Target scattering center extraction method and device, and electronic equipment

Abstract

The embodiment of the present application relates to the technical field of signal processing, in particular to a target scattering center extraction method, device and electronic equipment. The method comprises: obtaining radar echo data of a target to be measured; inputting the radar echo data of the target to be measured into a sparse estimation model to obtain an index set matrix of the radar echo data of the target to be measured; the sparse estimation model is trained by using radar echo samples of a plurality of known targets, each radar echo sample pair comprises sample echo data as input and sample index set matrix as output, the sample echo data comprises at least one scattering center, and the sample index set matrix is calculated according to the sample echo data; according to the index set matrix of the radar echo data of the target to be measured, the position of each scattering center in the target to be measured and the scattering amplitude coefficient of each scattering center are determined. The scheme can improve the extraction speed of the target scattering center.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of signal processing, in particular to a target scattering center extraction method, device and electronic equipment. BACKGROUND

[0002] Radar echoes mainly come from a small number of scattering centers of a target, so as long as the scattering center information of the target is extracted from the radar echoes, the target information such as distance, azimuth, speed and shape can be inverted. Generally, the scattering centers of the target are mainly generated at discontinuous point parts such as edges, inflection points, corners and tips of the target, resulting in strong sparsity of the radar echoes. The sparsity leads to that the extraction of the radar target scattering center itself belongs to an ill-posed problem, and the extraction accuracy is low.

[0003] In related technologies, the ill-posedness of inversion is eliminated by introducing sparse constraints to improve the inversion accuracy. However, the sparse method often needs to extract atoms representing scattering centers one by one through a loop iteration manner, and the estimation speed is slow, especially for complex targets with a large number of scattering centers, the feature inversion efficiency is lower.

[0004] Therefore, there is an urgent need for a target scattering center extraction method, device and electronic equipment to solve the above problems. SUMMARY

[0005] Based on the problem of low efficiency of the existing target scattering center extraction method, embodiments of the present application provide a target scattering center extraction method, device and electronic equipment, which can improve the extraction speed of the target scattering center.

[0006] In a first aspect, embodiments of the present application provide a target scattering center extraction method, comprising:

[0007] obtaining radar echo data of a target to be measured;

[0008] inputting the radar echo data of the target to be measured into a sparse estimation model to obtain an index set matrix of the radar echo data of the target to be measured; the sparse estimation model is trained by using radar echo samples of a plurality of known targets, each radar echo sample pair includes sample echo data as input and sample index set matrix as output, the sample echo data includes at least one scattering center, and the sample index set matrix is calculated according to the sample echo data;

[0009] determining the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target to be measured according to the index set matrix of the radar echo data of the target to be measured.

[0010] In a possible design, the sample index set matrix is calculated by the following way:

[0011] constructing a sparse representation of the radar target scattering centers based on a parametric model of the scattering centers;

[0012] calculating a sample index set matrix of the sample echo data based on the sparse representation by using an orthogonal matching pursuit method.

[0013] In a possible design, the parametric model of the scattering centers is:

[0014]

[0015] where S(k n ) is a radar echo signal, k n is a wave number, n=1, 2, …, N, N is a number of radar frequency sampling points, m=1, 2, …, M, M is a number of scattering centers, A m is a complex scattering amplitude coefficient of the mth scattering center, r m is a radial position of the mth scattering center, e n is measurement noise.

[0016] The sparse representation of the radar target scattering centers is:

[0017] S = Φσ + E

[0018] where S is a target observation echo column vector, S = [S(k1), S(k2), …, S(k N )] T ; Φ is a sparse dictionary matrix, Φ = [φ(k1), φ(k2), …, φ(k N )] T , φ(k n )(n=1, …N) is an L-dimensional column vector composed of exp(-j2k n r l )(l=1, …, L); r l (l=1, …, L) is a radial position r m of all possible values, and E is noise.

[0019] In a possible design, the calculating, based on the sparse representation, of the sample index set matrix of the sample echo data by using the orthogonal matching pursuit method includes:

[0020] S1, setting an orthogonal matching pursuit threshold based on the sparse representation;

[0021] S2, initializing parameters, the parameters including a residual vector, an index set matrix, an estimated sparse coefficient vector, a support set matrix, and an iteration number;

[0022] S3, calculating element index according to current parameters, and updating index set matrix, support set matrix, estimated sparse coefficient vector and residual error vector according to calculation result;

[0023] S4, judging whether convergence condition is met according to updated residual error vector and the orthogonal matching pursuit threshold; if the condition is met, stopping iteration and executing S5, otherwise updating iteration times and returning to execute S3;

[0024] S5, taking updated index set matrix as sample index set matrix of the sample echo data.

[0025] In a possible design, the expression of the orthogonal matching pursuit threshold a is: a = 0.01 * ||S|| 2, wherein ||·|| 2 is a 2-norm of a vector;

[0026] In the S2, the initial residual error vector R (0) = S, the initial index set matrix J (0) = Φ, the initial estimated sparse coefficient vector the initial support set matrix Ω (0) = [] and the initial iteration times t = 0;

[0027] In the S2, when the element index is calculated, the expression is:

[0028]

[0029] In the expression, <·> is an inner product operator, R (t-1) is a residual error vector calculated in the t-1th iteration, and Φ ( :, j) is the jth column of the sparse dictionary Φ;

[0030] In the S2, when the index set matrix is updated, the expression is:

[0031] In the S2, when the support set matrix is updated, the expression is: Ω (t) = [Ω （t-1) , Φ ( :, j) ] ;

[0032] In the S2, when the estimated sparse coefficient vector is updated, the expression is:

[0033]

[0034] In the expression, the superscript H represents Hermitian transpose;

[0035] In the S2, when the residual error vector is updated, the expression is:

[0036] In a possible design, the determining whether the convergence condition is met according to the updated residual vector and the OMP threshold comprises the following steps.

[0037] According to the updated residual vector R (t) , a residual signal energy W is calculated, and an expression of the residual signal energy W is W = ||R (t) ||2.

[0038] It is determined whether the residual signal energy is less than the OMP threshold, and if yes, it is determined that the convergence condition is met, and if not, it is determined that the convergence condition is not met.

[0039] In a possible design, the determining the position of each scattering center and the scattering amplitude coefficient of each scattering center according to the index set matrix of the radar echo data of the target to be measured comprises the following steps.

[0040] The row l D in which an element with a value of 1 in the index set matrix J D of the radar echo data of the target to be measured is determined as the position r

[0041] The row l D in which an element with a value of 1 in the index set matrix J D of the target to be measured is determined as the support set Ω D = [..., Φ ( :, l D ),... ] (l D ∈ J D ), and the scattering amplitude coefficient A of each scattering center is:

[0042] In a possible design, the sparse estimation model comprises a first layer neural network, a second layer neural network and a third layer neural network.

[0043] The first layer neural network is configured to calculate an inner product of input radar echo data and each atom, and each atom represents a column in a sparse dictionary.

[0044] The second layer neural network and the third layer neural network are configured to extract atoms with relatively high correlation, and output an index set matrix corresponding to the input radar echo data according to an extraction result.

[0045] In a second aspect, an embodiment of the present application further provides an extraction device of a target scattering center, comprising:

[0046] An acquisition module is configured to acquire radar echo data of a target to be measured.

[0047] The input module is configured to input radar echo data of the target to be detected into a sparse estimation model to obtain an index set matrix of the radar echo data of the target to be detected.

[0048] The determination module is configured to determine the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target to be detected according to the index set matrix of the radar echo data of the target to be detected.

[0049] In a third aspect, an electronic device is provided, which includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the method described in any of the embodiments of the present specification is implemented.

[0050] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program. When the computer program is executed in a computer, the computer is caused to perform the method described in any of the embodiments of the present specification.

[0051] The embodiments of the present application provide a target scattering center extraction method and device and electronic device. The method first constructs a sparse estimation model for inverting a radar echo data index set matrix, then trains the sparse estimation model by using radar echo sample pairs of known targets to obtain a trained sparse estimation model. In this way, for radar echo data of a target to be detected, the radar echo data of the target to be detected can be input into the sparse estimation model to obtain an index set matrix of the radar echo data of the target to be detected. Finally, according to the index set matrix, the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target to be detected can be determined. Therefore, the present solution can improve the extraction speed of the target scattering center. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0053] Figure 1 is a flow chart of a target scattering center extraction method provided by an embodiment of the present application;

[0054] Figure 2is a hardware architecture diagram of an electronic device provided by an embodiment of the present application.

[0055] Figure 3 is a structure diagram of a target scattering center extraction device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0056] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings of the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0057] The specific implementation of the above concept will be described below.

[0058] Reference is made to Figure 1 The embodiments of the present application provide a target scattering center extraction method, which comprises the following steps.

[0059] In step 100, radar echo data of a target to be measured is acquired.

[0060] In step 102, the radar echo data of the target to be measured is input into a sparse estimation model to obtain an index set matrix of the radar echo data of the target to be measured. The sparse estimation model is trained by using radar echo sample pairs of a plurality of known targets. Each radar echo sample pair comprises sample echo data as input and a sample index set matrix as output. The sample echo data comprises at least one scattering center, and the sample index set matrix is calculated according to the sample echo data.

[0061] In step 104, the position of each scattering center in the target to be measured and the scattering amplitude coefficient of each scattering center are determined according to the index set matrix of the radar echo data of the target to be measured.

[0062] In the embodiments of the present application, a sparse estimation model for inverting the index set matrix of radar echo data is first constructed, and then the sparse estimation model is trained by using radar echo sample pairs of known targets to obtain a trained sparse estimation model. In this way, for the radar echo data of a target to be measured, the index set matrix of the radar echo data of the target to be measured can be obtained by inputting the radar echo data into the sparse estimation model. Finally, according to the index set matrix, the position of each scattering center in the target to be measured and the scattering amplitude coefficient of each scattering center can be determined. As can be seen, the present solution can improve the extraction speed of target scattering centers.

[0063] In step 102, the sample echo data of the known target can be obtained by simulation calculation. For example, the full-space broadband data of a typical target is simulated, and the full-space refers to covering all the line-of-sight angles of the radar, such as the elevation angle from 0° to 180°, the azimuth angle from -180° to 180°, and 1° interval, so that 65341 data, i.e. sample echo data, can be obtained, which are used as the input of the sparse estimation model.

[0064] After the input of the sparse estimation model is determined, the output of the sparse estimation model needs to be constructed, i.e. the sample index set matrix corresponding to each input sample echo data, and the sample echo data and the sample index set matrix form a radar echo sample pair, forming 65341 sample pairs.

[0065] In some embodiments, the sample index set matrix is obtained by the following calculation:

[0066] Step A1, based on the parameterized model of point scattering center, a sparse expression form of the radar target scattering center is constructed;

[0067] Step A2, based on the sparse expression form, the sample index set matrix of the sample echo data is calculated by using the orthogonal matching pursuit method.

[0068] For step A1, in some embodiments, the parameterized model of point scattering center is:

[0069]

[0070] In the formula, S(k n ) is the radar echo signal, k n is the wave number, n=1, 2, …, N, N is the number of radar frequency sampling points, m=1, 2, …, M, M is the number of scattering centers, A m is the complex scattering amplitude coefficient of the mth scattering center, r m is the radial position of the mth scattering center, e n is the measurement noise;

[0071] The sparse expression form of the radar target scattering center is:

[0072] S=Φσ+E

[0073] In the formula, S is the target observation echo column vector, S=[S(k1), S(k2), …, S(k N )] T ; Φ is the sparse dictionary matrix, Φ=[φ(k1), φ(k2), …, φ(k N )] T , φ(k n )(n=1, …N) is exp(-j2k n rl ) form an L-dimensional column vector; r l ) is the radial position r m all possible values, E is noise.

[0074] For step A2, in some embodiments, based on the sparse representation form, the sample index set matrix of the sample echo data is calculated using the orthogonal matching pursuit method, comprising:

[0075] S1, based on the sparse representation form, setting the orthogonal matching pursuit threshold;

[0076] S2, initializing parameters, the parameters including a residual vector, an index set matrix, an estimated sparse coefficient vector, a support set matrix, and an iteration number;

[0077] S3, according to the parameters of the current time, calculating the element index, and updating the index set matrix, the support set matrix, the estimated sparse coefficient vector and the residual vector according to the calculation result;

[0078] S4, according to the updated residual vector and the orthogonal matching pursuit threshold, judging whether the convergence condition is met; if the condition is met, stopping iteration and executing S5, otherwise updating the iteration number and returning to execute S3;

[0079] S5, taking the updated index set matrix as the sample index set matrix of the sample echo data.

[0080] In some embodiments, in step S1, the expression of the orthogonal matching pursuit threshold a is: a = 0.01 * ||S||2, wherein ||·||2 is the 2-norm of the vector;

[0081] In step S2, the initial residual vector R (0) = S, the initial index set matrix J (0) = Φ, the initial estimated sparse coefficient vector the initial support set matrix Ω (0) = [] and the initial iteration number t = 0;

[0082] In step S2, when calculating the element index , the expression is:

[0083]

[0084] In the formula, <·> is the inner product operator, R (t-1) is the residual vector calculated in the t-1 iteration, Φ(:,j) is the jth column of the sparse dictionary Φ; in this step, by calculating the inner product of the current residual vector R (t-1) and each column of the sparse dictionary Φ, the Φ(:,j) that makes the inner product maximum, i.e. the atom, can be found.

[0085] In step S2, when updating the index set matrix, the expression is:

[0086] In step S2, when updating the support set matrix, the expression is: (t) (t-1)

[0087] In step S2, when updating the estimated sparse coefficient vector, the expression is:

[0088]

[0089] In the formula, the superscript H represents Hermitian transpose;

[0090] In step S2, when updating the residual vector, the expression is:

[0091] In this embodiment, the superscript t represents the result of the tth iteration calculation, and the superscript t-1 represents the result of the (t-1)th iteration calculation.

[0092] In some embodiments, according to the updated residual vector and the orthogonal matching pursuit threshold, it is judged whether the convergence condition is met, including:

[0093] According to the updated residual vector R (t) , the residual signal energy W is calculated, and the expression of the residual signal energy W is: W = ||R (t) ||2;

[0094] It is judged whether the residual signal energy is less than the orthogonal matching pursuit threshold, that is, W < a, if yes, it is determined to be converged, and if no, it is determined to be not converged.

[0095] After the above steps, 65,341 sample pairs can be generated, of which 33,000 are selected as a training set and the rest are selected as a test set.

[0096] In some embodiments, the sparse estimation model includes a first layer neural network, a second layer neural network and a third layer neural network.

[0097] The first layer neural network is used to calculate the inner product of the input radar echo data and each atom, and each atom represents a column in the sparse dictionary. This layer can use a fully connected layer, and the activation function is Relu. The output dimension of this layer is 9L.

[0098] ​​The second layer neural network and the third layer neural network are used to extract atoms with relatively high correlation, and output an index set matrix corresponding to the input radar echo data according to the extraction result. The second layer can adopt a convolution layer, the convolution kernel size is 3*1, the convolution step size is 3*1, and the activation function is sigmoid; the third layer is a pooling layer, the pooling window size is 3*1, and the step size is 3*1.

[0099] After determining the layers of the neural network, a suitable cost function needs to be selected. In some embodiments, a cross-entropy loss function is used as the cost function:

[0100]

[0101] In the formula, y is the output vector of the neural network.

[0102] When training and testing the sparse estimation model, first, the sample echo data in the training set and the test set are input into the model, and the model is trained and tested based on the above cost function. The network optimization method adopts a stochastic gradient descent method, and the parameter update formula corresponding to the stochastic gradient descent method is:

[0103]

[0104] Where θ t is the model parameter after the tth iteration, θ t+1 is the model parameter after the (t+1) th iteration, and η is a learning rate parameter.

[0105] The model is repeatedly trained using samples until the model converges, and a trained sparse estimation model is obtained.

[0106] In some embodiments, according to the index set matrix of the to-be-detected target radar echo data, the position of each scattering center in the to-be-detected target and the scattering amplitude coefficient of each scattering center are determined, including:

[0107] The row l D in which the element with a value of 1 in the index set matrix J D of the to-be-detected target radar echo data is determined as the position r

[0108] The row l D in which the element with a value of 1 in the index set matrix J D of the to-be-detected target scattering center is determined as the support set Ω D = [..., Φ(:, l D ),... ] (l D ∈ J D ), and the scattering amplitude coefficient A of each scattering center is:

[0109] For example,Figure 2 , Figure 3 As shown, this embodiment of the invention provides a device for extracting the scattering center of a target. The device embodiment can be implemented through software, hardware, or a combination of both. From a hardware perspective, such as... Figure 2 The diagram shown is a hardware architecture diagram of an electronic device for extracting the scattering center of a target, provided in an embodiment of the present invention. (Except for...) Figure 2 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing packets. Taking software implementation as an example, such as... Figure 3 As shown, a device in a logical sense is formed by the CPU of its electronic device reading the corresponding computer program from non-volatile memory into memory and running it. This embodiment provides a target scattering center extraction device, comprising:

[0110] The acquisition module 300 is used to acquire radar echo data of the target under test;

[0111] The input module 302 is used to input the radar echo data of the target to be measured into the sparse estimation model to obtain the index set matrix of the radar echo data of the target to be measured. The sparse estimation model is trained using radar echo sample pairs of several known targets. Each radar echo sample pair includes sample echo data as input and sample index set matrix as output. The sample echo data includes at least one scattering center, and the sample index set matrix is ​​calculated based on the sample echo data.

[0112] The determination module 304 is used to determine the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target under test based on the index set matrix of the radar echo data of the target under test.

[0113] In this embodiment of the invention, the acquisition module 300 can be used to execute step 100 in the above method embodiment, the input module 302 can be used to execute step 102 in the above method embodiment, and the determination module 304 can be used to execute step 104 in the above method embodiment.

[0114] In some implementations, the sample index set matrix is ​​calculated as follows:

[0115] Based on the parameterized model of point scattering centers, a sparse representation of radar target scattering centers is constructed.

[0116] Based on the sparse representation, the sample index set matrix of the sample echo data is calculated using the orthogonal matching pursuit method.

[0117] In some implementations, the parameterized model of the point scattering center is:

[0118]

[0119] where S(k n ) is the radar echo signal, k n is the wave number, n = 1, 2, …, N, N is the number of radar frequency sampling points, m = 1, 2, …, M, M is the number of scattering centers, A m is the complex scattering amplitude coefficient of the mth scattering center, r m is the radial position of the mth scattering center, e n is the measurement noise;

[0120] The sparse expression form of the radar target scattering center is:

[0121] S = Φσ + E

[0122] where S is the target observation echo column vector, S = [S(k1), S(k2), …, S(k N )] T ; Φ is a sparse dictionary matrix, Φ = [φ(k1), φ(k2), …, φ(k N )] T , φ(k n )(n = 1, … N) is an L-dimensional column vector composed of exp(-j2k n r l )(l = 1, …, L); r l (l = 1, …, L) is the radial position r m of all possible values, and E is the noise.

[0123] In some embodiments, based on the sparse expression form, the sample index set matrix of the sample echo data is calculated by using the orthogonal matching pursuit method, including:

[0124] S1, based on the sparse expression form, setting the orthogonal matching pursuit threshold;

[0125] S2, initializing parameters, the parameters including a residual vector, an index set matrix, an estimated sparse coefficient vector, a support set matrix, and an iteration number;

[0126] S3, according to the parameters of the current time, calculating the element index, and updating the index set matrix, the support set matrix, the estimated sparse coefficient vector, and the residual vector according to the calculation result;

[0127] S4, according to the updated residual vector and the orthogonal matching pursuit threshold, judging whether the convergence condition is met; if the condition is met, stopping iteration, and performing S5, otherwise updating the iteration number, and returning to perform S3;

[0128] S5, taking the updated index set matrix as a sample index set matrix of the sample echo data.

[0129] In some embodiments, in S1, the expression of the orthogonal matching pursuit threshold a is: a = 0.01 * ||S|| 2, wherein ||·|| 2 is the 2-norm of a vector;

[0130] In S2, the initial residual vector R (0) = S, the initial index set matrix J (0) = Φ, the initial estimated sparse coefficient vector The initial support set matrix Ω (0) = [] and the initial iteration time t = 0.

[0131] In S2, the element index j is calculated, and the expression is:

[0132]

[0133] In the formula, <·> is an inner product operator, R (t-1) is the residual vector calculated in the t-1th iteration, and Φ(:,j) is the jth column of the sparse dictionary Φ.

[0134] In S2, when the index set matrix is updated, the expression is:

[0135] In S2, when the support set matrix is updated, the expression is: Ω (t) = [Ω (t-1) , Φ(:,j)];

[0136] In S2, when the estimated sparse coefficient vector is updated, the expression is:

[0137]

[0138] In the formula, the superscript H represents Hermitian transpose;

[0139] In S2, when the residual vector is updated, the expression is:

[0140] In some embodiments, according to the updated residual vector and the orthogonal matching pursuit threshold, it is judged whether the convergence condition is met, comprising:

[0141] According to the updated residual vector R (t) , the residual signal energy W is calculated, and the expression of the residual signal energy W is: W = ||R (t) || 2.

[0142] It is judged whether the residual signal energy is less than the orthogonal matching pursuit threshold, if yes, it is determined to be converged, if not, it is determined to be not converged.

[0143] In some embodiments, the determining module 304 is configured to perform the following operations:

[0144] The index set matrix J of the to-be-detected target radar echo data D The row l in which the element with a value of 1 is located D The position r of the scattering center is determined,

[0145] The index set matrix J of the to-be-detected target scattering center D The row l in which the element with a value of 1 is located D The support set is Ω D = [..., Φ ( :, l D ),... ] (l D ∈ J D ), the scattering amplitude coefficient A of each scattering center is:

[0146] In some embodiments, the sparse estimation model comprises a first layer neural network, a second layer neural network and a third layer neural network.

[0147] The first layer neural network is configured to calculate the inner product of the input radar echo data and each atom, and each atom represents a column in the sparse dictionary.

[0148] The second layer neural network and the third layer neural network are configured to extract atoms with relatively large correlations, and output an index set matrix corresponding to the input radar echo data according to the extraction result.

[0149] It can be understood that the structure shown in the embodiments of the present application does not constitute a specific limitation on the target scattering center extraction device. In other embodiments of the present application, a target scattering center extraction device can include more or fewer components than the illustration, or combine certain components, or split certain components, or different component arrangements. The components shown can be implemented in hardware, software or a combination of software and hardware.

[0150] The information interaction, execution process and the like between the modules in the above device, since based on the same concept as the method embodiments of the present application, the specific content can be referred to the description in the method embodiments of the present application, which will not be repeated here.

[0151] The embodiments of the present application also provide an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the target scattering center extraction method in any of the embodiments of the present application.

[0152] The embodiment of the present application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program enables a processor to execute the target scattering center extraction method in any of the embodiments of the present application when the computer program is executed by the processor.

[0153] Specifically, a system or device provided with a storage medium storing software program codes for realizing the functions of any of the above embodiments can be provided, and the computer (or CPU or MPU) of the system or device reads out and executes the program codes stored in the storage medium.

[0154] In this case, the program codes read out from the storage medium can realize the functions of any of the above embodiments by themselves, and therefore the program codes and the storage medium storing the program codes constitute a part of the present application.

[0155] The storage medium for providing the program codes includes a floppy disk, a hard disk, a magneto-optical disk (for example, CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program codes can be downloaded from a server computer via a communication network.

[0156] In addition, it should be understood that not only the program codes read out by the computer, but also the operating system or the like operating on the computer can perform part or all of the actual operations based on the instructions of the program codes, so as to realize the functions of any of the above embodiments.

[0157] In addition, it should be understood that the program codes read out by the storage medium can be written into the memory provided in the expansion board inserted into the computer or the memory provided in the expansion module connected to the computer, and then part or all of the actual operations can be performed based on the instructions of the program codes by the CPU or the like installed on the expansion board or the expansion module, so as to realize the functions of any of the above embodiments.

[0158] It should be noted that, in the present document, the terms such as first and second are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. In addition, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device.

[0159] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the foregoing storage medium includes various storage media that can store program codes, such as ROM, RAM, magnetic disk or optical disk.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for extracting the scattering center of a target, characterized in that, include: Acquire radar echo data of the target under test; The radar echo data of the target to be measured is input into the sparse estimation model to obtain the index set matrix of the radar echo data of the target to be measured. The sparse estimation model is trained using radar echo sample pairs of several known targets. Each radar echo sample pair includes sample echo data as input and a sample index set matrix as output. The sample echo data includes at least one scattering center, and the sample index set matrix is ​​calculated based on the sample echo data. Based on the index set matrix of the radar echo data of the target under test, the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target under test are determined.

2. The method according to claim 1, characterized in that, The sample index set matrix is ​​calculated as follows: Based on the parameterized model of the point scattering center, a sparse representation of the radar target scattering center is constructed. Based on the sparse representation, the sample index set matrix of the sample echo data is calculated using the orthogonal matching pursuit method.

3. The method according to claim 2, characterized in that, The parameterized model of the point scattering center is: In the formula, S(k) n ) represents the radar echo signal, k n Let n be the wave number, n = 1, 2, ..., N, where N is the number of radar frequency sampling points, m = 1, 2, ..., M, where M is the number of scattering centers, and A m r is the complex scattering amplitude coefficient of the m-th scattering center. m Let e ​​be the radial position of the m-th scattering center. n For measuring noise; The sparse representation of the radar target's scattering center is as follows: S=Φσ+E In the formula, S is the target observation echo column vector, S=[S(k1),S(k2),…,S(k N )] T ; Φ is a sparse dictionary matrix, Φ=[φ(k1),φ(k2),…,φ(k N )] T ,φ(k n ) is derived from exp(-j2k n r l An L-dimensional column vector consisting of n = 1, ..., N, l = 1, ..., L; r l Radial position r m All possible values, E represents noise.

4. The method according to claim 3, characterized in that, The step of calculating the sample index set matrix of the sample echo data using the orthogonal matching pursuit method based on the sparse representation includes: S1. Based on the sparse representation, set an orthogonal matching tracking threshold; S2. Initialize parameters, including residual vector, index set matrix, estimated sparse coefficient vector, support set matrix, and number of iterations; S3. Calculate the element index based on the current parameters, and update the index set matrix, support set matrix, estimated sparse coefficient vector, and residual vector based on the calculation results. S4. Based on the updated residual vector and the orthogonal matching tracking threshold, determine whether the convergence condition is met; if the condition is met, stop the iteration and execute S5; otherwise, update the iteration count and return to execute S3. S5. Use the updated index set matrix as the sample index set matrix for the sample echo data.

5. The method according to claim 4, characterized in that, In S1, the expression for the orthogonal matching tracking threshold a is: a = 0.01 * ||S||2, where ||·||2 is the vector 2 norm; In S2, the initial residual vector R (0) =S, Initial index set matrix J (0) =Φ, Initial estimated sparse coefficient vector Initial support set matrix Ω (0) =[] and the initial iteration t=0; In step S2, the element index is calculated. When, the expression is: In the formula, <·> is the inner product operator, and R (t-1) Let Φ be the residual vector calculated in the (t-1)th iteration, and let Φ(:,j) be the j-th column of the sparse dictionary Φ. In S2, when updating the index set matrix, the expression is: In S2, when updating the support set matrix, the expression is: Ω (t) =[Ω (t-1) ,Φ(:,j)]; In S2, when updating the estimated sparse coefficient vector, the expression is: In the formula, the superscript H represents the Hermitian transpose; In S2, when updating the residual vector, the expression is:

6. The method according to claim 5, characterized in that, The step of determining whether the convergence condition is met based on the updated residual vector and the orthogonal matching pursuit threshold includes: Based on the updated residual vector R (t) Calculate the remaining signal energy W, which is expressed as: W = ||R (t) ||2; Determine whether the remaining signal energy is less than the orthogonal matched tracking threshold. If yes, it is considered converged; otherwise, it is considered non-converged.

7. The method according to claim 6, characterized in that, The step of determining the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target radar echo data based on the index set matrix of the target radar echo data includes: The index set matrix J of the radar echo data of the target under test D The row containing the element with a value of 1 D The location r of the scattering center is determined. l D ∈J D ; The target scattering center index set matrix J D The row containing the element with a value of 1 D The support set is Ω D =[…,Φ(:,l D ),...],l D ∈J D Then the scattering amplitude coefficient A of each scattering center is:

8. The method according to claim 1, characterized in that, The sparse estimation model includes a first layer neural network, a second layer neural network, and a third layer neural network; The first layer of the neural network is used to calculate the inner product of the input radar echo data and each atom, where each atom represents a column in the sparse dictionary; The second and third neural networks are used to extract atoms with high correlation and output an index set matrix corresponding to the input radar echo data based on the extraction results.

9. A device for extracting the scattering center of a target, characterized in that, include: The acquisition module is used to acquire radar echo data of the target under test; The input module is used to input the radar echo data of the target under test into the sparse estimation model to obtain the index set matrix of the radar echo data of the target under test; The sparse estimation model is trained using radar echo sample pairs of several known targets. Each radar echo sample pair includes sample echo data as input and a sample index set matrix as output. The sample echo data includes at least one scattering center, and the sample index set matrix is ​​calculated based on the sample echo data. The determination module is used to determine the position of each scattering center and the scattering amplitude coefficient of each scattering center in the target under test based on the index set matrix of the radar echo data of the target under test.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-8.

Images