Anti-mainlobe and sidelobe interference adaptive monopulse angle finding method for uniform linear array

Abstract

The application relates to an anti-mainlobe interference adaptive monopulse angle measurement method for a uniform linear array, which comprises the following steps: obtaining array echo data and calculating a covariance matrix; estimating the number of signal sources and the mainlobe interference angle by using the covariance matrix; constructing a mainlobe interference blocking matrix by using the mainlobe interference angle, and preprocessing the array echo data to obtain preprocessed array data; projecting a corresponding static steering vector of a beam direction to a signal subspace to obtain a sum beam projection vector; performing Wiener filtering on the preprocessed array data by using the sum beam projection vector to solve sum beam adaptive weights; calculating a difference beam projection vector by using the sum beam projection vector, and solving difference beam adaptive weights by using the difference beam projection vector; and calculating a target deflection angle by using the sum beam adaptive weights and the difference beam adaptive weights and adopting an amplitude comparison monopulse method. The method has the advantages of high anti-mainlobe interference capability and high angle measurement precision of a target near the mainlobe interference.

Description

Technical Field

[0001] This invention belongs to the field of signal and information processing technology, and specifically relates to an adaptive single-pulse angle measurement method for uniform linear arrays to resist main and sidelobe interference. Background Technology

[0002] With the increasingly fierce electromagnetic warfare environment, the emergence of mainlobe jamming (MLJ) poses a challenge to adaptive monopulse angle measurement technology. When countering mainlobe jamming, conventional adaptive monopulse angle measurement technology suffers from deep nulls formed by adaptive digital beamforming (ADBF) at the mainlobe jamming location. These nulls cause a shift in the sum (difference) beam pointing (null), resulting in severe distortion of the adaptive monopulse rate culve (MRC) near the beam pointing. In this situation, adaptive monopulse angle measurement technology cannot provide effective angle estimation results.

[0003] Currently, mature adaptive monopulse angle measurement methods for resisting main lobe and side lobe interference are mainly divided into two categories: data preprocessing-based and linearly constrained adaptive monopulse angle measurement algorithms. These algorithms can suppress main lobe interference while ensuring the validity of the angle measurement results. However, existing algorithms are implemented under the premise that the target signal is not present in the echo data. In reality, due to the inability to accurately know the true angle of the target and the limited length of the training samples, when the echo data contains the target signal, the algorithm will fail to measure the angle due to severe signal cancellation. Main lobe conformal processing is usually required to alleviate this problem. Moreover, although conformal processing can alleviate the signal cancellation phenomenon and ensure the validity of the angle measurement results, when the target is near the main lobe interference, the signal-to-noise ratio of the sum-difference adaptive channel output is too low. In this case, both types of algorithms measure the target deflection as 0, and the angle measurement result does not match the true direction of the target. At this time, the angle measurement accuracy is seriously insufficient. Summary of the Invention

[0004] To address the aforementioned problems in the prior art, this invention provides an adaptive single-pulse angle measurement method for uniform linear arrays, resistant to main and sidelobe interference. The technical problem to be solved by this invention is achieved through the following technical solution:

[0005] This invention provides an adaptive single-pulse angle measurement method for uniform linear arrays to resist main and sidelobe interference, comprising the following steps:

[0006] S1. Acquire array echo data and calculate the covariance matrix based on the array echo data;

[0007] S2. Use the covariance matrix to estimate the number of information sources and the main lobe interference angle;

[0008] S3. Using a pairwise cancellation method, a main lobe interference blocking matrix is ​​constructed using the main lobe interference angle, and the array echo data is preprocessed using the main lobe interference blocking matrix to obtain preprocessed array data.

[0009] S4. Construct a signal subspace using the covariance matrix and the number of sources, project the static steering vector corresponding to the beam pointing onto the signal subspace, and discard the last element of the beam steering vector after the subspace projection to obtain the beam projection vector.

[0010] S5. Using a reduced-rank generalized sidelobe cancellation structure, Wiener filtering is performed on the preprocessed array data using the sum beam projection vector to solve for the sum beam adaptive weights.

[0011] S6. The difference beam projection vector is calculated by using the symmetric inversion method on the sum beam projection vector, and the difference beam adaptive weights are solved by using the difference beam projection vector with the rank-reduced generalized sidelobe cancellation structure.

[0012] S7. Using the adaptive sum beam weights and the adaptive difference beam weights, construct adaptive sum channels and adaptive difference channels, and combine the adaptive sum channels and the adaptive difference channels to calculate the target deflection angle using the amplitude comparison single pulse method.

[0013] In one embodiment of the present invention, the main lobe interference blocking matrix is:

[0014]

[0015] Among them, u mlj u represents the wavenumber vector corresponding to the incident angle of the main lobe interference. mlj = -2πsin(θ) mlj ) / λ, θ mlj λ represents the main lobe interference angle, d represents the element spacing, and λ represents the radar operating wavelength.

[0016] The preprocessed array data is as follows:

[0017] y(k) = Bx(k).

[0018] In one embodiment of the present invention, step S4 includes:

[0019] Eigenvalue decomposition of the covariance matrix using the number of sources:

[0020]

[0021] Where N represents the number of array elements, M represents the number of information sources, and λ n σ represents the eigenvalues ​​of the first M largest covariance matrices. n Represents the remaining noise eigenvalues, un The eigenvectors of the covariance matrix are represented. Let U = [u1, ..., u] represent N eigenvalues ​​arranged in descending order. M ] represents the eigenvector corresponding to the first M largest eigenvalues;

[0022] The space spanned by the eigenvectors corresponding to the first M largest eigenvalues ​​is taken as the signal subspace:

[0023] P = UU H ;

[0024] Projecting the beam pointing to the corresponding static steering vector onto the signal subspace yields the beam steering vector projected into the subspace:

[0025] a p =Pa(θ0)=UU H a(θ0)

[0026] Where θ0 represents the beam pointing, and a(θ0) represents the static steering vector corresponding to the beam pointing;

[0027] Discarding the last element of the beam steering vector after subspace projection, we obtain the sum of the beam projection vectors:

[0028] w q ={a p} 1，...,N-1

[0029] in,{*} 1，...,N-1 This represents the first N-1 elements of the vector.

[0030] In one embodiment of the present invention, step S5 includes:

[0031] The preprocessed array data is weighted using the sum and beam projection vectors to obtain the target data for the first upper branch:

[0032]

[0033] A first lower branch blocking matrix is ​​constructed using the beam projection vector, and the preprocessed array data is processed using the first lower branch blocking matrix to obtain first and second-processed array data:

[0034] x0(k)=B0y(k)=B0Bx(k)

[0035]

[0036] Where B0 represents the first lower branch blocking matrix, I represents the identity matrix, and I has a size of (N-1)×(N-1);

[0037] Eigenvalue decomposition is performed on the covariance matrix of the first and second processed array data to obtain the first decomposition feature:

[0038]

[0039] in, Let x0(k) be the covariance matrix. The size is (N-1)×(N-1), where N represents the number of array elements, M represents the number of information sources, and λ n σ represents the eigenvalues ​​of the first M-1 largest covariance matrices corresponding to the number of information sources. n Represents the remaining noise characteristic values, v n The eigenvectors representing the covariance matrix;

[0040] The M-1 largest eigenvalues ​​in the first decomposition feature that are distinct from the noise eigenvalues ​​are used as the first rank reduction transformation matrix:

[0041] T = V H =[v1,…,v M-1 ] H ;

[0042] The first rank reduction transformation matrix is ​​used to reduce the rank of the first secondary processing array data to obtain the first lower branch interference data:

[0043] z(k)=Tx0(k)=TB0y(k)=TB0Bx(k);

[0044] According to Wiener filtering theory, the interference data of the first lower branch and the target data of the first upper branch are subjected to Wiener filtering, and then the adaptive weights of the first lower branch are solved:

[0045]

[0046] Among them, R z Let r represent the autocorrelation matrix of the first lower branch interference data. zd This represents the cross-correlation vector between the interference data of the first lower branch and the target data of the first upper branch.

[0047] The adaptive weights of the sum beam are solved using the first lower branch adaptive weights, the first rank reduction transformation matrix, the first lower branch blocking matrix, the sum beam projection vector, and the main lobe interference blocking matrix.

[0048] w sum =B H (w q -(TB0) H w z ).

[0049] In one embodiment of the present invention, the method for constructing the first lower branch blocking matrix includes the null space method, the eigenvector method, or the household method.

[0050] The first rank reduction transformation matrix can be constructed using methods such as principal component method, cross spectrum method, or multi-level Wiener filtering method.

[0051] In one embodiment of the present invention, step S6 includes:

[0052] The difference beam projection vector is calculated using the symmetric inversion method on the sum beam projection vector:

[0053]

[0054] Where ⊙ represents Haddonian accumulation;

[0055] The preprocessed array data is weighted using the difference beam projection vector to obtain the second upper branch target data:

[0056]

[0057] A second lower branch blocking matrix is ​​constructed using the difference beam projection vector, and the preprocessed array data is processed using the second lower branch blocking matrix to obtain second secondary processed array data:

[0058] x0′(k)=B0′y(k)=B0′Bx(k)

[0059]

[0060] Where B0′ represents the second lower branch blocking matrix, I represents the identity matrix, and I has a size of (N-1)×(N-1);

[0061] Eigenvalue decomposition is performed on the covariance matrix of the second-order processed array data to obtain the second decomposition features:

[0062]

[0063] in, Let x0′(k) be the covariance matrix. The size is (N-1)×(N-1), where N represents the number of array elements, M represents the number of information sources, and λ′ n Let v' represent the first M-1 largest covariance matrix eigenvalues ​​corresponding to the number of information sources, σ' represent the remaining noise eigenvalues, and v' represent the remaining noise eigenvalues. n The eigenvectors representing the covariance matrix;

[0064] The M-1 largest eigenvalues ​​in the second decomposition feature that are distinct from the noise eigenvalues ​​are used as the second rank reduction transformation matrix:

[0065] T′=V′ H =[v′1,…,v′ M-1 ] H

[0066] The second rank reduction transformation matrix is ​​used to reduce the rank of the second secondary processing array data to obtain the second lower branch interference data:

[0067] z′(k)=T′x0′(k)=T′B′0y(k)=T′B′0Bx(k)

[0068] Based on Wiener filtering theory, the interference data of the second lower branch and the target data of the second upper branch are subjected to Wiener filtering, and then the adaptive weights of the second lower branch are solved:

[0069]

[0070] Where R′ represents the autocorrelation matrix of the second lower branch interference data, r′ zd This represents the cross-correlation vector between the interference data of the second lower branch and the target data of the second upper branch.

[0071] The differential beam adaptive weights are solved using the second lower branch adaptive weights, the second rank reduction transformation matrix, the second lower branch blocking matrix, the differential beam projection vector, and the main lobe interference blocking matrix.

[0072] w diff =B H (w q-diff -(TB0) H w z ).

[0073] In one embodiment of the present invention, the method for constructing the second lower branch blocking matrix includes the null space method, the eigenvector method, or the household method;

[0074] The second rank reduction transformation matrix can be constructed using methods such as principal component method, cross spectrum method, or multi-level Wiener filtering method.

[0075] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0076] The angle measurement method of this invention first preprocesses the array echo data using a main lobe interference blocking matrix, then projects the static steering vector corresponding to the beam pointing onto the signal subspace. Based on this, a reduced-rank generalized sidelobe cancellation structure is used to solve for the sum and difference adaptive beam weights. Finally, the target deflection angle is calculated using the sum and difference adaptive beam weights. Compared with the traditional adaptive single-pulse angle measurement method against main and sidelobe interference, this method not only avoids signal cancellation when the echo data contains the target signal, ensuring the validity of the angle measurement results, but also ensures that the single-pulse comparison angle curve formed by the sum and difference adaptive beam weights will not be distorted near the main lobe interference angle. The deflection angle position at its zero-value response can adaptively change with the target incident angle, thereby correcting the mismatch error between the beam pointing and the true target angle to a certain extent. This method has the advantages of strong anti-main lobe interference capability and high angle measurement accuracy for targets near main lobe interference. Attached Figure Description

[0077] Figure 1 This is a flowchart illustrating an adaptive single-pulse angle measurement method for a uniform linear array to resist main and side lobe interference, provided in an embodiment of the present invention.

[0078] Figure 2 A schematic diagram of the coordinate system definition method provided in an embodiment of the present invention;

[0079] Figure 3 This is a schematic diagram of the process for solving and adaptive beam weights provided in an embodiment of the present invention;

[0080] Figure 4 This is a schematic diagram of the process for solving the differential beam adaptive weights provided in an embodiment of the present invention;

[0081] Figure 5 This is a structural diagram of the main lobe interference blocking matrix provided in an embodiment of the present invention;

[0082] Figure 6 This is a schematic diagram illustrating the correction effect of the subspace projection algorithm provided in this embodiment of the invention on angle mismatch errors of different degrees.

[0083] Figures 7a-7d These are the sum and difference projected beam vector patterns after projection processing, showing the beam direction of the target when it is incident from -3.2°, -1°, 1° and 3.2° respectively.

[0084] Figure 8 This is an example of an adaptive direction pattern for sums and differences provided in an embodiment of the present invention.

[0085] Figure 9 This is a curve showing the adaptive single-pulse ratio characteristic obtained from the sum-difference adaptive radiation pattern;

[0086] Figure 10This is a schematic diagram showing the angle measurement results of the adaptive single-pulse angle measurement method provided in this embodiment of the invention for targets with different incoming wave directions;

[0087] Figure 11 This refers to the sum and difference adaptive direction pattern of traditional adaptive single-pulse angle measurement technology;

[0088] Figure 12 The adaptive single-pulse ratio characteristic curve is for traditional adaptive single-pulse angle measurement technology;

[0089] Figure 13 This is a schematic diagram illustrating the change in angle measurement results of targets with different incoming wave directions as a function of the target's incident angle using traditional adaptive single-pulse angle measurement technology.

[0090] Figures 14a-14b Adaptive sum-difference patterns for two improved adaptive monopulse angle measurement techniques;

[0091] Figures 15a-15b The adaptive single-pulse ratio characteristic curves of two improved adaptive single-pulse angle measurement techniques are shown.

[0092] Figures 16a-16b The diagram illustrates the variation of angle measurement results with the target incident angle for two improved adaptive single-pulse angle measurement techniques.

[0093] Figure 17 This diagram illustrates the angle measurement results of the adaptive single-pulse angle measurement method provided in this embodiment of the invention and the traditional method for targets with different incoming wave directions. Detailed Implementation

[0094] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0095] Example 1

[0096] Please see Figure 1 , Figure 1 This is a flowchart illustrating an adaptive single-pulse angle measurement method for uniform linear arrays to resist main and side lobe interference, provided in an embodiment of the present invention. The method includes the following steps:

[0097] S1. Obtain array echo data and calculate the covariance matrix based on the array echo data.

[0098] In practical applications, the echo data received by the antenna is processed through signal sampling, digital down-conversion, and amplitude and phase error correction to obtain the required array echo data x(k).

[0099] Please see Figure 2 , Figure 2 This is a schematic diagram of a coordinate system definition method provided in an embodiment of the present invention. Figure 2In this embodiment, a one-dimensional linear array is placed on the x-axis, and the angle between the direction of arrival of the target wave and the normal direction of the linear array is defined as the angle of arrival θ. In this embodiment, let λ represent the radar operating wavelength, then the direction of arrival θ... m The corresponding wavenumber vector can be represented as u m = -2πsin(θ) m Let the element spacing of a uniform linear array (ULA) be d and the number of elements be N, then the direction of the incoming wave θ is... m The steering vector can be expressed as:

[0100] a(θ0)=[1,exp(jdu m ),...,exp(j(N-1)du m )] T .

[0101] Assume there are M far-field signals in space, including one target, one main lobe interference, and M-2 side lobe interferences, s m (k) represents the complex envelope of the m-th signal, and noise represents the additive white Gaussian noise of the array. Then, the echo data x(k) received by the array in a certain snapshot sampling is:

[0102]

[0103] Where x(k) represents the array echo data sampled in the kth snapshot, M represents the number of sources, and s m (k) represents the complex envelope of the m-th signal, θ m Let a(θ) represent its angle of incidence. m ) represents the signal steering vector, a(θ) m The structure is [1, exp(jdu)] m ),...,exp(j(N-1)du m )] T N represents the number of array elements, d represents the spacing between array elements, and u m = -2πsin(θ) m ) / λ represents the wavenumber vector, λ represents the radar operating wavelength, T represents the transpose, and noise represents the additive white Gaussian noise of the array.

[0104] The covariance matrix of the array echo data is:

[0105]

[0106] Where K represents the number of sampling snapshots, and H represents the conjugate transpose.

[0107] S2. Use the covariance matrix to estimate the number of sources and the main lobe interference angle.

[0108] Specifically, the Akaike information criterion (AIC) method is used to estimate the number of information sources using the covariance matrix. This method is an existing method and will not be described in detail in this embodiment.

[0109] The Capon method is used to estimate the main lobe jamming (MLJ) angle using the covariance matrix:

[0110]

[0111] Wherein, the main beam is the angular range within 3dB of the array pattern response, a(θ) represents the array steering vector with direction θ, and R x Let θ represent the covariance matrix. The value of the independent variable θ corresponding to the maximum value in the above equation is the main lobe interference angle θ. mlj .

[0112] S3. Using a pairwise cancellation method, a main lobe interference blocking matrix is ​​constructed using the main lobe interference angle, and the array echo data is preprocessed using the main lobe interference blocking matrix to obtain preprocessed array data.

[0113] Specifically, a pairwise cancellation method is used, utilizing the main lobe interference angle θ. mlj Construct the main lobe interference blocking matrix B (N-1)×N Its expression is:

[0114]

[0115] Among them, u mlj u represents the wavenumber vector corresponding to the incident angle of the main lobe interference. mlj = -2πsin(θ) mlj ) / λ, θ mlj This indicates the angle of interference on the main lobe.

[0116] The preprocessed array data obtained by preprocessing the array echo data using the main lobe interference blocking matrix is ​​as follows:

[0117] y(k)=Bx(k).

[0118] S4. Construct a signal subspace using the covariance matrix and the number of sources, project the static steering vector corresponding to the beam pointing onto the signal subspace, and discard the last element of the beam steering vector after the subspace projection to obtain the beam projection vector.

[0119] In this embodiment, the covariance matrix is ​​decomposed using eigenvalues ​​to construct the signal subspace P. Specifically, the covariance matrix is ​​decomposed using the number of sources:

[0120]

[0121] Where N represents the number of array elements, M represents the number of information sources, and λ n σ represents the eigenvalues ​​of the first M largest covariance matrices. n Represents the remaining noise eigenvalues, u n The eigenvectors of the covariance matrix are represented. Let U = [u1, ..., u] represent N eigenvalues ​​arranged in descending order. M ] represents the eigenvectors corresponding to the first M largest eigenvalues.

[0122] The eigenvectors corresponding to the first M largest eigenvalues ​​are U = [u1, ..., u2]. M The space spanned by the target and interference steering vectors is called the signal subspace P (signal subspace P belongs to the same space as the space spanned by the target and interference steering vectors), and its expression is:

[0123] P = UU H .

[0124] Let θ0 be the beam pointing direction, and a(θ0) be the static steering vector corresponding to the beam pointing direction. Projecting the static steering vector a(θ0) corresponding to the beam pointing direction onto the signal subspace yields the projected beam steering vector a. p :

[0125] a p =Pa(θ0)=UU H a(θ0)

[0126] Where θ0 represents the beam pointing, and a(θ0) represents the static steering vector corresponding to the beam pointing.

[0127] Discarding the last element of the beam steering vector after subspace projection, we obtain the beam projection vector w. q :

[0128] w q ={a p} 1，...,N-1

[0129] in,{*} 1，...,N-1 This represents the first N-1 elements of the vector.

[0130] S5. Using a reduced-rank generalized sidelobe cancellation (RR-GSC) structure, the preprocessed array data is optimally filtered using the sum-beam projection vector to solve for the sum-beam adaptive weights.

[0131] Please see Figure 3 , Figure 3 This is a schematic diagram illustrating the process of solving and adaptive beam weights provided in an embodiment of the present invention.

[0132] Specifically, using the beam projection vector w q The preprocessed array data y(k) is weighted to obtain the target data of the first upper branch. In the RR-GSC structure, the target data d(k) of the first upper branch can be expressed as:

[0133]

[0134] Then, using the null space method, the first lower branch blocking matrix B0 is constructed using the sum and beam projection vectors:

[0135]

[0136] Where I represents the identity matrix, and I has a size of (N-1)×(N-1).

[0137] It should be noted that the construction of the first lower branch blocking matrix in this embodiment is not limited to the null space method. Other existing methods can also be used to construct the first lower branch blocking matrix B0, such as the eigenvector method and the household method.

[0138] Finally, the first and second processed array data x0(k) are obtained by using the first lower branch blocking matrix B0 to process the preprocessed array data y(k), that is:

[0139] x0(k)=B0y(k)=B0Bx(k).

[0140] Next, the first rank-reduced transformation matrix T is constructed using the principal component method. Specifically, the covariance matrix of the first and second processed array data is:

[0141]

[0142] Eigenvalue decomposition is performed on the covariance matrix of the first and second processed array data to obtain the first decomposition feature:

[0143]

[0144] in, Let x0(k) be the covariance matrix. The size is (N-1)×(N-1), where N represents the number of array elements, M represents the number of information sources, and λ n This represents the eigenvalues ​​of the first M-1 largest covariance matrices corresponding to the number of signal sources (since the main lobe interference signal has been blocked, the number of signal sources is reduced by 1), σ n Represents the remaining noise characteristic values, v n The eigenvectors represent the covariance matrix.

[0145] After eigenvalue decomposition, there exists an eigenvalue σ that is distinct from the noise. 2Find the M-1 largest eigenvalues ​​and use them as the first rank-reducing transformation matrix:

[0146] T = V H =[v1,…,v M-1 ] H .

[0147] It should be noted that the method for constructing the first rank-reduced transformation matrix in this embodiment is not limited to the principal component method. Other existing methods can also be used to construct the first rank-reduced transformation matrix T, such as the cross-spectrum method and the multi-level Wiener filtering method.

[0148] Then, the first secondary processing array data x0(k) is reduced in rank using the first rank reduction transformation matrix T to obtain the first lower branch interference data z(k):

[0149] z(k)=Tx0(k)=TB0y(k)=TB0Bx(k).

[0150] Furthermore, based on Wiener filtering theory, the interference data z(k) of the first lower branch and the target data d(k) of the first upper branch are subjected to Wiener filtering, and then the adaptive weight w of the first lower branch is solved. z :

[0151]

[0152] Among them, R Z Let r represent the autocorrelation matrix of the first lower branch interference data. zd The cross-correlation vector between the interference data of the first lower branch and the target data of the first upper branch is represented; the adaptive weight w of the first lower branch is... z Let z(k) and d(k) be the optimal weight solution expressions.

[0153] Finally, the adaptive weights of the first lower branch are solved using the first rank reduction transformation matrix, the first lower branch blocking matrix, the sum beam projection vector, and the main lobe interference blocking matrix:

[0154] w sum =B H (w q -(TB0) H w z ).

[0155] S6. The difference beam projection vector is calculated using the symmetric inversion method on the sum beam projection vector, and the differential beam adaptive weight w is solved using the difference beam projection vector and a reduced-rank generalized sidelobe cancellation structure. diff .

[0156] Please see Figure 4 , Figure 4This is a schematic diagram of the process for solving the differential beam adaptive weights provided in an embodiment of the present invention.

[0157] Specifically, the difference beam projection weights w are first solved using the symmetric inversion method. q-diff Then, the RR-GSC structure is used to solve for w. diff .

[0158] Using the beam projection vector w q The difference beam projection vector w is calculated using the symmetric inversion method. q-diff The expression is:

[0159]

[0160] Here, ⊙ represents Haddonian.

[0161] Furthermore, using the difference beam projection vector w q-diff Replace the beam projection vector w in step S5 q And by performing the same steps as in step S5, the RR-GSC processing is completed to obtain the differential beam adaptive weight w. diff The specific process includes:

[0162] The preprocessed array data is weighted using the difference beam projection vector to obtain the second upper branch target data:

[0163]

[0164] A second lower branch blocking matrix is ​​constructed using the difference beam projection vector, and the preprocessed array data is processed using the second lower branch blocking matrix to obtain second secondary processed array data:

[0165] x0′(k)=B0′y(k)=B0′Bx(k)

[0166]

[0167] Where B0′ represents the second lower branch blocking matrix, I represents the identity matrix, and I has a size of (N-1)×(N-1);

[0168] Eigenvalue decomposition is performed on the covariance matrix of the second-order processed array data to obtain the second decomposition features:

[0169]

[0170] in, Let x0′(k) be the covariance matrix. The size is (N-1)×(N-1), where N represents the number of array elements, M represents the number of information sources, and λ′ nLet v' represent the first M-1 largest covariance matrix eigenvalues ​​corresponding to the number of information sources, σ' represent the remaining noise eigenvalues, and v' represent the remaining noise eigenvalues. n The eigenvectors representing the covariance matrix;

[0171] The M-1 largest eigenvalues ​​in the second decomposition feature that are distinct from the noise eigenvalues ​​are used as the second rank reduction transformation matrix:

[0172] T′=V′ H =[v′1,…,v′ M-1 ] H

[0173] The second rank reduction transformation matrix is ​​used to reduce the rank of the second secondary processing array data to obtain the second lower branch interference data:

[0174] z′(k)=T′x0′(k)=T′B′0y(k)=T′B′0Bx(k)

[0175] Based on Wiener filtering theory, the interference data of the second lower branch and the target data of the second upper branch are subjected to Wiener filtering, and then the adaptive weights of the second lower branch are solved:

[0176]

[0177] Where R′ represents the autocorrelation matrix of the second lower branch interference data, r′ zd This represents the cross-correlation vector between the interference data of the second lower branch and the target data of the second upper branch.

[0178] The differential beam adaptive weights are solved using the second lower branch adaptive weights, the second rank reduction transformation matrix, the second lower branch blocking matrix, the differential beam projection vector, and the main lobe interference blocking matrix.

[0179] w diff =B H (w q-diff -(T′B′0) H w′ z ).

[0180] It should be noted that the construction of the second lower branch blocking matrix in this embodiment is not limited to the null space method; other existing methods can also be used to construct the second lower branch blocking matrix B0, such as the eigenvector method and the household method. Similarly, the method for constructing the second rank-reduced transformation matrix T is not limited to the principal component method; other existing methods can also be used to construct the second rank-reduced transformation matrix T, such as the cross-spectrum method and the multi-stage Wiener filtering method.

[0181] S7. Construct adaptive sum channels and adaptive difference channels using the sum beam adaptive weights and the difference beam adaptive weights, and calculate the target deflection angle by combining the adaptive sum channels, the adaptive difference channels, and the ideal single-pulse ratio slope.

[0182] Specifically, firstly, the amplitude-comparison single-pulse method is used, utilizing the aforementioned beam adaptive weight w. sum The difference beam adaptive weight w diff Construction and pattern of radiation and difference patterns:

[0183] P sum (θ)=(w sum ) H a(θ)

[0184] P diff (θ)=(w diff ) H a(θ)

[0185] Among them, P sum (θ) represents the pattern and direction, P diff (θ) represents the difference direction pattern, and a(θ) represents the guide vector of the array in different directions.

[0186] Then, using the direction pattern P sum (θ), difference direction pattern P diff (θ) Solve for the single-pulse ratio characteristic curve MRC(θ) and perform a first-order fit to obtain the ideal single-pulse ratio slope r. The process is shown in the following equation:

[0187] MRC(θ)=P diff (θ) / P sum (θ)≈rθ

[0188] Where MRC(θ) represents the single-pulse ratio characteristic curve, r represents the ideal single-pulse ratio slope, and θ represents the target wave direction (incident angle).

[0189] Finally, the aforementioned beam adaptive weights w are utilized. sum The difference beam adaptive weight w diff The array echo data are weighted separately to construct an adaptive sum channel and an adaptive difference channel:

[0190] c sum (k)=w sum H x(k)

[0191] c diff (k)=w diff H x(k)

[0192] Among them, c sum(k) represents adaptive and channel, c diff (k) represents the adaptive difference channel, and x(k) represents the array echo data.

[0193] Then utilize the aforementioned adaptive and channel c sum (k) and the adaptive difference channel c diff (k) Calculate the adaptive difference and ratio MR:

[0194]

[0195] Where K represents the number of sampling snapshots.

[0196] Finally, the target deflection angle is calculated using the ideal single-pulse ratio slope r and the adaptive difference ratio MR:

[0197]

[0198] The angle measurement method in this embodiment first preprocesses the array echo data using the main lobe interference blocking matrix. Then, it projects the static steering vector corresponding to the beam pointing onto the signal subspace. Based on this, it uses a reduced-rank generalized sidelobe cancellation structure to solve for the sum and difference adaptive beam weights. Finally, it calculates the target deflection angle using the sum and difference adaptive beam weights. Compared with the traditional adaptive single-pulse angle measurement method against main and sidelobe interference, this method not only avoids signal cancellation when the echo data contains the target signal, ensuring the validity of the angle measurement results, but also ensures that the single-pulse comparison angle curve formed by the sum and difference adaptive beam weights will not be distorted near the main lobe interference angle. The deflection angle position at its zero-value response can adaptively change with the target incident angle, thereby correcting the mismatch error between the beam pointing and the true target angle to a certain extent. This method has the advantages of strong anti-main lobe interference capability and high angle measurement accuracy for targets near main lobe interference.

[0199] Example 2

[0200] Based on Example 1, this example utilizes the adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-side-lobe interference provided in the invention embodiment for angle measurement, specifically including the following steps:

[0201] S1. Obtain array echo data and calculate the covariance matrix based on the array echo data.

[0202] Specifically, the array and signal parameter settings are shown in Table 1:

[0203] Table 1. Parameter Setting Values

[0204] Number of array elements 16 Element spacing (λ) 1 / 2 Sampling snapshots 64 Beam pointing (°) 0 Incident angle range (°) [-3.2，3.2] Signal-to-noise ratio (dB) 0 Number of interferences 2 Interference incident angle (°) θ = (3, 10) Noise-to-interference ratio (dB) (40,40)

[0205] In this embodiment, the number of array elements is 16, and the element spacing is half a wavelength, resulting in a main lobe width of 6.4°. To demonstrate the improved accuracy of angle measurement results for targets in different directions, this embodiment considers dividing the main lobe range into 33 uniform angles from -3.2° to 3.2° as the incident angles of the target signal. 100 Monte Carlo experiments are performed at each angle, and the average value and root mean square error of the 100 experimental results at the same angle are used to characterize the performance of the adaptive single-pulse angle measurement method provided in this embodiment.

[0206] S2. Use the covariance matrix to estimate the number of sources and the main lobe interference angle.

[0207] S3. Using a pairwise cancellation method, a main lobe interference blocking matrix is ​​constructed using the main lobe interference angle, and the array echo data is preprocessed using the main lobe interference blocking matrix to obtain preprocessed array data.

[0208] Specifically, the estimated main lobe interference angle θ is used. mlj The main lobe interference blocking matrix B is constructed using a pairwise cancellation method. (N-1)×N Please refer to the method Figure 5 , Figure 5 This is a structural diagram of the main lobe interference blocking matrix provided in an embodiment of the present invention. Figure 5 It can be seen that with M array elements, the degrees of freedom after blocking preprocessing is M-1. In particular, when there are two main lobe interferences in space, the blocking matrix is ​​constructed in the same way, except that the loss of array degrees of freedom is 2.

[0209] S4. Construct a signal subspace using the covariance matrix and the number of sources, project the static steering vector corresponding to the beam pointing onto the signal subspace, and discard the last element of the beam steering vector after the subspace projection to obtain the beam projection vector.

[0210] Specifically, the degree of deviation of the target from the beam direction varies depending on the direction of the incoming wave, resulting in different degrees of constraint response mismatch in the adaptive beamforming. Therefore, the average peak (null) position of the sum (difference) beam projection vector pattern after projection processing is first used to demonstrate the effect of the subspace projection algorithm provided in this embodiment on correcting the angle mismatch error. Please refer to [reference needed]. Figure 6 , Figures 7a-7d . Figure 6 This is a schematic diagram illustrating the correction effect of the subspace projection algorithm provided in this embodiment of the invention on angle mismatch errors of different degrees. Figures 7a-7d These are the sum and difference projected beam vector patterns after projection processing, showing the beam patterns of the target incident at -3.2°, -1°, 1°, and 3.2° respectively. Figures 7a-7d The peak response and main lobe notch of the sum and difference pattern were also annotated.

[0211] like Figure 6 As shown, when the target is incident from the angle range of -3.2° to 0°, the correction effect of the subspace projection algorithm is not obvious. However, when the direction of the target wave is located in the angle range of 1.4° to 3°, that is, when the target is close to the main lobe interference, the corrected pointing error is maintained at 1°, the correction effect of the beam pointing error is obvious, and the peak response of the corrected beam projection vector is closer to the direction of the target wave.

[0212] Figures 7a-7d The peak / notch positions of the projection-processed sum and difference pattern are shown in the table below when the target is incident from different angles:

[0213] Table 2. Peak / Notch Positions of Feature Projection Vector and Difference Patterns at Different Target Angles

[0214] Target incident angle (°) -3.2 -1 1 3.2 Peak position (°) 0 0 0.2 2.1 Notch position (°) -0.1 -0.2 0.2 1.9

[0215] When the target is located at -3.2° and -1°, the projection algorithm's effect on correcting angle mismatch is generally poor. However, when the target is located at 1° and 3.2°, the peak / null position of the sum-difference projection weight pattern is located at 0.2°, and the projection processing provides some improvement, such as... Figure 7c As shown; and Figure 7d The values ​​are located at 2.1° and 1.9° respectively. It can be seen that compared with the static weight pointing to 0° when not projected, the angle mismatch has been greatly reduced at this time. This is because the guide vector of the weak target wave is more correlated with the guide vector of the strong main lobe interference wave, so the subspace projection result is better.

[0216] S5. Employing a reduced-rank generalized sidelobe cancellation (RR-GSC) structure, the preprocessed array data is optimally filtered using the sum-beam projection vector, and the adaptive sum-beam weight w is solved. sum The difference beam projection vector is calculated using the symmetric inversion method on the sum beam projection vector, and then the differential beam adaptive weight w is solved using the difference beam projection vector and a rank-reduced generalized sidelobe cancellation structure. diff .

[0217] Specifically, according to w sum w diff This allows us to obtain the sum-difference adaptive radiation pattern and the adaptive monopulse ratio characteristic curve. Please refer to [reference needed]. Figure 8 , Figure 9 Specifically, considering a target incident at 2°, Figure 8 This is a sum-difference adaptive direction pattern provided in an embodiment of the present invention. Figure 9 This is the adaptive single-pulse ratio characteristic curve (also known as the angle discrimination curve or angle error curve) obtained from the sum-difference adaptive direction pattern.

[0218] Figure 8The partial diagram shows the peak response of the adaptive sum pattern of the adaptive single-pulse angle measurement method provided in this embodiment when the target is incident at 2°. It can be seen that no signal cancellation occurs at this time, and due to the correction effect of the subspace projection algorithm, the peak position of the adaptive sum pattern response becomes 1.5°, while the main lobe null of the adaptive difference pattern is located at 0.7°, indicating that the beam pointing of the adaptive sum pattern of the algorithm proposed in Embodiment 1 can adaptively track the current direction of the target. Figure 9 The adaptive single-pulse ratio characteristic curve in the model exhibits high linearity without severe distortion, and its zero point occurs at 0.7°. At this point, the angle corresponding to the first-order fitted curve is 1.7°. The output signal-to-noise ratio and interference noise ratio of the difference beam are low at this time. The output response of the difference and ratio is mainly determined by the array noise, with a mean response of 0. The angle obtained at this point will fluctuate around 1.7°, a result close to the actual target incident angle of 2°.

[0219] S6. Construct adaptive sum channels and adaptive difference channels using the sum beam adaptive weights and the difference beam adaptive weights, and calculate the target deflection angle by combining the adaptive sum channels, the adaptive difference channels, and the ideal single-pulse ratio slope.

[0220] For the angle measurement performance of the adaptive single-pulse angle measurement method provided in this embodiment, please refer to [link to relevant documentation]. Figure 10 , Figure 10 This diagram illustrates the angle measurement results of the adaptive single-pulse angle measurement method provided in this embodiment of the invention for targets with different incoming wave directions. It can be seen that the adaptive single-pulse angle measurement algorithm provided in this embodiment has high angle measurement accuracy for targets within the entire main lobe range; and it overcomes the defect of previous algorithms where, when the target approaches main lobe interference, the adaptive radiation pattern forms a null at the main lobe interference point, causing a sharp drop in output SINR and a mean angle measurement result of 0°, resulting in angle measurement failure.

[0221] Building upon the aforementioned adaptive monopulse angle measurement method for uniform linear arrays to combat main and side lobe interference, this embodiment further demonstrates, through simulation experiments, the performance of traditional adaptive monopulse angle measurement technology in the presence of main lobe interference; and the performance degradation caused by signal cancellation in two improved adaptive monopulse angle measurement methods for combating main lobe interference when the echo data contains target signals, without robust adaptive beamforming (also known as main lobe conformal processing). For specific array and signal parameter settings, please refer to Table 1 in step S1 of this embodiment.

[0222] 1) In the presence of main lobe interference, with the target incident from 0°, please refer to [reference needed]. Figure 11 , Figure 12 and Figure 13 , Figure 11 This refers to the sum and difference adaptive direction pattern of traditional adaptive single-pulse angle measurement technology. Figure 12The adaptive single-pulse ratio characteristic curve is for traditional adaptive single-pulse angle measurement technology. Figure 13 This is a schematic diagram showing the change of angle measurement results for targets with different incoming wave directions as the target's incident angle is represented by traditional adaptive single-pulse angle measurement technology.

[0223] Traditional adaptive monopulse angle measurement techniques use the SMI criterion for optimal beamforming. Figure 11 The partial diagram shows the response of the sum-difference adaptive radiation pattern at the main lobe null in the traditional adaptive monopulse angle measurement technique. It can be seen that when the main lobe interference is present, the sum-difference adaptive radiation pattern is severely distorted, the sum beam direction is significantly shifted, and the difference beam cannot form a null at the beam direction. This indicates that the traditional method severely damages the characteristics of the static weights (such as the concavity of the difference weights and the beam direction of the sum weights) while suppressing the main lobe interference.

[0224] Figure 12 The results show that distortion in the sum-difference adaptive radiation pattern leads to severe distortion in the single-pulse ratio characteristic curve, manifested in its small first-order fitting slope, approximately 0.05 degrees; and a large fitting error. From... Figure 13 It can be seen that the traditional adaptive monopulse angle measurement technology has a large angle measurement error. When the target is located near the beam pointing or main lobe interference, its root mean square error of angle measurement is close to 1.8°. Only when the target is located in the angle range where the fitting error of the monopulse ratio characteristic curve is small, the angle measurement accuracy is high.

[0225] 2) Please see Figures 14a-14b , Figures 15a-15b and Figures 16a-16b , Figures 14a-14b The adaptive sum and difference patterns are shown for two improved adaptive monopulse angle measurement techniques. Figure 14a For data preprocessing, Figure 14b For linear constraint class, Figures 15a-15b The adaptive single-pulse ratio characteristic curves are shown for two improved adaptive single-pulse angle measurement techniques. Figure 15a For data preprocessing, Figure 15b For linear constraint class, Figures 16a-16b The diagram illustrates the variation of angle measurement results with the target incident angle for two improved adaptive single-pulse angle measurement techniques. Figure 16a shows the data preprocessing. Figure 16b It is a linear constraint class.

[0226] When the echo data contains the target signal, we first consider that the target is incident from -3.2°. For the sum-difference adaptive radiation pattern and adaptive single-pulse ratio characteristic curves of two relatively mature adaptive monopulse angle measurement techniques for anti-main-lobe interference without main-lobe conformal processing, please refer to [link to relevant documentation]. Figures 14a-14b , Figures 15a-15b For the angle measurement results of these two techniques for targets with different incoming wave directions, please refer to [link / reference needed]. Figures 16a-16b .

[0227] observe Figure 14a It can be seen that without main lobe conformal processing, the null of the adaptive difference pattern in the preprocessed adaptive monopulse angle measurement method will be severely shifted, with its main lobe null located at -3.1°. This is because the difference pattern forms a null depth at the target incident angle, thus suppressing the concavity of the difference beam at the beam pointing direction. Observation Figure 14b Due to the signal cancellation effect, the sum-difference adaptive radiation pattern of linear constraint algorithms without main lobe shape preservation forms a null at -4°. This causes the sum pattern to be unable to form a main lobe near the beam pointing, while the gain of the difference pattern is very low in the main lobe range, below -30dB, and the adaptive radiation pattern is severely distorted.

[0228] observe Figure 15a and Figure 15b It can be seen that, without main lobe conformal, the adaptive single pulse ratio characteristic curve of the preprocessing type adaptive single pulse angle measurement method is significantly distorted, which is caused by the offset main lobe null. At this time, the fitting error of the adaptive single pulse ratio characteristic curve is large. Although the adaptive single pulse ratio characteristic curve of the linear constraint type adaptive method has high linearity, its dynamic range is too small. At this time, the fluctuation of the single pulse ratio calculation value will lead to a large angle measurement deviation, that is, the root mean square error of the angle measurement result is large.

[0229] observe Figure 16a , Figure 16b It can be observed that, due to signal cancellation, the angle measurement results of both improved algorithms are invalid and cannot reflect the true position of the target to be detected (tracked).

[0230] In summary, when interference enters from the main lobe of the beam, the adaptive monopulse angle measurement technique using conventional adaptive beamforming techniques (such as SMI) will produce a deep notch in the adaptive pattern at the location of the main lobe interference. The main lobe shape, peak pointing, and gain level of the adaptive pattern will inevitably be affected, resulting in severe distortion of the monopulse ratio characteristic curve, making angle measurement impossible. When the echo data contains the target signal, without main lobe conformal processing, the two improved adaptive monopulse angle measurement techniques that specifically handle main lobe interference will experience severe performance degradation due to signal cancellation. The cancellation occurs for two reasons: the target's current direction cannot be accurately predicted, leading to a mismatch in the constraints of the SMI adaptive beamformer; and an insufficient number of snapshots results in the array covariance matrix not being accurately known. Only after applying conformal processing can these two improved adaptive monopulse angle measurement techniques suppress signal cancellation, thus ensuring the validity of the angle measurement results.

[0231] Example 3

[0232] This embodiment compares the performance of the adaptive single-pulse angle measurement method with main lobe shape-preserving processing to resist main lobe interference with that provided in Embodiment 1 through simulation experiments. The specific array and signal parameter settings in this embodiment are the same as in Embodiment 1. For the performance comparison results, please refer to [link to embodiment 1]. Figure 17 , Figure 17 This diagram illustrates the angle measurement results of the adaptive single-pulse angle measurement method provided in this embodiment of the invention and the traditional method for targets with different incoming wave directions.

[0233] Figure 17 In this paper, DL-3P-LCMV represents a conformal constrained adaptive monopulse angle measurement algorithm, BP-DL represents a conformal data preprocessing adaptive monopulse angle measurement algorithm, Traditional-SDC represents a traditional adaptive monopulse angle measurement algorithm, and Block-Proj-RR-GSC represents the adaptive monopulse angle measurement method provided in this embodiment of the invention. It can be seen that within the angle range of -3.2° to 0°, DL-3P-LCMV yields the best angle measurement results. This is because this algorithm does not use a blocking preprocessing matrix to stepwise resist main and side lobe interference, but directly uses constraint conditions to limit the amplitude and phase response of the adaptive difference beam. Its monopulse ratio characteristic curve has the highest linearity in this range, so its root mean square error (RMSE) is the lowest. The other three methods have comparable RMS errors. However, when the direction of the incoming wave is close to the direction of the main lobe interference, except for Block-Proj-RR-GSC, the other three adaptive monopulse methods cannot measure the target angle, and the angle measurement results converge to the vicinity of the beam direction. In this case, the angle measurement results are not reliable. The root mean square error of the angle measurement algorithm proposed in Example 1 did not increase significantly, indicating that the algorithm can effectively measure the angle of targets near the main lobe interference.

[0234] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for adaptive single-pulse angle measurement against main and sidelobe interference for uniform linear arrays, characterized in that, Including the following steps: S1. Acquire array echo data and calculate the covariance matrix based on the array echo data; S2. Use the covariance matrix to estimate the number of information sources and the main lobe interference angle; S3. Using a pairwise cancellation method, a main lobe interference blocking matrix is ​​constructed using the main lobe interference angle, and the array echo data is preprocessed using the main lobe interference blocking matrix to obtain preprocessed array data. S4. Construct a signal subspace using the covariance matrix and the number of sources, project the static steering vector corresponding to the beam pointing onto the signal subspace, and discard the last element of the beam steering vector after the subspace projection to obtain the beam projection vector. S5. Using a reduced-rank generalized sidelobe cancellation structure, Wiener filtering is performed on the preprocessed array data using the sum beam projection vector to solve for the sum beam adaptive weights. S6. The difference beam projection vector is calculated by using the symmetric inversion method on the sum beam projection vector, and the difference beam adaptive weights are solved by using the difference beam projection vector with the rank-reduced generalized sidelobe cancellation structure. S7. Using the adaptive sum beam weights and the adaptive difference beam weights, construct adaptive sum channels and adaptive difference channels, and combine the adaptive sum channels and the adaptive difference channels to calculate the target deflection angle using the amplitude comparison single pulse method.

2. The adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-side-lobe interference according to claim 1, characterized in that, The main lobe interference blocking matrix is: in, u mlj This represents the wavenumber vector corresponding to the incident angle of the main lobe interference. , Indicates the main lobe interference angle. d Indicates the spacing between array elements. Indicates the radar's operating wavelength; The preprocessed array data is as follows: y ( k )= Bx ( k ) in, x ( k ) indicates the first k Array echo data sampled in a quick snapshot.

3. The adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-side-lobe interference according to claim 1, characterized in that, Step S4 includes: Eigenvalue decomposition of the covariance matrix using the number of sources: in, N Indicates the number of array elements. M Indicates the number of information sources. Indicates the preceding M A large covariance matrix eigenvalue, Represents the remaining noise characteristic values. The eigenvectors of the covariance matrix are represented. λ 1> λ 2>…> λ M > ≈…≈ Indicates descending order N 1 eigenvalue, Indicates the preceding M The eigenvectors corresponding to the large eigenvalues; The former M The space spanned by the eigenvectors corresponding to the large eigenvalues ​​is called the signal subspace: ； Projecting the beam pointing to the corresponding static steering vector onto the signal subspace yields the beam steering vector projected into the subspace: in, θ 0 indicates beam direction. a ( θ 0) indicates that the beam is pointing to the corresponding static steering vector; Discarding the last element of the beam steering vector after subspace projection, we obtain the sum of the beam projection vectors: in, The front of the vector N -1 element.

4. The adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-side-lobe interference according to claim 1, characterized in that, Step S5 includes: The preprocessed array data is weighted using the sum and beam projection vectors to obtain the target data for the first upper branch: ； A first lower branch blocking matrix is ​​constructed using the beam projection vector, and the preprocessed array data is processed using the first lower branch blocking matrix to obtain first and second-processed array data: in, B 0 indicates the first lower branch blocking matrix. I Represents the identity matrix. I Size is ( N -1)×( N -1); Eigenvalue decomposition is performed on the covariance matrix of the first and second processed array data to obtain the first decomposition feature: in, express x 0( k The covariance matrix of ) Size is ( N -1)×( N -1), N Indicates the number of array elements. M Indicates the number of information sources. Indicates the number of sources corresponding to the previous M -1 is a larger eigenvalue of the covariance matrix. Represents the remaining noise characteristic values. The eigenvectors representing the covariance matrix; The first decomposition feature is distinguished from the noise feature value. M -1 large eigenvalues ​​are used as the first rank-reducing transformation matrix: ； The first rank reduction transformation matrix is ​​used to reduce the rank of the first secondary processing array data to obtain the first lower branch interference data: z ( k )= Tx 0( k )= TB 0 y ( k )= TB 0 Bx ( k )； According to Wiener filtering theory, the interference data of the first lower branch and the target data of the first upper branch are subjected to Wiener filtering, and then the adaptive weights of the first lower branch are solved: in, R z This represents the autocorrelation matrix of the first lower branch interference data. This represents the cross-correlation vector between the interference data of the first lower branch and the target data of the first upper branch. The adaptive weights of the sum beam are solved using the first lower branch adaptive weights, the first rank reduction transformation matrix, the first lower branch blocking matrix, the sum beam projection vector, and the main lobe interference blocking matrix. 。 5. The adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-side-lobe interference according to claim 4, characterized in that, The methods for constructing the first lower branch blocking matrix include the null space method, the eigenvector method, or the household method; The first rank reduction transformation matrix can be constructed using methods such as principal component method, cross spectrum method, or multi-level Wiener filtering method.

6. The adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-side-lobe interference according to claim 1, characterized in that, Step S6 includes: The difference beam projection vector is calculated using the symmetric inversion method on the sum beam projection vector: Where ⊙ represents Haddonian accumulation; The preprocessed array data is weighted using the difference beam projection vector to obtain the second upper branch target data: A second lower branch blocking matrix is ​​constructed using the difference beam projection vector, and the preprocessed array data is processed using the second lower branch blocking matrix to obtain second secondary processed array data: in, This represents the second lower branch blocking matrix. I Represents the identity matrix. I Size is ( N -1)×( N -1); Eigenvalue decomposition is performed on the covariance matrix of the second-order processed array data to obtain the second decomposition features: in, express The covariance matrix, Size is ( N -1)×( N -1), N Indicates the number of array elements. M Indicates the number of information sources. Indicates the number of sources corresponding to the previous M -1 is a larger eigenvalue of the covariance matrix. Represents the remaining noise characteristic values. The eigenvectors representing the covariance matrix; The second decomposition feature is distinguished from the noise feature value. M -1 large eigenvalues ​​are used as the second rank-reducing transformation matrix: The second rank reduction transformation matrix is ​​used to reduce the rank of the second secondary processing array data to obtain the second lower branch interference data: Based on Wiener filtering theory, the interference data of the second lower branch and the target data of the second upper branch are subjected to Wiener filtering, and then the adaptive weights of the second lower branch are solved: in, This represents the autocorrelation matrix of the second lower branch interference data. This represents the cross-correlation vector between the interference data of the second lower branch and the target data of the second upper branch. The differential beam adaptive weights are solved using the second lower branch adaptive weights, the second rank reduction transformation matrix, the second lower branch blocking matrix, the differential beam projection vector, and the main lobe interference blocking matrix. 。 7. The adaptive single-pulse angle measurement method for uniform linear arrays with anti-main-sidelobe interference according to claim 6, characterized in that, The methods for constructing the second lower branch blocking matrix include the null space method, the eigenvector method, or the household method; The second rank reduction transformation matrix can be constructed using methods such as principal component method, cross spectrum method, or multi-level Wiener filtering method.

Images