Pull-type interference confrontation method based on spatial morphological characteristics

Through the dragged interference adversarial method based on spatial morphological characteristics, single-pulse three-dimensional imaging and soft-spaced linear SVM discriminator are used to solve the target identification problem of radar under broadband signal interference, and improve the interference identification accuracy and target imaging quality.

CN119024278AActive Publication Date: 2024-11-26XIDIAN UNIV

Patent Information

Application Number
CN202411210184.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-26
Estimated Expiration
2044-08-30

AI Technical Summary

Technical Problem

In the prior art, radars are difficult to effectively combat broadband signal interference in the context of towed interference, especially when one-dimensional distance images are difficult to distinguish between targets and interference, the interference identification accuracy is low, and the target imaging quality after anti-interference is not high. .

Method used

Using drag interference confrontation method based on spatial morphological characteristics, echo data is obtained through single-pulse three-dimensional imaging technology, and noise removal and point cloud clustering are used to calculate the spatial covariance characteristic value of point cloud clusters. , build a soft-spaced linear SVM target discriminator, and use the differences between targets and interference in spatial morphology to achieve target identification and interference suppression.

Benefits of technology

It realizes efficient noise removal and target identification under broadband conditions, improves the accuracy of interference identification, improves the target imaging quality after anti-interference, and makes full use of the differences between interference and target in spatial morphological characteristics.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024278A_ABST
    Figure CN119024278A_ABST
Patent Text Reader

Abstract

The invention relates to a towed interference countermeasure method based on spatial morphological characteristics, which comprises the following steps: acquiring echo data of a sum channel, an azimuth difference channel and a pitch difference channel, and performing three-dimensional imaging of a radar foresight area on the echo data by using a monopulse three-dimensional imaging technology to obtain a three-dimensional point cloud; performing clustering and noise elimination on the three-dimensional point clouds by using a spatial filtering algorithm based on density clustering to obtain a plurality of point cloud clusters; calculating a spatial morphological feature descriptor of each point cloud cluster based on the spatial covariance feature values of the point cloud clusters; and performing target identification and interference suppression on each point cloud cluster in the point cloud cluster sample set by using a soft interval linear SVM target identifier based on the difference between the target and the interference in the spatial form to obtain an anti-interference target imaging result. According to the method, pull-type interference confrontation is realized by fully utilizing the difference between the interference and the target in spatial morphological characteristics, the accuracy of interference identification is relatively high, and the quality of an imaging result of the target after interference resistance is relatively high.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of radar technology, and in particular relates to a towed interference countermeasure method based on spatial morphological characteristics. Background Art

[0002] As a main lobe jamming style, towed decoys transmit high-power jamming signals to achieve angle deception of guidance radars to protect target aircraft. Thanks to their low cost, convenient equipment, and simple operation, towed decoys have become a common technical equipment for self-defense jamming of aircraft carriers, playing an important role.

[0003] Target tracking under towed decoy jamming conditions has always been a research hotspot at home and abroad. Due to a series of advantages such as low computational complexity, high tracking accuracy, and simple engineering implementation, single pulse measurement technology is widely used in terminal guidance. In view of the disadvantage that radar is easily deceived by decoy signals, resulting in radar loss of lock or tracking decoys, Chen Boxiao et al. used the corresponding differences between jamming signals and target echoes in different polarization channels to realize jamming presence detection and suppression. Chen Anna proposed a towed decoy detection method based on adaptive smoothing (Wigner-Hough transform), which realizes jamming presence detection by detecting the difference in peak characteristics before and after jamming is released. Yang Cheng et al. proposed to use single pulse diagonal difference channel signals to realize decoy presence identification, and realized target angle measurement through an adaptive one-dimensional conformal angle measurement algorithm. K. Yu et al. realized towed jamming countermeasure under EPC radar system by combining element pulse phase coding (Element-Pulse Code).

[0004] As a forward-looking imaging technology, single-pulse 3D imaging technology fills the blind spot where synthetic aperture technology cannot work in forward-looking scenes. Diao Guijie and others have realized the reconstruction of the strong scattering center point cloud of ship targets through single-pulse 3D imaging technology. Chen Boxiao and others have realized the interference countermeasure of outboard active interference and cross-eye interference through single-pulse 3D imaging technology to ensure stable tracking of targets.

[0005] However, for a long time, research on countermeasures against towed jamming has been mostly based on the premise that radars transmit narrowband signals, and there has been little research on countermeasures against radars transmitting wideband signals in the context of towed jamming. Summary of the invention

[0006] In order to solve the above problems existing in the prior art, the present invention provides a towed interference countermeasure method based on spatial morphological features. The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0007] The embodiment of the present invention provides a towed interference countermeasure method based on spatial morphological features, comprising the steps of:

[0008] Acquire echo data of the sum channel, azimuth difference channel and elevation difference channel, and use single pulse three-dimensional imaging technology to perform three-dimensional imaging of the radar forward viewing area on the echo data to obtain a three-dimensional point cloud;

[0009] Using a density-based spatial filtering algorithm to cluster and remove noise from the three-dimensional point cloud, a number of point cloud clusters are obtained;

[0010] Based on the spatial covariance eigenvalues ​​of the point cloud clusters, the spatial morphological feature descriptor of each point cloud cluster is calculated, and the spatial morphological features of multiple point cloud clusters constitute a point cloud cluster sample set;

[0011] Based on the difference between the target and the interference in spatial morphology, a soft-margin linear SVM target discriminator is used to perform target identification and interference suppression on each point cloud cluster in the point cloud cluster sample set to obtain the anti-interference target imaging result.

[0012] In one embodiment of the present invention, the echo data of the sum channel, the azimuth difference channel and the pitch difference channel are acquired, and the single pulse three-dimensional imaging technology is used to perform three-dimensional imaging of the radar forward viewing area on the echo data to obtain a three-dimensional point cloud, including:

[0013] Acquire echo data of sum channel, azimuth difference channel and elevation difference channel;

[0014] Performing range-Doppler high-resolution processing on the echo data to obtain high-resolution processing results of a sum channel, an azimuth difference channel, and a pitch difference channel;

[0015] Performing two-dimensional constant false alarm detection on the high-resolution processing results of the sum channel to extract distance information corresponding to a number of resolution units occupied by strong scattering points;

[0016] The high-resolution processing results of the sum channel, azimuth difference channel and elevation difference channel are used to measure the angle at the amplitude of the detection threshold unit to obtain the azimuth and elevation angle of each strong scattering point resolution unit relative to the radar;

[0017] Combined with the distance information, the azimuth angle and elevation angle of each strong scattering point resolution unit relative to the radar are converted into an azimuth-dimensional relative distance and an elevation-dimensional relative distance to obtain the three-dimensional point cloud.

[0018] In one embodiment of the present invention, the azimuth relative distance and the elevation relative distance are respectively:

[0019]

[0020] Among them, d az is the relative distance in the azimuth dimension, is the azimuth of each strong scattering point resolution unit relative to the radar, R = [R 1,R 2 ,...,R P ] is the distance information corresponding to the number of resolution units occupied by the strong scattering points, d el is the relative distance in the pitch dimension, θ=[θ 1 ,θ 2 ,...,θ P ] is the elevation angle of each strong scattering point resolution unit relative to the radar.

[0021] In one embodiment of the present invention, a density-based spatial filtering algorithm is used to cluster and remove noise from the three-dimensional point cloud to obtain a number of point cloud clusters, including:

[0022] Calculating the distance between any two scattering points in the three-dimensional point cloud to obtain a distance matrix;

[0023] A scattering point is selected as a reference point of a new cluster, and the number of first scattering points of the reference point within a preset neighborhood radius is counted. When the number of the first scattering points is less than the minimum number of elements in the preset domain, the reference point is removed as a noise point. When the number of the first scattering points is greater than or equal to the minimum number of elements in the preset domain, the scattering points within the preset neighborhood radius are included in the current cluster; wherein,

[0024] Selecting a scattering point in the current cluster that is not used as a reference point as a new reference point;

[0025] Counting the number of second scattering points within the preset neighborhood radius of the new reference point, when the number of the second scattering points is less than the minimum number of elements in the preset domain, selecting a new reference point in the current cluster for judgment; when the number of the second scattering points is greater than or equal to the minimum number of elements in the preset domain, classifying the scattering points within the preset neighborhood radius of the new reference point into the current cluster;

[0026] Traversing all scattering points in the current cluster;

[0027] When there are unclustered points, a scattering point is selected from the unclustered points as a reference point of a new cluster to be removed or clustered to obtain a new point cloud cluster; when there are no unclustered points, the plurality of point cloud clusters after spatial domain filtering are output.

[0028] In one embodiment of the present invention, based on the spatial covariance eigenvalues ​​of the point cloud clusters, the spatial morphological feature descriptor of each point cloud cluster is calculated, and the spatial morphological features of multiple point cloud clusters constitute a point cloud cluster sample set, including:

[0029] The spatial covariance matrix of the point cloud cluster is defined based on the scattering point positions and cluster center positions within the point cloud cluster:

[0030]

[0031] Among them, K is the number of scattering points in the point cloud cluster, p kc is the spatial position of the kth scattering point relative to the cluster center, p c is the cluster center position, p k is the position of the kth scattering point, p k =(x k ,y k ,z k ) T (k=1,...,K), x k ,y k ,z k represents the coordinates of the kth scattering point;

[0032] Perform eigendecomposition on the spatial covariance matrix to obtain the spatial covariance eigenvalue, and the expression of the eigendecomposition is:

[0033]

[0034] Where Λ=diag(λ 1 ,λ 2 ,λ 3 ) is a diagonal matrix of spatial covariance eigenvalues ​​arranged from large to small Three spatial covariance eigenvalues ​​λ 1 ,λ 2 ,λ 3 The corresponding eigenvector matrix is ​​V LRF =[v 1 ,v 2 ,v 3 ],v 1 、v 2 、v 3 is the characteristic column vector;

[0035] The spatial covariance eigenvalue is used to calculate the spatial morphological feature descriptor of each point cloud cluster to obtain the spatial morphological feature of each point cloud cluster:

[0036] x i =[s 1,i ,s 2,i ,…,s N,i ] T

[0037] Among them, x i Represents the N-dimensional feature vector formed by the spatial morphological features of the i-th point cloud cluster, s 1,i ,s 2,i ,…,s N,i Represents the N spatial morphological feature descriptors corresponding to the i-th point cloud cluster;

[0038] The spatial morphological features of all point cloud clusters form a point cloud cluster sample set:

[0039]

[0040] Where M represents the number of point cloud clusters, is an N-dimensional real number space.

[0041] In one embodiment of the present invention, the spatial morphological feature descriptors of each point cloud cluster include anisotropy, sum of eigenvalues, linearity, flatness, sphericity, feature entropy, curvature and total variance.

[0042] In one embodiment of the present invention, the calculation formula of the anisotropy is:

[0043] The calculation formula for the sum of the eigenvalues ​​is: 2 =λ 1 +λ 2 +λ 3 ;

[0044] The linearity calculation formula is:

[0045] The calculation formula of the flatness is:

[0046] The calculation formula of the sphericity is:

[0047] The calculation formula of the characteristic entropy is: 6 =-λ 1 lnλ 1 -λ 2 lnλ 2 -λ 3 lnλ 3 ;

[0048] The calculation formula of the curvature is:

[0049] The calculation formula of the total variance is:

[0050] In one embodiment of the present invention, the design method of the soft-margin linear SVM target discriminator comprises the steps of:

[0051] Obtain a point cloud cluster sample set and its corresponding label set;

[0052] The target identification problem in the point cloud cluster sample set is transformed into a convex quadratic programming binary classification problem:

[0053]

[0054] in, is the separating hyperplane w T x i +b=0 normal vector, b is the offset of the linear classifier, C is the penalty factor satisfying C>0, ζ i is a slack variable, M is the number of point cloud clusters in the sample set, x i =[s 1,i ,s 2,i ,…,s N,i ] T is the N-dimensional feature vector of the i-th point cloud cluster, is a set of M point cloud cluster samples, is an N-dimensional real number space;

[0055] By introducing Lagrange multipliers to construct Lagrange functions and combining strong duality, the convex quadratic programming binary classification problem is transformed into a dual classification problem:

[0056]

[0057] Among them, λ i is the first Lagrange multiplier of the i-th point cloud cluster, μ i is the second Lagrange multiplier of the i-th point cloud cluster, is the label set corresponding to the point cloud cluster sample set, -1 represents "bait" and "1" represents "target";

[0058] Solve the dual form of the classification problem and combine it with the convex quadratic programming binary classification problem to obtain the hyperplane And the corresponding decision function The soft-margin linear SVM target discriminator is obtained.

[0059] In one embodiment of the present invention, based on the difference between the target and the interference in spatial morphology, a soft-margin linear SVM target discriminator is used to perform target identification and interference suppression on each point cloud cluster in the point cloud cluster sample set to obtain an anti-interference target imaging result, including:

[0060] The point cloud cluster samples in the point cloud cluster sample set are input into the interval linear SVM target discriminator in sequence. When the output of the interval linear SVM target discriminator is 1, the point cloud cluster sample is the target. When the output of the interval linear SVM target discriminator is -1, the point cloud cluster sample is interference. The target imaging result is formed by the point cloud cluster samples whose output is the target.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] The towed interference countermeasure method based on spatial morphological features provided by the present invention realizes noise removal and strong scattering point cloud clustering by performing spatial domain filtering on the three-dimensional point cloud, extracts spatial morphological features of each point cloud cluster, realizes target identification and interference suppression through a soft margin linear SVM classifier, and obtains a three-dimensional image result of the target after interference resistance. The interference signal emitted by the bait under broadband conditions cannot completely simulate the spatial morphology of the target in the three-dimensional imaging space. When it is difficult to distinguish the target from the interference in a one-dimensional range image, the difference in spatial morphological features between the interference and the target is fully utilized to realize towed interference countermeasure. The interference identification accuracy is high, and the quality of the target imaging result after interference resistance is high. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A schematic flow chart of a towed interference countermeasure method based on spatial morphological features provided by an embodiment of the present invention;

[0064] Figure 2 A schematic flow chart of another method for countering towed interference based on spatial morphological features provided by an embodiment of the present invention;

[0065] Figure 3 A schematic diagram of a process flow of single pulse three-dimensional imaging provided by an embodiment of the present invention;

[0066] Figure 4 A schematic diagram of a flow chart of a spatial domain filtering algorithm based on density clustering provided in an embodiment of the present invention;

[0067] Figure 5 A diagram showing the spatial position relationship and signal transmission of a drag jamming scenario provided by an embodiment of the present invention;

[0068] Figure 6a-6d The three-dimensional imaging result and ISAR image of the target under the towed interference condition provided by the embodiment of the present invention;

[0069] Figure 7a-7c Three views of the point cloud clustering result provided by the embodiment of the present invention;

[0070] Figure 8a-8c Three views of target imaging results after interference suppression provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0071] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0072] Embodiment 1

[0073] See also Figure 1 and Figure 2 , Figure 1A flowchart of a towed interference countermeasure method based on spatial morphological features provided by an embodiment of the present invention is provided. Figure 2 A flowchart of another towed interference countermeasure method based on spatial morphological features provided by an embodiment of the present invention. The towed interference countermeasure method based on spatial morphological features comprises the following steps:

[0074] S1. Acquire the echo data of the sum channel, azimuth difference channel and pitch difference channel, and use the single pulse three-dimensional imaging technology to perform three-dimensional imaging of the radar forward viewing area on the echo data to obtain a three-dimensional point cloud.

[0075] See also Figure 3 , Figure 3 A schematic diagram of a process of single pulse three-dimensional imaging provided by an embodiment of the present invention, wherein step S1 specifically includes:

[0076] S11, radar receiving and echo data of channel, azimuth difference channel and elevation difference channel.

[0077] S12, performing range-Doppler high-resolution processing on the echo data to obtain high-resolution processing results of the sum channel, azimuth difference channel and elevation difference channel, that is, the focused ISAR image. Specifically, the range-Doppler high-resolution processing includes stretch processing, envelope alignment, phase focusing and pulse-to-pulse FFT.

[0078] S13, performing two-dimensional constant false alarm rate (2-D CFAR) detection on the high-resolution processing results of the sum channel, and extracting distance information corresponding to a number (P) of resolution units occupied by the strong scattering points.

[0079] S14, using the high-resolution processing results of the sum channel, azimuth difference channel and pitch difference channel to measure the angle at the amplitude of the detection threshold unit, and obtain the azimuth of each strong scattering point resolution unit relative to the radar and pitch angle θ = [θ 1 ,θ 2 ,...,θ P ].

[0080] S15. Combined with the distance information, the azimuth and elevation angles of each strong scattering point resolution unit relative to the radar are converted into azimuth-dimensional relative distance and elevation-dimensional relative distance to obtain a three-dimensional point cloud, thereby achieving three-dimensional imaging of the forward-looking area.

[0081] Specifically, the relative distance in azimuth and the relative distance in elevation are:

[0082]

[0083] Among them, d az is the relative distance in the azimuth dimension, d el is the relative distance in the pitch dimension.

[0084] S2. Use a density-based spatial filtering algorithm to cluster the three-dimensional point cloud and remove noise to obtain several point cloud clusters.

[0085] Angular glitter is the phase wavefront distortion of the echo caused by the superposition of random changes in scattering intensity and phase at different parts of a complex target, which manifests as abnormal angle measurement points. There are angular glitter points that are seriously deviated from the point cloud cluster in the three-dimensional image space and angle measurement singular points caused by noise. In this embodiment, they are uniformly classified as noise points and eliminated during spatial filtering to avoid negative effects on subsequent processing. Based on the idea of ​​density clustering, this embodiment proposes a spatial filtering algorithm for realizing three-dimensional point cloud clustering and noise removal.

[0086] See also Figure 4 , Figure 4 A schematic diagram of a flow chart of a spatial domain filtering algorithm based on density clustering provided by an embodiment of the present invention. The spatial domain filtering algorithm based on density clustering comprises the following steps:

[0087] S21. Input the 3D point cloud obtained by single pulse 3D imaging processing, complete the algorithm initialization, and pre-set the algorithm hyperparameters based on prior knowledge: neighborhood radius ξ, minimum number of elements contained in the domain N min ;

[0088] S22. Calculate the distance between any two scattering points in the three-dimensional point cloud to obtain a distance matrix D; use the matrix element D(i,p) in the distance matrix D to represent the distance between the i-th scattering point and the p-th scattering point (p=1,2,…,P).

[0089] S23, randomly select a scattering point q as the reference point of the new cluster, and count the number of the first scattering points N of the reference point within the preset neighborhood radius (D(q,p)≤ξ) near , when the number of first scattering points N near Less than the minimum number of elements N in the preset domain min , that is, N near <N min When the reference point q is a noise point, it is classified into the noise point cluster and removed. When the number of the first scattering points N near Greater than or equal to the minimum number of elements N contained in the preset domain min , that is, N near ≥N min When , the current cluster is established, and the scattering points within the preset neighborhood radius are classified into the current cluster.

[0090] S24, selecting a scattering point m in the current cluster that is not used as a reference point as a new reference point.

[0091] S25. Count the number of second scattering points N within the preset neighborhood radius of the new reference point m. near , when the number of the second scattering points is less than the minimum number of elements N in the preset domainmin , that is, N near <N min When the number of the second scattering points N is greater than , step S24 is repeated to select a new reference point in the current cluster for judgment; near Greater than or equal to the minimum number of elements N contained in the preset domain min , that is, N near ≥N min When , the scattering points within the preset neighborhood radius of the new reference point are classified into the current cluster and the cluster is updated.

[0092] S26. Repeat steps S24-S25 until all scattering points in the current cluster are traversed.

[0093] S27, determine whether there are still unclustered points. If there are unclustered points, select a scattering point from the unclustered points as a reference point for a new cluster, repeat steps S23-S26, remove or cluster the reference points of the new cluster, and form a new point cloud cluster.

[0094] S28. When there are no unclustered points, output a number of point cloud clusters after spatial filtering.

[0095] The spatial domain filtering algorithm of this embodiment does not require a priori knowledge of the number of target types when clustering the point cloud, and can remove noise points that are free from the cluster.

[0096] S3. Based on the spatial covariance eigenvalues ​​of the point cloud cluster, the spatial morphological feature descriptor of each point cloud cluster is calculated, and the spatial morphological features of multiple point cloud clusters constitute a point cloud cluster sample set.

[0097] Step S3 further realizes the identification of target clusters and the suppression of interference clusters by extracting the spatial morphological features of point cloud clusters, which specifically includes:

[0098] S31. Define the spatial covariance matrix of the point cloud cluster based on the scattering point positions and cluster center positions within the point cloud cluster.

[0099] Specifically, in three-dimensional space, the position of the scattering points in a point cloud cluster P containing K scattering points can be expressed as p k =(x k ,y k ,z k ) T (k=1,…,K), where x k ,y k ,z k represents the coordinates of the kth scattering point, then the position of the cluster center p c It is expressed as:

[0100]

[0101] The spatial covariance matrix of the point cloud cluster is defined as:

[0102]

[0103] Among them, p kc is the spatial position of the kth scattering point relative to the cluster center.

[0104] S32. Perform eigendecomposition on the spatial covariance matrix to obtain the spatial covariance eigenvalue. The expression of the eigendecomposition is:

[0105]

[0106] Where Λ=diag(λ 1 ,λ 2 ,λ 3 ) is a diagonal matrix of spatial covariance eigenvalues ​​arranged from large to small Three spatial covariance eigenvalues ​​λ 1 ,λ 2 ,λ 3 The corresponding eigenvector matrix is ​​V LRF =[v 1 ,v 2 ,v 3 ],v 1 、v 2 、v 3 is the characteristic column vector. The characteristic column vector v 1 、v 2 、v 3 They are mutually orthogonal and can be used as the three coordinate axes of the local reference frame (LPF). The point cloud feature description based on LPF description is only related to the geometric shape of the target, and has nothing to do with the position and posture of the target. The eigenvalues ​​of the spatial covariance matrix show a strong ability in describing complex geometric structure information.

[0107] S33, using the spatial covariance eigenvalue to calculate the spatial morphological feature descriptor of each point cloud cluster, and obtain the spatial morphological features of each point cloud cluster:

[0108] x i =[s 1,i ,s 2,i ,…,s N,i ] T

[0109] Among them, x i Represents the N-dimensional feature vector formed by the spatial morphological features of the i-th point cloud cluster, s 1,i ,s 2,i ,…,s N,i Represents the N spatial morphological feature descriptors corresponding to the i-th point cloud cluster.

[0110] Specifically, the spatial morphological feature descriptors of each point cloud cluster include anisotropy, sum of eigenvalues, linearity, flatness, sphericity, feature entropy, curvature and total variance. Please refer to Table 1, which shows several spatial morphological feature descriptors provided in this embodiment.

[0111] Table 1 Spatial morphological feature descriptors

[0112]

[0113] The spatial morphological feature descriptor of point cloud clusters can characterize point cloud clusters of different physical sizes and geometric shapes, and the features remain robust after the point cloud clusters undergo translation and rotation transformations.

[0114] S34. The spatial morphological features of all point cloud clusters form a point cloud cluster sample set:

[0115]

[0116] Where M represents the number of point cloud clusters, is an N-dimensional real number space.

[0117] S4. Based on the difference in spatial morphology between the target and the interference, the soft-margin linear SVM target discriminator is used to perform target identification and interference suppression on each point cloud cluster in the point cloud cluster sample set to obtain the target imaging result after anti-interference.

[0118] The spatial form of the three-dimensional point cloud reflects the physical size and shape information of the target. Under broadband conditions, the point cloud cluster distribution of the interference signal emitted by the towed decoy and the target echo in the three-dimensional image signal space shows obvious differences. The point cloud generated by the towed decoy signal shows the characteristics of high energy, small physical size, and concentrated point cloud distribution; while the scattered point cloud cluster of the target shows obvious three-dimensional geometric characteristics, large physical size, large distribution range of scattering points, and obvious anisotropy.

[0119] Based on the spatial morphological feature descriptor, this embodiment proposes a multi-feature soft-margin linear SVM target discriminator to achieve the identification of bait and target. The design method of the margin linear SVM target discriminator includes the following steps:

[0120] 1) Obtain a point cloud cluster sample set and its corresponding label set.

[0121] Specifically, let the input M point cloud cluster sample set be x i =[s 1,i ,s 2,i ,…,s N,i ] T is the N-dimensional feature vector of the i-th point cloud cluster, s 1,i ,s 2,i,…,s N,i is the N spatial morphological feature descriptors corresponding to the i-th point cloud cluster, is the label set of the corresponding input feature vector. It is an N-dimensional real number space, -1 represents "bait" and "1" represents "target".

[0122] 2) Convert the target identification problem in the point cloud cluster sample set into a convex quadratic programming binary classification problem:

[0123]

[0124] in, is the separating hyperplane w T x i +b=0 normal vector, b is the offset of the linear classifier, C is the penalty factor satisfying C>0, ζ i is the slack variable.

[0125] 3) By introducing Lagrange multipliers to construct Lagrangian functions and combining strong duality, the convex quadratic programming binary classification problem is transformed into a dual form of classification problem.

[0126] Specifically, by introducing the Lagrange multiplier λ i , μ i , construct the Lagrangian function:

[0127]

[0128] st i ≥0,μ i ≥0,ζ i ≥0

[0129] The station meets:

[0130]

[0131] Substitute the conditions satisfied at the stationary point into the Lagrangian function and, based on the strong duality, transform the convex quadratic programming binary classification problem into a classification problem in a dual form that is easier to calculate:

[0132]

[0133] Among them, λ i is the first Lagrange multiplier of the i-th point cloud cluster, μ i is the second Lagrange multiplier of the i-th point cloud cluster.

[0134] 4) Solve the dual form of the classification problem, and combine it with the convex quadratic programming binary classification problem to obtain the hyperplane and the corresponding decision function, and obtain the soft margin linear SVM target discriminator.

[0135] Specifically, the Sequential Minimal Optimization (SMO) algorithm is used to solve the dual form of the solution is a vector formed by the Lagrange multipliers of multiple point cloud clusters, and then Substitute it into the convex quadratic programming binary classification problem and find the solution to the problem and Finally, the separating hyperplane is obtained And the corresponding decision function Obtain a soft-margin linear SVM target discriminator.

[0136] Further, the point cloud cluster sample set obtained by processing steps S1, S2, and S3 is input into the soft-margin linear SVM target discriminator, and the soft-margin linear SVM target discriminator outputs the target identification result. When the output of the margin linear SVM target discriminator is 1, the point cloud cluster sample is the target, and when the output of the margin linear SVM target discriminator is -1, the point cloud cluster sample is interference, and the target imaging result is formed by the point cloud cluster sample output as the target.

[0137] The towed interference countermeasure method based on spatial morphological features provided in this embodiment performs spatial filtering on the three-dimensional point cloud to achieve noise removal and strong scattering point cloud clustering, and extracts spatial morphological features of each point cloud cluster, and realizes target identification and interference suppression through a soft-margin linear SVM classifier to obtain a three-dimensional image of the target after interference resistance. The interference signal emitted by the bait under broadband conditions cannot completely simulate the spatial morphology of the target in the three-dimensional imaging space. When it is difficult to distinguish the target from the interference in the one-dimensional range image, the difference in spatial morphological features between the interference and the target is fully utilized to achieve towed interference countermeasure, and the interference identification accuracy is high, and the quality of the target imaging result after interference resistance is high.

[0138] Furthermore, in order to verify the effectiveness of the above method, this embodiment takes the single pulse terminal guidance tracking system as an example to simulate a typical towed jamming scenario. The main simulation parameters are shown in Table 2.

[0139] Table 2 Main simulation parameters

[0140]

[0141] Towed decoy jamming scenarios such as Figure 5 As shown, Figure 5The spatial position relationship and signal transmission diagram of the towed jamming scenario provided by the embodiment of the present invention. The radar is located at the coordinate origin O(0,0,0). At the initial moment, the center of the radar beam is irradiated directly at the target's centroid, and the beam center points at a horizontal angle of 30° to the target's flight direction. The towed decoy is located directly behind the target. The jammer transmits a modulated false target signal.

[0142] By stretching the received signals of the three channels of sum, azimuth difference, and pitch difference, it can be seen from the processing results that the modulated forwarding interference signal emitted by the decoy jammer occupies multiple distance units, and its amplitude is higher than that of the real target. It has good distance deception characteristics, and it is difficult to distinguish the target from the interference from the one-dimensional range image.

[0143] After further motion compensation, single pulse three-dimensional imaging of the target under towed interference conditions is performed. According to the angle measurement results of three-dimensional imaging, it can be seen that the measurement singularity points caused by interference and the strong scattering points of the target are both located in the main lobe beam, which seriously affects the guidance and tracking.

[0144] See also Figure 6a-6d , Figure 6a-6d The three-dimensional imaging result of the target under the towed interference condition provided by the embodiment of the present invention is three-dimensional and ISAR image, Figure 6a is the front view, Figure 6b is a side view, Figure 6c is a top view, Figure 6d The angle measurement result is combined with the distance information to be converted into a three-dimensional image of the forward-looking scene through the calculation formula of the relative distance in the azimuth dimension and the relative distance in the elevation dimension, such as Figure 6a-6c As shown in the figure, from the results of the three-dimensional image, it can be obtained that the front-view area is composed of two point cloud clusters with large shapes, which are composed of strong scattering point echoes, representing the real target and the false target generated by the interference signal. From the 6d ISAR imaging results, it can also be obtained that there is a strong interference with an amplitude far higher than the target echo outside the target size. Since the speed of the target and the decoy is far greater than the speed measurement range, there is speed ambiguity, and the ISAR results also reflect this phenomenon.

[0145] Furthermore, the imaging results are subjected to spatial noise removal and point cloud clustering to obtain the point cloud clustering results, such as Figure 7a-7c As shown, Figure 7a-7c The three views of the point cloud clustering result provided by the embodiment of the present invention are: Figure 7a is the front view, Figure 7b is a side view, Figure 7c As shown in Figure 7, the false target is generated by the interference signal, which presents high-density and linear features in spatial form, and has obvious three-dimensional shape feature differences with the real target in the three-dimensional image space.

[0146] Under the current simulation parameter conditions, 1000 Monte Carlo experiments were performed to obtain cluster point cloud clusters. Feature extraction and classifier training were performed on the cluster point cloud clusters generated by 800 experiments, and the parameters of the designed discriminator decision function were obtained as follows:

[0147]

[0148] The point cloud clusters obtained from the remaining 200 experiments were used for target identification test verification. According to the verification results, the interference signal of the decoy and the target echo have significant spatial morphological characteristics after reconstruction in the three-dimensional imaging space. The interference signal has higher linearity, smaller size, higher aggregation, and weaker irregularity in three-dimensional space, while the target has obvious shape characteristics, larger volume, higher flatness, more dispersed scattering point distribution, and stronger irregularity. From the test results, the accuracy of the designed classifier in identifying targets and interference reached 100%.

[0149] The target imaging result after interference suppression is as follows: Figure 8a-8c As shown, Figure 8a-8c The three-view image of the target imaging result after interference suppression provided by the embodiment of the present invention, Figure 8a 8b is a front view, and 8b is a side view. Figure 8c It can be seen from Figure 8 that the three-view image of the target after anti-interference is basically consistent with the three-view image of the real target.

[0150] In view of the problem that the terminal guidance radar is easily deceived by the towed decoy and causes the attack failure, this embodiment proposes a method for terminal guidance radar to resist the towed decoy based on spatial morphological features. The method uses single pulse three-dimensional imaging technology to realize three-dimensional imaging of the front-view area; clustering and noise removal of strong scattering point cloud clusters are realized through spatial domain filtering; by extracting the spatial morphological features of each point cloud cluster, the soft margin linear SVM classifier is used to realize target identification and remove interference points to obtain a three-dimensional image of the real target. Simulation results verify the effectiveness of this method in combating towed decoys.

[0151] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A towing interference countermeasure method based on spatial morphological characteristics, characterized in that: Includes steps: Acquire echo data of the sum channel, azimuth difference channel and elevation difference channel, and use single pulse three-dimensional imaging technology to perform three-dimensional imaging of the radar forward viewing area on the echo data to obtain a three-dimensional point cloud; Using a density-based spatial filtering algorithm to cluster and remove noise from the three-dimensional point cloud, a number of point cloud clusters are obtained; Based on the spatial covariance eigenvalues ​​of the point cloud clusters, the spatial morphological feature descriptor of each point cloud cluster is calculated, and the spatial morphological features of multiple point cloud clusters constitute a point cloud cluster sample set; Based on the difference between the target and the interference in spatial morphology, a soft-margin linear SVM target discriminator is used to perform target identification and interference suppression on each point cloud cluster in the point cloud cluster sample set to obtain the anti-interference target imaging result.

2. The towed jamming countermeasure method based on spatial morphological features according to claim 1 is characterized in that: Acquire the echo data of the sum channel, the azimuth difference channel and the pitch difference channel, and use the single pulse three-dimensional imaging technology to perform three-dimensional imaging of the radar forward viewing area on the echo data to obtain a three-dimensional point cloud, including: Acquire echo data of sum channel, azimuth difference channel and elevation difference channel; Performing range-Doppler high-resolution processing on the echo data to obtain high-resolution processing results of a sum channel, an azimuth difference channel, and a pitch difference channel; Performing two-dimensional constant false alarm detection on the high-resolution processing results of the sum channel to extract distance information corresponding to a number of resolution units occupied by strong scattering points; The high-resolution processing results of the sum channel, azimuth difference channel and elevation difference channel are used to measure the angle at the amplitude of the detection threshold unit to obtain the azimuth and elevation angle of each strong scattering point resolution unit relative to the radar; Combined with the distance information, the azimuth angle and elevation angle of each strong scattering point resolution unit relative to the radar are converted into an azimuth-dimensional relative distance and an elevation-dimensional relative distance to obtain the three-dimensional point cloud.

3. The towed interference countermeasure method based on spatial morphological features according to claim 2 is characterized in that: The relative distance in the azimuth dimension and the relative distance in the elevation dimension are respectively: Among them, d az is the relative distance in the azimuth dimension, is the azimuth of each strong scattering point resolution unit relative to the radar, R = [R1, R2, ..., R P ] is the distance information corresponding to the number of resolution units occupied by the strong scattering points, d el is the relative distance in the pitch dimension, θ=[θ1,θ2,...,θ P ] is the elevation angle of each strong scattering point resolution unit relative to the radar.

4. The towed interference countermeasure method based on spatial morphological features according to claim 1 is characterized in that: The three-dimensional point cloud is clustered and noise is removed using a spatial domain filtering algorithm based on density clustering to obtain several point cloud clusters, including: Calculating the distance between any two scattering points in the three-dimensional point cloud to obtain a distance matrix; A scattering point is selected as a reference point of a new cluster, and the number of first scattering points of the reference point within a preset neighborhood radius is counted. When the number of the first scattering points is less than the minimum number of elements in the preset domain, the reference point is removed as a noise point. When the number of the first scattering points is greater than or equal to the minimum number of elements in the preset domain, the scattering points within the preset neighborhood radius are included in the current cluster; wherein, Selecting a scattering point in the current cluster that is not used as a reference point as a new reference point; Counting the number of second scattering points within the preset neighborhood radius of the new reference point, when the number of the second scattering points is less than the minimum number of elements in the preset domain, selecting a new reference point in the current cluster for judgment; when the number of the second scattering points is greater than or equal to the minimum number of elements in the preset domain, classifying the scattering points within the preset neighborhood radius of the new reference point into the current cluster; Traversing all scattering points in the current cluster; When there are unclustered points, a scattering point is selected from the unclustered points as a reference point of a new cluster to be removed or clustered to obtain a new point cloud cluster; when there are no unclustered points, the plurality of point cloud clusters after spatial domain filtering are output.

5. The towed jamming countermeasure method based on spatial morphological features according to claim 1 is characterized in that: Based on the spatial covariance eigenvalues ​​of the point cloud cluster, the spatial morphological feature descriptor of each point cloud cluster is calculated, and the spatial morphological features of multiple point cloud clusters constitute a point cloud cluster sample set, including: The spatial covariance matrix of the point cloud cluster is defined based on the scattering point positions and cluster center positions within the point cloud cluster: Among them, K is the number of scattering points in the point cloud cluster, p kc is the spatial position of the kth scattering point relative to the cluster center, p c is the cluster center position, p k is the position of the kth scattering point, p k =(x k ,y k ,z k ) T (k=1,...,K), x k ,y k ,z k represents the coordinates of the kth scattering point; Perform eigendecomposition on the spatial covariance matrix to obtain the spatial covariance eigenvalue, and the expression of the eigendecomposition is: Among them, Λ=diag(λ1,λ2,λ3) is a diagonal matrix with spatial covariance eigenvalues ​​arranged from large to small. The eigenvector matrix corresponding to the three spatial covariance eigenvalues ​​λ1, λ2, λ3 is V LRF =[v1,v2,v3], v1, v2, v3 are feature column vectors; The spatial covariance eigenvalue is used to calculate the spatial morphological feature descriptor of each point cloud cluster to obtain the spatial morphological feature of each point cloud cluster: x i =[s 1,i ,s 2,i ,…,s N,i ] T Among them, x i Represents the N-dimensional feature vector formed by the spatial morphological features of the i-th point cloud cluster, s 1,i ,s 2,i ,…,s N,i Represents the N spatial morphological feature descriptors corresponding to the i-th point cloud cluster; The spatial morphological features of all point cloud clusters form a point cloud cluster sample set: Where M represents the number of point cloud clusters, is an N-dimensional real number space.

6. The towed jamming countermeasure method based on spatial morphological features according to claim 5 is characterized in that: The spatial morphological feature descriptors of each point cloud cluster include anisotropy, sum of eigenvalues, linearity, flatness, sphericity, characteristic entropy, curvature and total variance.

7. The towed interference countermeasure method based on spatial morphological features according to claim 6 is characterized in that: The calculation formula of the anisotropy is: The calculation formula of the sum of the eigenvalues ​​is: s2=λ1+λ2+λ3; The linearity calculation formula is: The calculation formula of the flatness is: The calculation formula of the sphericity is: The characteristic entropy is calculated as follows: s6 = -λ1lnλ1-λ2lnλ2-λ3lnλ3; The calculation formula of the curvature is: The calculation formula of the total variance is:

8. The towed interference countermeasure method based on spatial morphological features according to claim 1 is characterized in that: The design method of the soft-margin linear SVM target discriminator comprises the steps of: Obtain a point cloud cluster sample set and its corresponding label set; The target identification problem in the point cloud cluster sample set is transformed into a convex quadratic programming binary classification problem: s.t.y i (w T x i +b)≥1-ζ i ,i=1,2,…,M g i ≥0,i=1,2,…,M in, is the separating hyperplane w T x i +b=0 normal vector, b is the offset of the linear classifier, C is the penalty factor satisfying C>0, ζ i is a slack variable, M is the number of point cloud clusters in the sample set, x i =[s 1,i ,s 2,i ,…,s N,i ] T is the N-dimensional feature vector of the i-th point cloud cluster, is a set of M point cloud cluster samples, is an N-dimensional real number space; By introducing Lagrange multipliers to construct Lagrange functions and combining strong duality, the convex quadratic programming binary classification problem is transformed into a dual classification problem: stC-λ i -m i =0 l i ≥0 m i ≥0,i=1,2,…,M Among them, λ i is the first Lagrange multiplier of the i-th point cloud cluster, μ i is the second Lagrange multiplier of the i-th point cloud cluster, is the label set corresponding to the point cloud cluster sample set, -1 represents "bait" and "1" represents "target"; Solve the dual form of the classification problem and combine it with the convex quadratic programming binary classification problem to obtain the hyperplane And the corresponding decision function The soft-margin linear SVM target discriminator is obtained.

9. The towed interference countermeasure method based on spatial morphological features according to claim 8 is characterized in that: Based on the difference between the target and the interference in spatial morphology, a soft margin linear SVM target discriminator is used to perform target identification and interference suppression on each point cloud cluster in the point cloud cluster sample set, and the anti-interference target imaging result is obtained, including: The point cloud cluster samples in the point cloud cluster sample set are input into the interval linear SVM target discriminator in sequence. When the output of the interval linear SVM target discriminator is 1, the point cloud cluster sample is the target. When the output of the interval linear SVM target discriminator is -1, the point cloud cluster sample is interference. The target imaging result is formed by the point cloud cluster samples whose output is the target.

Citation Information

Patent Citations

  • One-dimensional distance image three-dimensional scene recovery method and device based on multi-radar cooperation

    CN115113162A

  • Towed interference identification method based on transmitting and receiving spatial frequency spectrum

    CN117724055A

  • Angle anti-array interference confrontation method and device based on monopulse three-dimensional imaging

    CN118549893A

Cited By

  • Passive synthetic aperture method based on R-L deconvolution beam domain processing

    CN120468860A