage of the image and then quantize it. The quantization mapping matrix element S(a,b) becomes : 。 6. The multi-source information mapping method for a radar signal processing system according to claim 4, characterized in that: In step S4, the target position continuity constraint is established in combination with the speed information: Let the time interval be , in the x direction, by the velocity and displacement Relationship: ； In the y direction, the velocity and displacement Relationship: ； In the z direction, the velocity and displacement Relationship: ； Establish target position continuity constraints: Define an error threshold , the target position is considered to be continuous when the following conditions are met: ； If this condition is met, a temporal edge connection is established between the corresponding nodes at two moments in the radar point cloud image.

Images