Target angle super-resolution estimation method based on array radar

Abstract

The invention relates to a target angle super-resolution estimation method based on array radar. Comprising the steps that array radar subarray distribution is determined, an array radar comprises M rows and N columns of subarrays, and digital zero intermediate frequency echo data of each subarray is acquired; taking pitching dimension angle super-resolution as an example, calculating a digital beam forming weight coefficient according to the distribution condition of the subarrays, and performing digital beam forming on the subarrays in rows to obtain M pieces of pitching dimension beam forming data; carrying out two-dimensional target information extraction on the M pieces of pitching-dimension beam synthesis data to obtain two-dimensional one-time snapshot data; and carrying out decoherence source MUSIC direction finding on the obtained two-dimensional one-time snapshot data to obtain a target pitching incident angle. According to the method, the target angle estimation resolution and angle estimation precision of the array radar in a complex electromagnetic environment can be improved, the stability and reliability in the target tracking process are enhanced, meanwhile, the method mainly depends on the digital signal processing technology, additional hardware resources do not need to be added, and the implementation cost is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of radar signal processing, and in particular to a target angle super-resolution estimation method based on array radar. Background Art

[0002] Modern radars face extremely complex electromagnetic environments. Unfavorable conditions such as stealth aircraft, enemy active interference, multipath reflection, ground and sea clutter, etc., all threaten the performance of existing algorithms. The target angle estimation method of conventional array radar is beam scanning method (CBF). The angular resolution of this method cannot break through the limitation of Rayleigh criterion. Rayleigh criterion means that for two signal sources with equal amplitude in space, if the angle is within a beam width, they are said to be indistinguishable in angle. It is well known that the beam width is inversely proportional to the effective aperture of the array. Therefore, increasing the number of antenna elements in the array or increasing the spacing between the elements can improve the angular resolution. However, in actual engineering applications, increasing the number of elements will increase the cost, and increasing the spacing between the elements (the maximum cannot exceed half the wavelength) will introduce grating lobes. In a complex electromagnetic environment, when two targets are very close (measured by beam width), the conventional beam scanning method can no longer distinguish the angle of the target. In addition, the conventional beam scanning method has a large error in target angle estimation, and there is a possibility of target tracking being unstable or lost. Therefore, it is necessary to obtain accurate angle information of the target in practical applications. Super-resolution direction of arrival estimation is born to meet the above needs.

[0003] Spatial spectrum estimation is a technology of spatial signal processing and a branch of array signal processing. The energy distribution of the signal source in all directions of space can be represented by its spatial spectrum, that is, the direction of arrival (DOA) of the signal can be represented by the spatial spectrum, so spatial spectrum estimation is often called DOA estimation. Spatial spectrum estimation technology applies super-resolution technology to DOA estimation, which can break through the constraint of Rayleigh limit and obtain very high angular resolution. Among them, the super-resolution algorithm is represented by the multiple signal classification (MUSIC) algorithm. The MUSIC algorithm performs eigendecomposition on the covariance matrix of the array received data to obtain the signal subspace and noise subspace, and then uses the orthogonality between the noise subspace and the signal's steering vector to search for spectrum peaks to obtain the signal's arrival angle, which has high angular resolution and angle estimation accuracy. In addition, due to the complex electromagnetic environment in practical applications, the target signal source may be correlated. If the correlation of the target signal source is not decorrelated, the target's direction of arrival cannot be effectively estimated. If the target moves very fast, the angle will change. In practical applications, multiple snapshots are not allowed, but only one snapshot can be obtained, resulting in large target angle estimation errors and unstable or lost target tracking. In view of the practical problems encountered in the above projects, the array radar direction finding system urgently needs an effective array direction finding super-resolution algorithm to perform fast and high-precision direction finding in complex signal environments.

[0004] Therefore, it is necessary to provide a new technical solution to improve one or more problems existing in the above solutions.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present application, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] The purpose of the present application is to provide a target angle super-resolution estimation method based on array radar, thereby overcoming one or more problems caused by limitations and defects of related technologies at least to a certain extent.

[0007] According to an embodiment of the present application, a target angle super-resolution estimation method based on array radar is provided, the method comprising:

[0008] Determine the distribution of array radar subarrays, the array radar includes M rows and N columns of subarrays, obtain the received echo signal of each subarray, and perform digital down-conversion processing to obtain digital zero intermediate frequency echo data;

[0009] Taking the pitch-dimensional angle super-resolution as an example, the weight coefficient of digital beam synthesis is calculated, and the digital zero intermediate frequency echo data is digitally beam synthesized row by row by using the weight coefficient of digital beam synthesis to obtain M pitch beam synthesis data;

[0010] Extracting two-dimensional target information from the M elevation beamformation data respectively to obtain two-dimensional snapshot data;

[0011] The two-dimensional one-shot data is subjected to decorrelated source MUSIC direction finding to obtain the target pitch incident angle.

[0012] In an embodiment of the present application, the received echo signal formula of each subarray is as follows:

[0013] x(t)=s(t)+n(t)(1);

[0014] Where t represents the receiving time, x(t) represents the received echo signal of each subarray, s(t) represents the received target echo signal of interest, and n(t) represents Gaussian white noise;

[0015] The formula of the zero intermediate frequency echo signal after digital down-conversion of each subarray is as follows:

[0016] x l (m) = I l (m)+j×Q l (m)(2);

[0017] Where m represents the number of sampling points in each pulse repetition period, m∈[1,Q], l represents one sampling pulse repetition period, l∈[1,L], x l (m) represents the zero intermediate frequency echo signal after digital down-conversion, I l (m) represents the real part of the zero-IF echo signal after digital down-conversion, Q l (m) represents the imaginary part of the zero intermediate frequency echo signal after digital down conversion, and j represents the complex imaginary part unit.

[0018] In an embodiment of the present application, taking the pitch-dimensional angle super-resolution as an example, calculating the weight coefficient of digital beam synthesis, and performing digital beam synthesis on the digital zero intermediate frequency echo data row by row through the weight coefficient of digital beam synthesis to obtain M pitch beam synthesis data, the step includes:

[0019] The formula for the weight coefficient of the digital beamforming is:

[0020]

[0021] Where λ represents the wavelength of the radar operating carrier frequency, d x and d yrepresents the center coordinate value of each subarray, β represents the azimuth receiving beam width, Indicates the elevation receiving beam center direction, W′ M×N Represents the weight coefficient of digital beam synthesis;

[0022] Based on the amplitude and phase calibration coefficients of each subarray receiving channel, the actual weight coefficients of digital beam synthesis are obtained;

[0023] The actual weight coefficient formula of the digital beamforming is:

[0024]

[0025] In the formula, Indicates the amplitude and phase calibration coefficients of each radar subarray receiving channel.

[0026] In an embodiment of the present application, taking the pitch-dimensional angle super-resolution as an example, calculating the weight coefficient of digital beam synthesis, and performing digital beam synthesis on the digital zero intermediate frequency echo data row by row through the weight coefficient of digital beam synthesis to obtain M pitch beam synthesis data, the step includes:

[0027] Digital beamforming is based on the pitch dimension d of each row of subarrays. y The digital zero-IF echo data of the same N sub-arrays are DBF synthesized to form M elevation beam synthesis data;

[0028] Taking the i-th pitch beam data synthesis as an example, the pitch beam data synthesis formula is:

[0029]

[0030] In the formula, the subscript (i, k) represents the (i, k)th sub-matrix, W (i，k) represents the actual weight coefficient of the digital beam synthesis of the (i, k)th subarray, I (i，k) (m) represents the real part of the digital zero-IF echo data of the (i, k)th subarray, Q (i，k) (m) represents the imaginary part of the digital zero-IF echo data of the (i, k)th subarray, i∈[1,M], k∈[1,N].

[0031] In an embodiment of the present application, the step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes:

[0032] Performing digital pulse compression processing on the M elevation beam synthesis data respectively to obtain M digital pulse compression results;

[0033] The M digital pulse pressure results are respectively subjected to moving target detection processing to obtain M two-dimensional distance and speed matrices.

[0034] In an embodiment of the present application, the step of extracting two-dimensional target information from the pitch beam synthesis data to obtain two-dimensional one-shot data includes:

[0035] The M distance-speed two-dimensional matrices are respectively subjected to constant false alarm detection processing to obtain the constant false alarm detection result.

[0036] In an embodiment of the present application, the step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes:

[0037] Performing target differentiation on the constant false alarm detection result, selecting the target position with the largest amplitude, extracting the target spectrum peaks of the range unit and Doppler unit corresponding to the target position, performing target angle super-resolution processing, and obtaining two-dimensional one-shot data;

[0038] The expression of the two-dimensional one-shot data is:

[0039] X=[Y1(p,q),Y2(p,q)…Y M (p, q)] H (6);

[0040] The matrix form that can be rewritten is:

[0041] X(f) = A(θ)S(f) + N(f) (7);

[0042] Where f(p, q) represents the target information of the qth range unit on the pth Doppler unit, M represents the number of elevation beams formed by the M-row and N-column subarrays, A(θ) is the array steering vector, S(f) is the signal vector, and N(f) is the noise vector.

[0043] In an embodiment of the present application, the step of performing decorrelation source MUSIC direction finding on the two-dimensional primary snapshot data to obtain the target pitch incidence angle includes:

[0044] The signal source correlation is solved by forward smoothing the two-dimensional one-shot data, and the covariance matrix is ​​estimated by statistical processing of the decorrelated data;

[0045] The covariance matrix calculation formula of the forward smoothing sub-matrix is:

[0046] R k =E[X k X k H ] (8);

[0047] The covariance matrix estimation expression after solving the signal source correlation using the forward smoothing correction method is:

[0048]

[0049] In the formula, E represents the mathematical expectation, X k represents the kth forward spatial smoothing submatrix, R k represents the covariance matrix of the kth forward spatial smoothing submatrix, and P represents the number of submatrices selected for forward spatial smoothing.

[0050] In an embodiment of the present application, the step of performing decorrelation source MUSIC direction finding on the two-dimensional primary snapshot data to obtain the target pitch incidence angle includes:

[0051] The theoretical expression of the covariance matrix of the array output can be written as:

[0052] R=A(θ)R S A(θ) H +δ 2 I(10);

[0053] Where A(θ) represents the array steering vector, R S represents the target echo signal covariance matrix, δ 2 represents the noise power, I represents the unit matrix;

[0054] The covariance matrix output by the array is subjected to eigendecomposition to obtain eigenvalues ​​and eigenvectors; wherein the process of eigendecomposition of the covariance matrix is ​​as follows:

[0055]

[0056] Where U S represents the signal subspace spanned by the eigenvectors corresponding to the large eigenvalues, S represents the number of eigenvalues ​​in the signal subspace, and U N represents the noise subspace spanned by the eigenvectors corresponding to small eigenvalues, and N represents the number of eigenvalues ​​in the noise subspace;

[0057] In an embodiment of the present application, the step of performing decorrelation source MUSIC direction finding on the two-dimensional primary snapshot data to obtain the target pitch incidence angle includes:

[0058] The spectrum estimation calculation formula of the MUSIC algorithm is:

[0059]

[0060] In the formula, θ∈[-θ B +θ0:0.1:θ B +θ0]; θ represents the pitch angle, θ0 is the direction of the pitch beam center, θ Bis the pitch beam scanning range, 0.1 is temporarily set as the beam scanning step; a(θ) represents the signal direction vector, U N represents the noise subspace.

[0061] After calculating the cepstrum of all pitch angles, a spectrum peak search is performed, and the point where the spectrum peak has the maximum position is the target pitch incident angle.

[0062] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0063] In one embodiment of the present application, through the above method, by determining a radar array including M rows and N columns of subarrays, and obtaining the received echo signal of each subarray for digital down-conversion processing, digital zero intermediate frequency echo data is obtained; taking the pitch dimension angle super-resolution as an example, the digital zero intermediate frequency echo data is digitally beamformed row by row through the weight coefficient of digital beam synthesis to obtain the pitch beam synthesis data; the pitch beam synthesis data is subjected to two-dimensional target information extraction to obtain two-dimensional one-shot data; the two-dimensional one-shot data is subjected to decorrelated source MUSIC direction finding to obtain the target pitch incidence angle. Through this method, the target angle estimation resolution and angle estimation accuracy of the array radar in a complex electromagnetic environment can be improved, and the stability and reliability of the target tracking process can be enhanced. At the same time, this method mainly relies on digital signal processing technology, does not require additional hardware resources, and reduces the implementation cost.

[0064] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0066] Figure 1 A flowchart schematically shows a method for super-resolution estimation of target angle based on array radar in an exemplary embodiment of the present application;

[0067] Figure 2 The flowchart of digital beam synthesis processing in the exemplary embodiment of the present application is schematically shown;

[0068] Figure 3 Schematically showing a block diagram of a frequency domain digital pulse compression process of a linear frequency modulation signal in an exemplary embodiment of the present application;

[0069] Figure 4Schematically showing the unit average constant false alarm principle diagram in the exemplary embodiment of the present application;

[0070] Figure 5 Schematically showing a radar super-resolution estimation block diagram in an exemplary embodiment of the present application;

[0071] Figure 6 The schematic diagram shows the principle diagram of forward spatial smoothing in an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0072] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be more comprehensive and complete and fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0073] In addition, the accompanying drawings are only schematic illustrations of the present application and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0074] According to the existing methods, the target angle is within a beam width, and the angle is indistinguishable. In actual applications, the signal sources may be correlated. If the correlation of the target signal source is not de-correlated, the target's incoming wave direction cannot be effectively estimated. If the target moves very fast, the angle will change. In actual applications, multiple snapshot data are not allowed, and only one snapshot data can be obtained, resulting in large target angle estimation errors, and unstable or lost target tracking.

[0075] Based on this, this example embodiment provides a target angle super-resolution estimation method based on array radar. Figure 1 As shown in , the method includes: steps S101 to S104.

[0076] Wherein, step S101: determine a radar array, the radar array includes M rows and N columns of sub-arrays, obtain a received echo signal of each sub-array, and perform digital down-conversion processing to obtain digital zero intermediate frequency echo data.

[0077] Step S102: Taking elevation angle super-resolution as an example, calculate the weight coefficients of digital beam synthesis, and perform digital beam synthesis on the digital zero intermediate frequency echo data row by row using the weight coefficients of digital beam synthesis to obtain M elevation beam synthesis data.

[0078] Step S103: extracting two-dimensional target information from the M elevation beam synthesis data respectively to obtain two-dimensional one-shot data.

[0079] Step S104: performing decorrelation source MUSIC direction finding on the two-dimensional primary snapshot data to obtain the target pitch incidence angle.

[0080] In one embodiment of the present application, through the above method, by determining a radar array including M rows and N columns of subarrays, and obtaining the received echo signal of each subarray for digital down-conversion processing, digital zero intermediate frequency echo data is obtained; digital beam synthesis is performed on the digital zero intermediate frequency echo data through the weight coefficient of digital beam synthesis to obtain elevation beam synthesis data; two-dimensional target information is extracted from the elevation beam synthesis data to obtain two-dimensional one-shot data; and decorrelated source MUSIC direction finding is performed on the two-dimensional one-shot data to obtain the target elevation incident angle. Through this method, the target angle estimation resolution and angle estimation accuracy of the array radar in a complex electromagnetic environment can be improved, and the stability and reliability of the target tracking process can be enhanced. At the same time, this method mainly relies on digital signal processing technology, does not require additional hardware resources, and reduces the implementation cost.

[0081] Next, we will refer to Figures 2 to 6 Each step of the above method in this example implementation is described in more detail.

[0082] In step S101, in a multifunctional phased array radar, signals are picked up by an antenna array arranged in a certain space, which often contains tens of thousands of array elements. Therefore, a subarray structure is usually required. By multi-channel reception and digital processing of the subarray, the digital output of the subarray is obtained, and the digital output signal of the spatial subarray is synthesized into one or more beams through transformation, and the synthesized beam data is used to perform spatial spectrum estimation.

[0083] In the present application, when digital beamforming is performed, the digital beamforming subarray distribution is first established. Specifically, it is assumed that the planar phased array is composed of M rows and N columns of subarrays, and the directional pattern characteristics of each subarray unit are the same. It is assumed that each subarray is placed on a rectangular grid in the XOY plane. x and d y Indicates the center coordinate value of each subarray. The radar transmit signal is a linear frequency modulation signal, and the transmit signal pulse repetition period is T r There are L sampling pulse repetition periods in total, and the number of sampling points in each pulse repetition period is Q. For a specific subarray, the formula for receiving the echo signal at a specific time is:

[0084] The formula for receiving echo signals of each subarray is as follows:

[0085] x(t)=s(t)+n(t)(1);

[0086] Where t represents the receiving time, x(t) represents the received echo signal of each subarray, s(t) represents the received target echo signal of interest, and n(t) represents Gaussian white noise;

[0087] The echo signal of each subarray is received and digitally processed to obtain the zero intermediate frequency echo signal of each subarray after digital down-conversion. The formula is as follows:

[0088] x l (m) = I l (m)+j×Q l (m)(2);

[0089] Where m represents the number of sampling points in each pulse repetition period, m∈[1,Q], l represents a total of 1 sampling pulse repetition period, l∈[1,L], x l (m) represents the zero intermediate frequency echo signal after digital down-conversion, I l (m) represents the real part of the zero-IF echo signal after digital down-conversion, Q l (m) represents the imaginary part of the zero intermediate frequency echo signal after digital down conversion, and j represents the complex imaginary part unit.

[0090] In step S102, taking the elevation angle super-resolution as an example, the weight coefficients of digital beam synthesis are calculated, and digital beam synthesis is performed row by row on the digital zero intermediate frequency echo data using the weight coefficients of digital beam synthesis to obtain M elevation beam synthesis data, including:

[0091] The formula for the weight coefficient of digital beam synthesis is:

[0092]

[0093] Where λ represents the wavelength of the radar operating carrier frequency, d x and d y represents the center coordinate value of each subarray, β represents the azimuth receiving beam width, Indicates the elevation receiving beam center direction, W′ M×N Represents the weight coefficient of digital beam synthesis;

[0094] When calculating the complete digital beamforming weight coefficients, the influence of the amplitude and phase consistency of each subarray channel needs to be considered. Therefore, the actual weight coefficients of digital beamforming are obtained based on the amplitude and phase calibration coefficients of each subarray receiving channel.

[0095] The formula for the actual weight coefficients of digital beamforming is:

[0096]

[0097] In the formula, Indicates the amplitude and phase calibration coefficients of each radar subarray receiving channel.

[0098] It can be understood that after the digital beamforming distribution is established, the weight coefficient of the digital beamforming is calculated. Specifically, taking the elevation angle super-resolution rate as an example, digital beamforming is performed once for N sub-arrays in each row of sub-arrays (with different azimuths) with the same elevation dimension dyv, then M rows and N columns of sub-arrays need to perform M beamformings, and finally M digital beamformings are obtained. The weight coefficients of the M digital beamformings are calculated by the above formula 3.

[0099] Further, taking the pitch-dimensional angle super-resolution as an example, the weight coefficient of digital beam synthesis is calculated, and the digital zero intermediate frequency echo data is digitally beamformed row by row by the weight coefficient of digital beam synthesis to obtain M pitch beam synthesis data, including:

[0100] The digital beamforming process is as follows: Figure 2 As shown, digital beamforming is based on the elevation dimension d of each row of subarrays. y The digital zero intermediate frequency echo data of the same N sub-arrays are synthesized by digital beam forming (DBF) to form M elevation beam synthesis data;

[0101] Taking the i-th elevation beam data synthesis as an example, according to the actual weight coefficient of the digital beam synthesis of the i-th beam, taking the first pulse repetition period as an example, the formula for the elevation beam data synthesis is:

[0102]

[0103] In the formula, the subscript (i, k) represents the (i, k)th sub-matrix, W (i，k) represents the actual weight coefficient of the digital beam synthesis of the (i, k)th subarray, I (i,k) (m) represents the real part of the digital zero-IF echo data of the (i, k)th subarray, Q (i，k) (m) represents the imaginary part of the digital zero-IF echo data of the (i, k)th subarray, i∈[1,M], k∈[1,N].

[0104] In step S103, after the elevation beam synthesis data is obtained, it is necessary to extract two-dimensional target information thereon to obtain two-dimensional one-shot data, so as to facilitate the subsequent decorrelated source MUSIC direction finding.

[0105] Further, the step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes:

[0106] Performing digital pulse compression processing on M elevation beam synthesis data respectively to obtain M digital pulse compression results;

[0107] The M digital pulse pressure results are processed for moving target detection respectively to obtain M two-dimensional distance and velocity matrices.

[0108] The step of extracting two-dimensional target information from the pitch beam synthesis data to obtain two-dimensional one-shot data includes:

[0109] The M two-dimensional range-velocity matrices are respectively subjected to constant false alarm detection processing to obtain constant false alarm detection results.

[0110] It can be understood that the digital pulse compression is realized by using the frequency domain pulse compression method for the M elevation beam synthesis data. The matching coefficient is the product of the frequency response of the matching filter and the weighting function (Hamming window). The received echo signal is processed by Fast Fourier Transformation (FFT) and multiplied by the matching coefficient, and then processed by Inverse Fast Fourier Transformation (Invers e Fast Fourier Transform (IFFT) processing to obtain the pulse pressure result, see Figure 3 As shown, the pulse compression results of M elevation synthetic beam data are obtained.

[0111] When the target detection processing is performed on the M pitch synthetic beam digital pulse compression results respectively, according to the change of the wave position detection number L, in order to obtain lower side lobes in practical applications, the Chebyshev window weighted Fourier transform is used to complete it. Since FFT is easy to implement through programming in the modular digital signal processing platform, FFT transform is used to realize moving target detection, and M range and velocity two-dimensional matrices are obtained, in which each row of the M two-dimensional matrix can obtain the range information of the target, and each column can obtain the Doppler information on the corresponding range gate.

[0112] The M pitch beam synthesis data are subjected to constant false alarm detection processing. The unit average method is used in the distance-speed dimension to realize constant false alarm detection processing. In order to reduce the problem of significant increase in false alarm of this type of detector inside the clutter edge, the improved circuit - the two-side unit average selection circuit (selection of small or average) is generally used. The unit average method processing block diagram is shown in Figure 4 shown.

[0113] Further, the step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes:

[0114] like Figure 5 As shown, the constant false alarm detection result is used to distinguish the target, the target position with the largest amplitude is selected, the target spectrum peak of the range unit and Doppler unit corresponding to the target position is extracted, and the target angle super-resolution processing is performed to obtain two-dimensional one-shot data;

[0115] The expression of the two-dimensional one-shot data is:

[0116] X=[Y1(p,q),Y2(p,q)…Y M (p, q)] H (6).

[0117] This is equivalent to a snapshot in DOA estimation. The array signal DOA is estimated in the frequency domain for this snapshot data. The above formula contains the target signal information and noise. The rewritten matrix form is:

[0118] X(f) = A(θ)S(f) + N(f) (7);

[0119] Where f(p, q) represents the target information of the qth range unit on the pth Doppler unit, M represents the number of beams formed by the M-row and N-column subarrays, A(θ) is the array steering vector, S(f) is the signal vector, and N(f) is the noise vector.

[0120] It can be understood that FFT is a linear operation. After two-dimensional processing of distance and Doppler, the phase difference of the signal part between the M elevation synthetic beams still exists, and Gaussian white noise is still Gaussian white noise after two-dimensional processing.

[0121] In step S104, after the two-dimensional one-shot data is obtained, it is necessary to perform decorrelation source MUSIC direction finding on it to obtain the target pitch incidence angle.

[0122] Furthermore, the step of performing decorrelation source MUSIC direction finding on the two-dimensional one-shot data to obtain the target pitch incidence angle includes:

[0123] The signal source correlation is solved by forward smoothing the two-dimensional one-shot data, and the covariance matrix is ​​estimated by statistical processing of the decorrelated data;

[0124] The covariance matrix calculation formula of the forward smoothing sub-matrix is:

[0125] R k =E[X k X k H ] (8);

[0126] The covariance matrix estimation expression after solving the signal source correlation using the forward smoothing correction method is:

[0127]

[0128] In the formula, E represents the mathematical expectation, X k represents the kth forward spatial smoothing submatrix, R k represents the covariance matrix of the kth forward spatial smoothing submatrix, and P represents the number of submatrices selected for forward spatial smoothing.

[0129] It is understandable that the only basis for spatial spectrum estimation is based on the analysis of the array covariance matrix characteristics. By decomposing the array covariance matrix, the signal subspace and noise subspace can be obtained. The signal zero point, that is, the direction of the signal wave, is determined by using the orthogonal characteristics of the signal direction vector and the noise subspace.

[0130] If some of the D signal sources in space are completely correlated, the coherent signal source entering the array will reduce the rank of the signal subspace of the array covariance matrix, resulting in the signal subspace dimension being smaller than the number of signal sources and the noise subspace estimation error. The correlated signals can be combined into one signal, and the number of independent sources reaching the array will be reduced. In order to perform spatial spectrum estimation under the condition of correlated sources, it is necessary to perform decorrelation preprocessing before spectrum estimation, and then use the method of processing independent sources to perform spatial spectrum estimation.

[0131] Decorrelation processing uses the forward smoothing technology of the reduced-dimensional array space, sacrificing the number of effective beams in exchange for the irrelevance of the signal. The rank of the covariance matrix can be restored, so as to correctly estimate the direction of the signal source. The basic idea is to divide the equidistant linear array into several overlapping sub-arrays. If the array manifolds of each sub-array are the same, the average value of the sub-array covariance matrix can represent the estimated value of the array covariance matrix.

[0132] The two-dimensional one-shot data obtained after two-dimensional processing constitutes an equally spaced spatial sequence, and the covariance matrix can be estimated by performing statistical processing on the spatial sequence.

[0133] like Figure 6 As shown in the figure, the array space forward smoothing technique divides the uniform linear array containing M array elements into P (P>D) interlaced sub-arrays, and the number of array elements in each sub-array is L (L>D), that is, M=L+P-1. Assume that the incoming directions of the D signal sources are set to θ1, θ2, ..., θ d …, θ D , then take the first forward smoothing sub-matrix as the reference sub-matrix, and the data model for the Kth forward smoothing sub-matrix is:

[0134] X k =[x k , x k+1 …x k+L-1]

[0135] =A(θ)D (k-1) S+N k (13);

[0136] Where A(θ) is the steering vector matrix of the subarray;

[0137] A(θ)=[a(θ1), a(θ2),…, a(θ d ),…,a(θ D )] (14);

[0138]

[0139] θ d represents the direction of the wave from the dth signal source, a(θ d ) represents θ d Steering vector of incoming wave direction;

[0140]

[0141] N k =[n k , n k+1 ,…,n k+L-1 ] H (17)

[0142] S=[S1,S2,...,S d , ..., S D ] (18);

[0143] S represents the complex envelope of D signal sources;

[0144] Therefore, the data covariance matrix of the Kth forward smoothing sub-matrix is:

[0145]

[0146] The forward spatial smoothing method recovers the full-rank covariance matrix by finding the mean of the covariance matrices of each sub-matrix, that is, the covariance matrix corrected by forward smoothing is:

[0147]

[0148] in:

[0149] Although the forward smoothing algorithm can estimate the coherent source, the number of subarrays is reduced after smoothing, the antenna degree of freedom is reduced, and the dimension of the covariance matrix is ​​reduced, and the number of estimable signal sources is reduced.

[0150] Furthermore, the step of performing decorrelation source MUSIC direction finding on the two-dimensional one-shot data to obtain the target pitch incidence angle includes:

[0151] The theoretical expression of the covariance matrix of the array output can be written as:

[0152] R=A(θ)R S A(θ) H +δ 2 I(10);

[0153] Where A(θ) represents the array steering vector, R S represents the target echo signal covariance matrix, δ 2 represents the noise power, I represents the unit matrix;

[0154] The covariance matrix output by the array is subjected to eigendecomposition to obtain eigenvalues ​​and eigenvectors; wherein the process of eigendecomposition of the covariance matrix is ​​as follows:

[0155]

[0156] Where U S represents the signal subspace spanned by the eigenvectors corresponding to the large eigenvalues, S represents the number of eigenvalues ​​in the signal subspace, and U N represents the noise subspace spanned by the eigenvectors corresponding to small eigenvalues, and N represents the number of eigenvalues ​​in the noise subspace;

[0157] If the echo signal received by the radar subarray is composed of D independent sources, then rank{R S}=D, and by performing eigendecomposition on R, we can obtain D relatively large eigenvalues ​​and LD eigenvalues ​​equal to δ 2 The eigenvalues ​​of D relatively large eigenvalues ​​correspond to the eigenvectors that form a signal subspace U S , which is the same as the column vector a(θ d )(d=1…D) is the same as the subspace formed by the direction vector a(θ d ) and the orthogonal characteristics of the noise subspace determine the signal zero point, that is, the direction of the signal arrival.

[0158] Furthermore, the step of performing decorrelation source MUSIC direction finding on the two-dimensional one-shot data to obtain the target pitch incidence angle includes:

[0159] The spectrum estimation calculation formula of the MUSIC algorithm is:

[0160]

[0161] In the formula, θ∈[-θ B +θ0:0.1:θB +θ0]; θ represents the pitch angle, θ0 is the direction of the pitch beam center, θ B is the pitch beam scanning range, 0.1 is temporarily set as the beam scanning step; a(θ) represents the signal direction vector, U N represents the noise subspace.

[0162] After calculating the cepstrum of all pitch angles, a spectrum peak search is performed, and the point where the spectrum peak has the maximum position is the target pitch incident angle.

[0163] It can be understood that, in an ideal situation (sufficient number of snapshots and large enough signal-to-noise ratio), the signal subspace and noise subspace in the received data space are completely orthogonal, that is, the steering vector in the signal subspace is orthogonal to the noise subspace, so:

[0164] a H (θ)U N =0(20);

[0165] Due to the existence of noise interference in the actual environment and the finite length of the received data, the estimation of the covariance matrix is ​​inconsistent with the estimation under ideal conditions, and complete orthogonality is difficult to achieve. This is caused by the inconsistency of the array channels and the error in the R matrix estimation. Therefore, the estimation of the incoming wave direction is obtained by minimizing the following cost function:

[0166]

[0167] Therefore, the spectrum estimation calculation formula of the above MUSIC algorithm is 12.

[0168] After calculating the cepstrum of all pitch angles, perform a spectrum peak search to find the first few maximum values ​​(based on the number of targets). The angle corresponding to the maximum value is the target pitch incident angle.

[0169] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application.

Claims

1. A target angle super-resolution estimation method based on array radar, characterized in that: The method includes: Determine the distribution of array radar subarrays, the array radar includes M rows and N columns of subarrays, obtain the received echo signal of each subarray, and perform digital down-conversion processing to obtain digital zero intermediate frequency echo data; Taking the elevation angle super-resolution as an example, the weight coefficient of digital beam synthesis is calculated, and the digital zero intermediate frequency echo data is digitally beam synthesized row by row by using the weight coefficient of digital beam synthesis to obtain M elevation beam synthesis data; Extracting two-dimensional target information from the M elevation beamformation data respectively to obtain two-dimensional snapshot data; The two-dimensional one-shot data is subjected to decorrelated source MUSIC direction finding to obtain the target pitch incidence angle.

2. The target angle super-resolution estimation method based on array radar according to claim 1 is characterized in that: The received echo signal formula of each subarray is as follows: x(t)=s(t)+n(t)(1); Where t represents the receiving time, x(t) represents the echo signal received by each subarray, s(t) represents the received target echo signal of interest, and n(t) represents Gaussian white noise; The formula of the zero intermediate frequency echo signal after digital down-conversion of each subarray is as follows: x l (m)=I l (m)+j×Q l (m)(2); Where m represents the number of sampling points in each pulse repetition period, m∈[1,Q], l represents a total of l sampling pulse repetition periods, l∈[1,L], x l (m) represents the zero intermediate frequency echo signal after digital down-conversion, I l (m) represents the real part of the zero-IF echo signal after digital down-conversion, Q l (m) represents the imaginary part of the zero intermediate frequency echo signal after digital down conversion, and j represents the complex imaginary part unit.

3. The target angle super-resolution estimation method based on array radar according to claim 2 is characterized in that: The step of taking the pitch-dimensional angle super-resolution as an example, calculating the weight coefficient of digital beam synthesis, and performing digital beam synthesis on the digital zero intermediate frequency echo data row by row by using the weight coefficient of digital beam synthesis to obtain M pitch beam synthesis data includes: The formula for the weight coefficient of the digital beamforming is: Where λ represents the wavelength of the radar operating carrier frequency, d x and d y represents the center coordinate value of each subarray, β represents the azimuth receiving beam width, Indicates the elevation receiving beam center direction, W′ M×N Represents the weight coefficient of digital beam synthesis; Based on the amplitude and phase calibration coefficients of each subarray receiving channel, the actual weight coefficients of digital beam synthesis are obtained; The actual weight coefficient formula of the digital beamforming is: In the formula, Indicates the amplitude and phase calibration coefficients of each radar subarray receiving channel.

4. The target angle super-resolution estimation method based on array radar according to claim 3 is characterized in that: The step of taking the pitch-dimensional angle super-resolution as an example, calculating the weight coefficient of digital beam synthesis, and performing digital beam synthesis on the digital zero intermediate frequency echo data row by row by using the weight coefficient of digital beam synthesis to obtain M pitch beam synthesis data includes: Digital beamforming is based on the pitch dimension d of each row of subarrays. y The digital zero-IF echo data of the same N sub-arrays are DBF synthesized to form M elevation beam synthesis data; Taking the i-th pitch beam data synthesis as an example, the pitch beam data synthesis formula is: In the formula, the subscript (i, k) represents the (i, k)th sub-matrix, W (i,k) represents the actual weight coefficient of the digital beam synthesis of the (i, k)th subarray, I (i,k) (m) represents the real part of the digital zero-IF echo data of the (i, k)th subarray, Q (i,k) (m) represents the imaginary part of the digital zero-IF echo data of the (i, k)th subarray, i∈[1,M], k∈[1,N].

5. The target angle super-resolution estimation method based on array radar according to claim 4 is characterized in that: The step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes: Performing digital pulse compression processing on the M elevation beam synthesis data respectively to obtain M digital pulse compression results; The M digital pulse pressure results are respectively subjected to moving target detection processing to obtain M two-dimensional distance and speed matrices.

6. The target angle super-resolution estimation method based on array radar according to claim 5 is characterized in that: The step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes: The M two-dimensional distance-speed matrices are respectively subjected to constant false alarm detection processing to obtain the constant false alarm detection result.

7. The target angle super-resolution estimation method based on array radar according to claim 6 is characterized in that: The step of extracting two-dimensional target information from the M pitch beam synthesis data to obtain two-dimensional one-shot data includes: Performing target differentiation on the constant false alarm detection result, selecting the target position with the largest amplitude, extracting the target spectrum peaks of the range unit and Doppler unit corresponding to the target position, performing target angle super-resolution processing, and obtaining two-dimensional one-shot data; The expression of the two-dimensional one-shot data is: X=[Y1(p,q),Y2(p,q)…Y M (p,q)] H (6); The matrix form that can be rewritten is: X(f) = A(θ)S(f) + N(f) (7); Where Y(p,q) represents the target information of the qth range unit on the pth Doppler unit, M represents the number of elevation beams formed by the M-row and N-column subarrays, A(θ) is the array steering vector, S(f) is the signal vector, and N(f) is the noise vector.

8. The target angle super-resolution estimation method based on array radar according to claim 7 is characterized in that: The step of performing forward smoothing and decoherent source MUSIC direction finding on the two-dimensional one-shot data to obtain the target pitch incidence angle includes: The signal source correlation is solved by forward smoothing the two-dimensional one-shot data, and the covariance matrix is ​​estimated by statistical processing of the decorrelated data; The covariance matrix calculation formula of the forward smoothing sub-matrix is: R k =E[X k X k H ] (8); The covariance matrix estimation expression after solving the signal source correlation using the forward smoothing correction method is: In the formula, E represents the mathematical expectation, X k represents the kth forward spatial smoothing submatrix, R k represents the covariance matrix of the kth forward spatial smoothing submatrix, and P represents the number of submatrices selected for forward spatial smoothing.

9. The target angle super-resolution estimation method based on array radar according to claim 8, characterized in that: The step of performing decorrelation source MUSIC direction finding on the two-dimensional one-shot data to obtain the target pitch incidence angle includes: The theoretical expression of the covariance matrix of the array output can be written as: R=A(θ)R S A(θ) H +d 2 I(10); Where A(θ) represents the array steering vector, R S represents the target echo signal covariance matrix, δ 2 represents the noise power, I represents the unit matrix; The covariance matrix output by the array is subjected to eigendecomposition to obtain eigenvalues ​​and eigenvectors; wherein the process of eigendecomposition of the covariance matrix is ​​as follows: Where U S represents the signal subspace spanned by the eigenvectors corresponding to the large eigenvalues, S represents the number of eigenvalues ​​in the signal subspace, and U N represents the noise subspace spanned by the eigenvectors corresponding to small eigenvalues, and N represents the number of eigenvalues ​​in the noise subspace.

10. The target angle super-resolution estimation method based on array radar according to claim 9, characterized in that: The step of performing decorrelation source MUSIC direction finding on the two-dimensional one-shot data to obtain the target pitch incidence angle includes: The spectrum estimation calculation formula of the MUSIC algorithm is: In the formula, θ∈[-θ B +θ0:0.1:θ B +θ0]; θ represents the pitch angle, θ0 is the direction of the pitch beam center, θ B is the pitch beam scanning range, 0.1 is temporarily set as the beam scanning step; a(θ) represents the signal direction vector, U N represents the noise subspace. After calculating the cepstrum of all pitch angles, a spectrum peak search is performed, and the point where the spectrum peak has the maximum position is the target pitch incident angle.

Images