CN117763832A - Digital array pattern null optimization control method based on channel amplitude and phase error dynamic compensation - Google Patents

Abstract

The invention discloses a digital array pattern null optimization control method based on dynamic compensation of channel amplitude and phase errors. This method determines the amplitude-phase self-calibration dry-noise ratio INR threshold ψ according to the expected null depth. Based on the two-dimensional Fourier Based on the interference detection and angle estimation accelerated by transform FFT, and the orthogonal projection algorithm based on Schmidt orthogonalization, the interference subspace is compensated for the amplitude and phase, thereby dynamically optimizing the depth and width of the pattern anti-interference null. control. The present invention can dynamically calibrate the amplitude and phase errors and achieve optimal control of the interference null depth and width; avoid matrix inversion in the interference subspace projection matrix through Schmidt orthogonalization, and reduce the computational complexity; and achieve the desired null depth through Schmidt orthogonalization. Determine the amplitude and phase self-calibration threshold, use two-dimensional FFT interference detection and Schmidt orthogonalization orthogonal projection algorithm, and use strong interference sources that meet the threshold to calculate the amplitude and phase error self-calibration in parallel, so as to update the amplitude and phase error in real time while working. the goal of.

Description

Digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors

Technical Field

The invention belongs to the technical field of digital array antenna spatial domain anti-interference, and in particular relates to a digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors.

Background Art

Digital array antennas can achieve spatial anti-interference by controlling the radiation pattern to form a null in the interference direction. They also have the advantages of fast beam switching, flexible beam pattern control, and strong spatial interference suppression capabilities. Since each RF channel of a digital array antenna is composed of linear or nonlinear active devices such as mixers, amplifiers, and filters. There are inevitably amplitude and phase errors between RF channels, and they will change slowly with changes in ambient temperature and temperature differences between channels. Channel amplitude and phase errors will affect the accuracy of signal source angle estimation and array gain, and seriously affect the interference null depth and sidelobe level of the array radiation pattern. Therefore, for digital array antennas with low sidelobe and deep null anti-interference requirements, dynamic channel amplitude and phase error calibration must be performed.

The existing calibration algorithms are mainly divided into two categories: active calibration and self-calibration. The active calibration method requires entering a special calibration mode before the system works normally, and uses an auxiliary signal source with a precisely known spatial position for calibration. It is not possible to "calibrate while working". The self-calibration method can simultaneously perform a joint estimate of the channel amplitude and phase error parameters and the direction of arrival (DOA) of the source. Most self-calibration methods obtain more accurate DOA estimates through amplitude and phase calibration. For situations where the interference angle changes rapidly, it is not possible to dynamically achieve the optimization control of the depth and width of the anti-interference pattern null of amplitude and phase compensation.

Summary of the invention

The purpose of the present invention is to solve the problems raised in the above-mentioned background technology, and to provide a digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors. According to the desired pattern nulling depth and width requirements, the channel amplitude and phase errors are dynamically calibrated using a non-cooperative interference source that meets the self-calibration interference-to-noise ratio threshold, thereby effectively controlling the pattern nulling depth and width.

In order to achieve the purpose of the present invention, the present invention discloses a digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors, comprising the following steps:

Step 1: Perform two-dimensional FFT processing on the array received data vector x(k). According to the relationship between the two-dimensional FFT output and the beamforming output and the calculated interference detection threshold T, perform fast interference detection and estimate the number of interferences P and the interference angle. and INR _p ,(p＝1,…,P);

Step 2(a), when P = 1 and INR _P ≥ ψ, calculate the covariance matrix of x(k) by the power method The principal eigenvector of use Interference-oriented vector The difference in the estimated amplitude and phase error

Step 2(b): When P>0, that is, there is effective interference, according to P, The interference subspace C with null-sinking characteristics is constructed by using the null-sinking widths Δu and Δv in the u and v directions, and the amplitude and phase errors calculated in step 2(a) are used to By making correction C, we get

Step 3: Schmidt orthogonalization is performed, and the spatial nulling anti-interference weight w of the nulling depth and nulling width optimization control is calculated using the orthogonal projection algorithm.

The auxiliary amplitude-phase self-calibration threshold calculation module determines the INR threshold ψ of the amplitude-phase error self-calibration by analyzing the relationship between the expected null depth and the channel amplitude-phase error, and between the channel amplitude-phase error and the calibration source INR.

Furthermore, the specific process of step 1 is:

Step 1-1, two-dimensional FFT interference detection and angle estimation;

For a two-dimensional rectangular grid array, the array receives the data vector x(k) and performs two-dimensional FFT processing to obtain

in, is the output signal of the nth array element, N _x , N _y are the number of array elements in the x-axis and y-axis directions respectively;

The corresponding relationship between the two-dimensional FFT output channel (l _x , _ly ) and the angle ( _up ,v _p ) is:

Where _dx and _dy are the array element spacings in the x-axis and y-axis directions respectively, and λ is the wavelength;

Step 1-2, calculate the interference detection threshold T;

First, calculate the noise power _Pn

Where, I is the number of two-dimensional FFT row and column points, I _P is the number of large peaks to be eliminated, and I ₀ is the set of eliminated point positions;

Next, calculate the constant false alarm detection threshold factor α when the false alarm probability is _PF

Finally, the interference detection threshold T is calculated based on the above _Pn and α

T＝ _αPn

Step 1-3: Estimate the number of interferences P and the interference angle and INR _p ,(p＝1,…,P);

When d _x , d _y is greater than λ/2, due to the existence of periodic blur, in the visible area (u ² +v ² ≤1), the number of all possible positions corresponding to an interference after two-dimensional FFT processing is

in, is the floor rounding symbol;

Search for the peak value that exceeds the interference detection threshold T in the data after two-dimensional FFT processing, and estimate the interference angle according to the channel position (l _x , _ly ) corresponding to the peak value The estimation results are grouped; the difference between the interference angle estimation values in the same group is an integer multiple of the fuzzy spacing, and only one angle in the group of interference fuzzy angles needs to be selected. Finally, the number of irrelevant interferences P is determined;

Calculate the interference power according to the peak position (l _x , _ly ) corresponding to the interference

P _p = |F(l _x , _ly ,k)| ²

The corresponding interference-to-noise ratio is INR _p =10log(P _p /P _n ).

Furthermore, step 2 consists of two parallel computing sub-steps, specifically:

Step 2(a): When P = 1 and INR _P ≥ ψ, calculate and update the amplitude and phase errors.

Calculate the covariance matrix of x(k) Iterative estimation by power method The principal eigenvector of The interference directivity vector is: Where x _n , y _n are array element positions, n = 1, 2, ..., N, N is the number of array elements,

and The relationship is

in, Calculate the normalized amplitude and phase error estimates of other array elements divided by the reference array element (the first array element)

in,

Step 2(b): When P>0, calculate the interference subspace for compensating the amplitude and phase errors

According to P. And the null-sinking widths Δu and Δv in the u and v directions are used to construct the interference subspace C = [C ₁ ,C ₂ ,…,C _P ] with null-sinking characteristics, where

The amplitude and phase errors calculated using step 2(a) are By making correction C, we get

Furthermore, step 3 is specifically as follows:

right The column vector of is Schmidt orthogonalized to obtain

The interference orthogonal subspace projection matrix Z ^⊥ is

Orthogonal projection algorithm is used to calculate the spatial nulling anti-interference weights for nulling depth and width optimization control

Among them, _w0 is a static weight without zero sink.

Compared with the prior art, the significant progress of the present invention is that: 1) it can dynamically estimate and compensate for channel amplitude and phase errors according to the expected null depth and width of the directional pattern, and realize the optimization control of the anti-interference null depth and width of the directional pattern; 2) the matrix inversion operation in the interference subspace projection matrix is avoided through Schmidt orthogonalization calculation, thereby reducing the calculation complexity; 3) the amplitude and phase self-calibration threshold is determined by the expected null depth, and an orthogonal projection algorithm based on two-dimensional FFT and an orthogonal projection algorithm based on Schmidt orthogonalization are adopted. The amplitude and phase error self-calibration is calculated in parallel using strong interference sources that meet the self-calibration threshold, so as to achieve the purpose of updating the amplitude and phase errors in real time while working.

In order to more clearly illustrate the functional characteristics and structural parameters of the present invention, further description is given below in conjunction with the accompanying drawings and specific implementation methods.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

Fig. 1 is a flow chart of algorithm implementation of the present invention;

FIG2 is a schematic diagram of the array surface distribution of a 64-element rectangular grid plane array;

3 is a schematic diagram of the relationship between the null depth and the amplitude-phase error, and the amplitude-phase error and the interference source INR through simulation analysis in one embodiment of the present invention;

4 is a schematic diagram of two-dimensional FFT interference detection and angle estimation results when there is one sidelobe interference at different array element spacings in an embodiment of the present invention;

FIG. 5 is a comparison diagram of interference pattern nulling with and without amplitude and phase compensation and output signal to interference noise ratio changes with the number of snapshots in an embodiment of the present invention.

DETAILED DESCRIPTION

The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

In conjunction with FIG1 , a digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors includes the following main steps:

Step 1: Perform two-dimensional FFT processing on the array received data vector x(k). According to the relationship between the two-dimensional FFT output and the beamforming output and the calculated interference detection threshold T, perform constant false alarm detection, peak detection, and angle ambiguity discrimination on the two-dimensional FFT processed data to estimate the number of interferences P and the interference angle and INR _p ,(p＝1,…,P);

Step 2(a), when P = 1 and INR _P ≥ ψ, estimate the covariance matrix of x(k) by the power method The principal eigenvector of use Interference-oriented vector The difference in the estimated amplitude and phase error is a diagonal matrix.

Step 2(b), when P>0, that is, when effective interference exists, according to P, The interference subspace C with null-sinking characteristics is constructed by using the null-sinking widths Δu and Δv in the u and v directions, and the amplitude and phase errors calculated in step 2(a) are used to By making correction C, we get:

Step 3, Schmidt orthogonalization is performed, and the spatial nulling anti-interference weight w of the nulling depth and nulling width optimization control is calculated using the orthogonal projection algorithm.

The specific detailed processing process of step 1 is as follows:

Step 1-1, two-dimensional FFT interference detection and angle estimation;

For a two-dimensional rectangular grid array, the beam output in the ( _up , _vp ) direction is:

in, is the output signal of the nth array element, _Nx , _Ny are the number of array elements in the x-axis and y-axis directions respectively, _dx , _dy are the array element spacing in the x-axis and y-axis directions respectively, and λ is the wavelength.

Performing two-dimensional FFT processing on x(k) yields:

It can be seen that there is a direct correspondence between the beamforming output y( _up , _vp , k) of x(k) and the two-dimensional FFT output F( _lx , _ly , k), namely:

Then we get:

The two-dimensional FFT output is detected and judged, and the interference detection and angle estimation results are obtained through the corresponding relationship between the two-dimensional FFT output channel (l _x , _ly ) and the angle ( _up ,v _p ) given by the above formula.

Step 1-2, calculating the interference detection threshold T;

First, calculate the noise power P _n :

Among them, I is the number of two-dimensional FFT row and column points, _IP is the number of large peaks to be eliminated, and _I0 is the set of eliminated point positions.

Next, calculate the constant false alarm detection threshold factor α when the false alarm probability is _PF :

Finally, the interference detection threshold T is calculated based on the above _Pn and α:

T＝ _αPn (7)

Step 1-3, estimate the number of interferences P and the interference angle and INR _p ,(p＝1,…,P);

When d _x , d _y is greater than λ/2, due to the periodicity of the two-dimensional FFT, we have:

F(l _x ,l _y ,k)＝F(l _x ±r ₁ N _x ,l _y ±r ₂ N _y ,k), (r ₁ =0,1,2,…; r ₂ =0,1 ,2,…) (8)

Then, the correspondence between the two-dimensional FFT output channel (l _x , _ly ) and the angle ( _up ,v _p ) becomes:

There are cases where the same interference corresponds to multiple blur angles, where the row blur angle interval u _dim and the column blur angle interval v _dim are respectively: In the visible area (u ² +v ² ≤1), the number of all possible positions corresponding to an interference is: in, The floor symbol.

Search for the peak value that exceeds the interference detection threshold T in the data after two-dimensional FFT processing, and estimate the interference angle according to the channel position (l _x , _ly ) corresponding to the peak value The estimation results are grouped, and the angle estimation values in the same group differ by an integer multiple of the fuzzy spacing. Suppressing one of the angles will suppress the other fuzzy angles. Therefore, any interference estimation angle in each group is selected to finally determine the number of irrelevant interferences P.

According to the peak position (l _x , _ly ) corresponding to the interference, calculate the interference power:

P _p =|F(l _x ,l _y ,k)| ² (10)

The corresponding interference-to-noise ratio is: INR _p =10log(P _p /P _n ) wherein, when calculating P _n , the number of eliminated points _IP is generally selected to be slightly larger than the product of the maximum number of interferences and the number of blur angles thereof.

Step 2 is divided into two parallel computing sub-steps, and the specific detailed processing process is as follows:

First, assuming that the amplitude and phase errors of the nth channel of the digital array channel are α _n and β _n respectively, the array amplitude and phase error matrix is expressed as: N is the number of array elements.

Step 2(a): When P = 1 and INR _P ≥ ψ, calculate and update the amplitude and phase errors.

Iteratively estimate the covariance matrix of x(k) by the power method The largest eigenvector of The iterative formula for the lth iteration of the power method is:

in, means The largest element in the model.

First, randomly initialize the iteration vector Set the iteration count l = 1; then calculate make If V ^(l) is less than the selected threshold Terminate the iteration, otherwise set l = l + 1 and continue iterating; finally, after L iterations, the main eigenvalue is obtained The corresponding principal eigenvector is

The interference directivity vector is: Where x _n , y _n are array element positions, n = 1, 2, ..., N,

and The relationship is: Right now:

in, The normalized amplitude and phase error estimates of other array elements divided by the reference array element (the first array element) are calculated as follows:

Step 2(b): when P>0, calculate the interference subspace that compensates for the amplitude and phase errors

According to P. And the null-stretch widths Δu and Δv in the u and v directions are used to construct an interference subspace C = [C ₁ ,C ₂ ,…,C _P ] with null-stretch characteristics, where:

The amplitude and phase errors calculated using step 2(a) are By making correction C, we get:

Furthermore, step 3 is specifically as follows:

right The column vector ε _i of the Schmidt orthogonalization is obtained in, The calculation method of each column vector γ _i (i＝1,2,…,5P) is as follows:

at this time, The interference orthogonal subspace projection matrix Z ^⊥ is:

The orthogonal projection algorithm is used to calculate the spatial nulling anti-interference weights of the nulling depth and width optimization control:

Wherein, w ₀ is a static weight without nulling, such as the steering vector a(u ₀ ,v ₀ ) of the beam pointing direction (u ₀ ,v ₀ ) or an amplitude-weighted low sidelobe weight.

According to the above description, and in combination with the algorithm implementation flow chart of the present invention in FIG1 , the implementation method of the present invention mainly includes the following two parts:

1. Interference detection and estimation:

Perform two-dimensional FFT processing on the array received data vector x(k), perform fast interference detection based on the relationship between the two-dimensional FFT output and the beamforming output and the calculated interference detection threshold, and estimate the number of interferences P and the interference angle and INR _p ,(p＝1,…,P);

2. Anti-interference weight calculation:

When P = 1 and INR _P ≥ ψ, the covariance matrix of x(k) is calculated by the power method The principal eigenvector of use Interference-oriented vector The difference in the estimated amplitude and phase error

When P>0, that is, there is effective interference, according to P, The interference subspace C with null-sinking characteristics is constructed by using the null-sinking widths Δu and Δv in the u and v directions, and the amplitude and phase errors calculated in step 2(a) are used to By making correction C, we get

right Schmidt orthogonalization is performed, and the spatial nulling anti-interference weight w of the nulling depth and nulling width optimization control is calculated using the orthogonal projection algorithm.

Among them, the auxiliary step is the amplitude and phase self-calibration threshold calculation module: by analyzing the relationship between the expected null depth and the channel amplitude and phase error, and the channel amplitude and phase error and the calibration source INR, the INR threshold ψ of the amplitude and phase error self-calibration is determined.

The present invention is described in detail below with reference to specific embodiments.

Example

The present invention proposes a method for dynamically calibrating channel amplitude and phase errors according to the desired null depth and width of the directional pattern by using a strong interference source that meets the self-calibration threshold, thereby effectively controlling the null depth and width of the directional pattern. See Figure 1 for a flowchart of the method.

This example uses a matrix grid array with N=64 elements. The unit antennas are isotropic omnidirectional antennas. The mutual coupling between elements is not considered. Figure 2 shows the rectangular grid array model used. Beam pointing Interference from The INR is 10dB, the noise is an additive white Gaussian noise vector, and the noise is unrelated to the signal. There are amplitude and phase errors between channels, and the amplitude error α _n and the phase error β _n both obey uniform distribution, α _n =1+r _n ,β _n =s _n , where That is, the RMS values of α _n and β _n are A -35dB Chebyshev amplitude weighting is used for low sidelobe processing. The expected null depth in this example is -45dB, and the null width requirements Δu and Δv are both 0.0348, that is, the null width is about ±2°. The two-dimensional FFT angle ambiguity problem is analyzed and verified at different array element spacings, including: (1) the array element spacing is equal to the wavelength, that is, the array element spacing in the x-axis and y-axis directions is _dx = λ, _dy = λ respectively; (2) the array element spacing is equal to half the wavelength, that is, the array element spacing in the x-axis and y-axis directions is _dx = 0.5λ, _dy = 0.5λ respectively.

For a 64-element matrix grid planar array with two array element spacings, the implementation of a digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors includes the following processing steps:

First, through the auxiliary steps, the relationship between the channel amplitude and phase error and INR, as well as the relationship between the null depth and the amplitude and phase error are simulated, and the amplitude and phase self-calibration INR threshold that meets the expected null depth of -45dB is determined to be: ψ = 5dB. The simulation results of the determined threshold are shown in Figure 3.

Figure (a) in Figure 3 simulates the relationship between the residual amplitude and phase error value and INR when the root mean square of the four channel amplitude and phase errors (σ _α ,σ _β ) are (0.25dB, 2.5°), (0.5dB, 5°), (0.75dB, 7.5°), and (1.0dB, 10°). Figure (b) in Figure 3 shows the relationship between the null depth and the residual amplitude and phase error. Figure (b) marks the range of the amplitude and phase calibration residual when the null depth is better than -45dB. For the array used in the simulation, the root mean square of the residual amplitude error cannot exceed 0.18dB, and the root mean square of the residual phase error cannot exceed 8°. At the same time, combined with the simulation results of Figure (a), it is obtained that when the channel residual amplitude and phase error meets the above requirements, the self-calibration INR threshold ψ＝5dB.

Step 1: Set the false alarm probability of interference detection to _PF = ^10-4 , the number of two-dimensional FFT row and column points I to 128, and use equation (6) to calculate the threshold factor α = 9.21. According to the two-dimensional FFT interference detection and angle estimation algorithm, calculate the number of interference angle estimation ambiguity angles under two array element spacing conditions.

(1) In the case of large array element spacing, calculate the number of blur angles in the visible area (u ² +v ² ≤1) Therefore, one interference will be detected at two angles in the visible area, with intervals of u _dim = λ/d _x = 1, v _dim = λ/d _y = 1. After angle ambiguity discrimination, it can be determined that there is only one interference, P = 1.

(2) When the array element spacing is small, the number of blurred angles in the visible area is calculated to be 0, so there is no angle blur problem, and P = 1.

Figure 4 (a) and (b) simulate the two-dimensional FFT interference detection and angle estimation results under large array element spacing and normal array element spacing, respectively. The area surrounded by the red dotted line is the visible area. The simulation results show that in the first case, when the array element spacing is greater than λ/2, angle blur occurs in the visible area. Select any angle in the group of blurred angles, such as Right now The interference power is calculated using equation (10) and the interference-to-noise ratio INR _P ≈10 dB. In the second case, the array element spacing is equal to λ/2, and there is no angle ambiguity in the visible area. The interference estimation result is Right now INR _P ≈10dB.

Step 2(a): P = 1, INR _P ≥ 5dB, calculate the amplitude and phase errors

The number of snapshots is set to 256, and the covariance matrix of x(k) is calculated. Select the threshold for terminating the iteration of the power method Estimated The principal eigenvector of and interference-oriented vector The relationship is:

The simulation sets the initial amplitude and phase error σα＝1dB, σβ＝10°, INR＝10dB, and estimates the amplitude and phase error matrix according to formula (13): Table 1 shows the comparison between the actual amplitude and phase errors of the set channels and the amplitude and phase errors estimated by the algorithm.

Table 1 Comparison between the actual channel amplitude and phase error and the estimated amplitude and phase error

From the data analysis in Table 1, it can be seen that the amplitude and phase errors estimated by the amplitude and phase calibration algorithm are basically close to the actual amplitude and phase errors.

Step 2(b), when P>0, calculate the interference subspace that compensates for the amplitude and phase errors

The array element spacing is set to λ/2, and the interference subspace C with null-stretching characteristics is constructed using equation (14). The two-dimensional FFT angle estimation result is obtained from step 1. P = 1. The null expansion widths Δu and Δv are both 0.0348, so C ₁ in equation (14) is expressed as:

C ₁ =[a(0.5625,0),a(0.5625,0-0.0348),a(0.5625,0+0.0348),a(0.5625+0.0348,0),a(0.5625-0.0348,0)]

use After modifying C ₁ , we get:

Step 3: Use formula (15) to After Schmidt orthogonalization, we get The interference orthogonal subspace projection matrix is: The orthogonal projection algorithm is used to calculate the spatial nulling anti-interference weights of the nulling depth and width optimization control:

Wherein, _w0 is the low sidelobe weight of -35dB Chebyshev amplitude weighting. Table 2 gives the specific value of w.

Table 2 Spatial null anti-interference weight w coefficient

Figure 5 (a) and (b) simulate the interference null pattern comparison with and without amplitude and phase compensation and the output signal to noise ratio change with the number of snapshots. It can be seen that the null depth of the directional pattern after compensating for the amplitude and phase errors is better than -45dB, a widened null of about ±2° is formed near the interference, and the output signal to noise ratio is greater than the case without amplitude and phase compensation.

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

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims (4)

1. A digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors, characterized in that it comprises the following steps: Step 1: Perform two-dimensional FFT processing on the array received data vector x(k). According to the relationship between the two-dimensional FFT output and the beamforming output and the calculated interference detection threshold T, perform fast interference detection and estimate the number of interferences P and the interference angle. and INR _p ,(p＝1,…,P); Step 2(a), when P = 1 and INR _P ≥ ψ, calculate the covariance matrix of x(k) by the power method The principal eigenvector of use Interference-oriented vector The difference in the estimated amplitude and phase error Step 2(b): When P>0, that is, there is effective interference, according to P, The interference subspace C with null-sinking characteristics is constructed by using the null-sinking widths Δu and Δv in the u and v directions, and the amplitude and phase errors calculated in step 2(a) are used to By making correction C, we get Step 3: Schmidt orthogonalization is performed, and the spatial nulling anti-interference weight w of the nulling depth and nulling width optimization control is calculated using the orthogonal projection algorithm. 2. According to the method for optimizing the control of nulling of a digital array pattern based on dynamic compensation of channel amplitude and phase errors in claim 1, the specific process of step 1 is as follows: Step 1-1, two-dimensional FFT interference detection and angle estimation; For a two-dimensional rectangular grid array, the array receives the data vector x(k) and performs two-dimensional FFT processing to obtain in, is the output signal of the nth array element, N _x , N _y are the number of array elements in the x-axis and y-axis directions respectively; The corresponding relationship between the two-dimensional FFT output channel (l _x , _ly ) and the angle ( _up ,v _p ) is: Where _dx and _dy are the array element spacings in the x-axis and y-axis directions respectively, and λ is the wavelength; Step 1-2, calculate the interference detection threshold T; First, calculate the noise power _Pn Where, I is the number of two-dimensional FFT row and column points, I _P is the number of large peaks to be eliminated, and I ₀ is the set of eliminated point positions; Next, calculate the constant false alarm detection threshold factor α when the false alarm probability is _PF Finally, the interference detection threshold T is calculated based on the above _Pn and α T＝ _αPn Step 1-3: Estimate the number of interferences P and the interference angle and INR _p ,(p＝1,…,P); When d _x , d _y is greater than λ/2, due to the existence of periodic blur, in the visible area (u ² +v ² ≤1), the number of all possible positions corresponding to an interference after two-dimensional FFT processing is in, is the floor rounding symbol; Search for the peak value that exceeds the interference detection threshold T in the data after two-dimensional FFT processing, and estimate the interference angle according to the channel position (l _x , _ly ) corresponding to the peak value The estimation results are grouped; the interference angle estimation values in the same group differ from each other by an integer multiple of the fuzzy spacing, and only one angle in the group of interference fuzzy angles needs to be selected; and finally the number of irrelevant interferences P is determined; Calculate the interference power according to the peak position (l _x , _ly ) corresponding to the interference P _p = |F(l _x , _ly ,k)| ² The corresponding interference-to-noise ratio is INR _p =10log(P _p /P _n ). 3. The digital array pattern nulling optimization control method based on dynamic compensation of channel amplitude and phase errors according to claim 1 is characterized in that step 2 is two parallel calculation sub-steps, specifically: Step 2(a): When P = 1 and INR _P ≥ ψ, calculate and update the amplitude and phase errors. Calculate the covariance matrix of x(k) Iterative estimation by power method The principal eigenvector of The interference guidance vector is Where x _n , y _n are array element positions, n = 1, 2, ..., N, N is the number of array elements, and The relationship is in, Calculate the normalized amplitude and phase error estimates of other array elements divided by the reference array element (the first array element) in, Step 2(b): When P>0, calculate the interference subspace for compensating the amplitude and phase errors According to P. And the null-sinking widths Δu and Δv in the u and v directions are used to construct the interference subspace C = [C ₁ ,C ₂ ,…,C _P ] with null-sinking characteristics, where The amplitude and phase errors calculated using step 2(a) are By making correction C, we get 4. The digital array pattern nulling optimization control method based on channel amplitude and phase error dynamic compensation according to claim 1, characterized in that step 3 specifically comprises: right The column vector of is Schmidt orthogonalized to obtain The interference orthogonal subspace projection matrix Z ^⊥ is Orthogonal projection algorithm is used to calculate the spatial nulling anti-interference weights for nulling depth and width optimization control Among them, _w0 is a static weight without zero sink.