Array antenna fault correction method and device based on alternate projection, and storage medium

Abstract

The invention relates to an array antenna fault correction method and device based on alternate projection and a storage medium, and the method comprises the following steps: determining the positions and number of fault antenna units in an antenna array, and setting the fault antenna units to be in a closed state; obtaining an original radiation pattern before the antenna array fails and a radiation pattern after the antenna array is damaged, setting correction target parameters, and establishing a target pattern mask according to the target parameters; and based on the target directional diagram mask, performing excitation compensation iterative optimization on residual non-fault array elements through an improved alternating projection method combined with a Bayesian compressed sensing technology, and realizing fault correction. Compared with the prior art, the method has the advantages that the high-quality repair of the damaged array antenna pattern can be realized by using the fewest reconstructed array elements, and the efficiency is higher.

Description

Technical Field

[0001] The present invention belongs to the field of electronic technology and relates to an array antenna fault correction method, in particular to an array antenna fault correction method, device and storage medium based on alternating projection, which is suitable for array antenna system fault correction in the fields of modern wireless communication systems, radar and remote sensing. Background Art

[0002] With the rapid development of modern communication and radar systems, the performance stability of antenna arrays has received increasing attention. However, environmental factors and physical damage often cause individual elements in the antenna array to fail. These faulty elements can lead to problems such as reduced radiation gain, increased sidelobe levels, and directional pattern distortion, which seriously affect the radiation pattern and efficacy of the entire array.

[0003] Traditional fault repair methods usually require physical replacement of damaged antenna elements, which is costly and inefficient. Therefore, it has become an important research direction to correct the radiation pattern of the damaged array by reconstructing the remaining lossless array element excitations and to achieve efficient and low-cost repair of the damaged array radiation performance.

[0004] In the prior art, F. Zardi et al. proposed an array repair method based on the minimum complexity fault correction (MCFC) framework in 2020, which achieved minimum complexity repair of the array pattern through compression processing. However, this method has low efficiency for fault correction of large-scale arrays and is not suitable for real-time repair. Chinese patent CN113420431B discloses an array pattern repair method based on iterative weighted L1 norm, but this method is sensitive to parameter settings and cannot guarantee the optimal repair effect.

[0005] Therefore, the prior art still has the following problems:

[0006] 1) The repair efficiency is low and is not suitable for real-time repair of large-scale arrays;

[0007] 2) The parameters are highly sensitive and it is difficult to ensure the optimal repair effect;

[0008] 3) It is easy to fall into the local optimal solution, which affects the repair quality;

[0009] 4) The repair cost and complexity are high, and the practicality is limited. Summary of the invention

[0010] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide an array antenna fault correction method based on alternating projection, which can use the least reconstructed array elements to achieve high-quality repair of the damaged array antenna radiation pattern and is more efficient, and is suitable for real-time repair of large-scale antenna arrays.

[0011] The purpose of the present invention can be achieved by the following technical solutions:

[0012] An array antenna fault correction method based on alternating projection comprises the following steps:

[0013] Determine the location and number of faulty antenna units in the antenna array, and set the faulty antenna units to a shutdown state;

[0014] Obtaining an original radiation pattern of the antenna array before a fault occurs and a radiation pattern after the antenna array is damaged, setting correction target parameters, and establishing a target pattern mask according to the target parameters;

[0015] Based on the target pattern mask, an improved alternating projection method combining Bayesian compressed sensing technology is used to iteratively optimize the excitation compensation of the remaining non-faulty array elements to achieve fault correction.

[0016] Furthermore, the correction target parameters include a target sidelobe level and / or a mainlobe beamwidth.

[0017] Furthermore, when performing iterative optimization of excitation compensation on the remaining non-faulty array elements, the objective function expression used is:

[0018] minimize‖Δw‖0s.tR(w+Δw)≤M

[0019] Among them, ‖Δw‖0 represents the L0 norm of Δw, that is, the number of non-zero elements in the excitation compensation vector Δw; w is the original excitation vector; M is the discrete representation of the target pattern mask; the matrix R is Discretization representation of x n represents the position of the nth unit, represents the imaginary unit, k is the wave number of free space, u=sin(θ), θ represents the pitch angle in space.

[0020] Furthermore, the improved alternating projection method comprises the following steps:

[0021] Step S301: Initialize the number of iterations i = 0 and set the initial directional pattern F [0] is the original radiation pattern before the fault occurs;

[0022] Step S302: The initial directional pattern F of the i-th iteration [i] Project it onto the target pattern mask to get the forward projection

[0023] Step S303: Based on the forward projection The Bayesian compressed sensing method is used to solve the compensation excitation vector of the i-th iteration, and then the correction direction map F obtained by the i-th iteration is calculated. [i+1], determine the correction direction diagram F [i+1] Whether the target pattern condition is met, if so, proceed to step S304, if not, set the number of iterations i=i+1, and return to step S302;

[0024] Step S304: Set the non-zero element with the smallest absolute value in the current compensation excitation vector to 0, recalculate the correction direction map, and determine whether the iteration termination condition is met. If so, proceed to step S305. If not, set the number of iterations i=i+1 and return to step S302.

[0025] Step S305: Select the compensation excitation vector with the least number of non-zero elements and the best corresponding correction pattern radiation performance from all compensation excitation vectors as the global optimal solution, and perform excitation compensation on the remaining non-faulty array elements.

[0026] Furthermore, in step S302, the forward projection The calculation formula is:

[0027]

[0028] in, Represents F [i] The forward projection of ; |.| represents taking the absolute value of the complex number; α is the scaling factor to adaptively adjust the absolute size of the directional pattern mask; δ dB It is a penalty factor with dB as the dimension; ∠ represents the angle calculation of complex numbers.

[0029] Furthermore, in step S303, the compensation excitation vector is calculated by solving the following matrix equation:

[0030]

[0031] Among them, Δw [i] represents the compensation excitation vector of the i-th iteration.

[0032] Furthermore, in step S303, when the Bayesian compressed sensing method is used to solve the compensation excitation vector, the matrix equation is first transformed into:

[0033]

[0034] Among them, Re and Im represent the real part and imaginary part operations respectively;

[0035] Then, the Bayesian compressed sensing method is used to solve Re(Δw [i] ) and Im(Δw [i] );

[0036] Finally, the compensation excitation vector Δw of the i-th iteration is obtained [i] =Re(Δw[i] )+jIm(Δw [i] ).

[0037] Further, in step S304, the iteration termination condition includes any one of the following:

[0038] The number of iterations reaches the preset maximum value;

[0039] A compensation excitation vector with a smaller number of non-zero elements cannot be obtained in several consecutive iterations.

[0040] The present invention also provides an electronic device, comprising:

[0041] one or more processors;

[0042] Memory; and

[0043] One or more programs stored in a memory, the one or more programs comprising instructions for executing the array antenna fault correction method as described above.

[0044] The present invention also provides a computer-readable storage medium, comprising one or more programs for execution by one or more processors of an electronic device, wherein the one or more programs include instructions for executing the array antenna fault correction method as described above.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The present invention can effectively avoid falling into local optimum by improving the alternating projection method, combining Bayesian compressed sensing technology with innovative iterative termination conditions, ensure the best directional pattern correction effect by using the least reconstruction units, and reduce the correction cost and complexity.

[0047] 2. The present invention uses an efficient Bayesian compressed sensing algorithm in the alternating projection iterative process and introduces an innovative adaptive pattern mask and penalty factor strategy, which can accelerate the convergence of the iterative projection process, reduce the number of iterations, and greatly improve the efficiency of the algorithm. It has important application value for fault correction of large-scale antenna arrays. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the process of the present invention;

[0049] Figure 2 A schematic diagram of the process flow of the improved alternating projection method proposed by the present invention;

[0050] Figure 3 It is a schematic diagram of the directional pattern correction effect of a 32-unit Chebyshev linear array in an embodiment of the present invention, including a comparison of the original directional pattern, the damaged directional pattern and the directional pattern after correction;

[0051] Figure 4 It is a schematic diagram of the correction result of the array element excitation amplitude of the 32-unit Chebyshev linear array in an embodiment of the present invention, showing the comparison between the original excitation, the damaged excitation and the corrected excitation. DETAILED DESCRIPTION

[0052] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0053] refer to Figure 1 As shown, this embodiment provides an array antenna fault correction method based on alternating projection, and the specific implementation steps of the method include:

[0054] Step S1: Use existing diagnostic technology to determine the location and number of faulty antenna units in the array.

[0055] Let the total number of array elements be N, the number of faulty units be S, and their serial numbers form set S, and the number of remaining intact units be Q, and their serial numbers form set Q. Therefore, S+Q=N and S+Q={1,2,…,N}. Set all faulty units to the off state, that is, set their excitation to 0.

[0056] Step S2, obtaining the original radiation pattern of the antenna array before the fault occurs and the pattern after the damage, setting correction target parameters, including but not limited to the sidelobe level and the mainlobe beamwidth, and defining the target pattern mask according to the target parameters.

[0057] The original pattern of the array is expressed as:

[0058]

[0059] Where u = sin(θ), θ represents the spatial pitch angle, n is the antenna unit number, and w n and x n represent the original excitation and position of the nth unit, respectively, represents the imaginary unit and k is the wave number of free space.

[0060] Since all faulty unit excitations are set to 0 in step S1, the directional pattern of the damaged antenna array is expressed as:

[0061]

[0062] Where F(u) is the directivity pattern of the damaged array.

[0063] According to the original radiation pattern of the array, the correction target parameters are set, including but not limited to the target sidelobe level and the main lobe beam width. According to the target parameters, the normalized upper mask function M(u) of the target radiation pattern can be determined, wherein the maximum value of M(u) is 1. Since the user can set a variety of different correction target parameters, there is no unified strict mathematical expression for M(u). In this embodiment, only the single beam basic situation of the sidelobe level SLL and the main lobe beam width is set, and M(u) can be expressed as:

[0064]

[0065] Wherein, SLL is the sidelobe level value in dB.

[0066] Step S3: By using the improved alternating projection method combined with the Bayesian compressed sensing technology, the excitation compensation of the remaining non-faulty array elements is iteratively optimized until the termination condition is met, so that the correction pattern meets the mask restriction of the target pattern and the best repair effect is achieved using the least reconstruction units.

[0067] In order to correct the damaged radiation pattern and achieve the repair effect, it is necessary to excite and compensate the remaining intact antenna units. Let Δw = {Δw n ; n∈Q} is the excitation compensation vector, where Δw n represents the excitation for compensating the nth unit. Therefore, the corrected directivity pattern F C (u) can be expressed as

[0068]

[0069] Discretized sampling is performed on the observation domain u, and this formula can be converted into a matrix equation F = R (w + Δw). Among them, the vector F is F C The discretization representation of (u), the matrix R is The discretization representation of the original excitation vector w = {w n ; n∈Q}.

[0070] Therefore, the directional pattern correction problem to be solved can be described in the following mathematical form:

[0071] minimize||Δw||0 stR(w+Δw)≤M

[0072] Among them, ‖Δw‖0 represents the L0 norm of Δw, that is, the number of non-zero elements in Δw; M is the discrete representation of the upper mask function M(u).

[0073] Furthermore, the improved alternating projection method is used to iteratively solve the problem. The improved alternating projection method iteratively optimizes the corrected directional map by repeatedly performing the forward projection and backward projection processes. The algorithm flow is as follows: Figure 2 As shown, the following steps are included:

[0074] Step S301: Initialize the number of iterations to i←0, the initial directional pattern F [0] is the original pattern before the array is damaged. Define the penalty factor δ dB and the normalized target direction map mask M.

[0075] Step S302: forward projection operation, F [i] Projected onto the target pattern mask M, where F [i] Represents the initial directional pattern of the i-th iteration. Specifically:

[0076]

[0077] in, Represents F [i] forward projection; |.| represents the absolute value of the complex number; α is the scaling factor to adaptively adjust the absolute size of the directional pattern mask, set to α = max(|F [i] |);δ dB It is a penalty factor with dB as the dimension, usually set between 3dB and 10dB, which can accelerate the convergence of the algorithm and reduce the number of iterations; ∠ represents the complex number argument operation.

[0078] Step S303: Back projection operation. Specifically:

[0079] Direction map obtained by forward projection The compensation excitation vector Δw is calculated by solving the following matrix equation,

[0080]

[0081] Among them, Δw [i] represents the compensation excitation vector of the i-th iteration.

[0082] The following transformation of the above equation is solved using the Bayesian compressed sensing method:

[0083]

[0084] Where Re and Im represent the real and imaginary part operations respectively. Solving for Re(Δw [i] ) and Im(Δw [i] ), the compensation excitation vector can be expressed as Δw [i] =Re(Δw [i] )+jIm(Δw [i] ).

[0085] Next, calculate the corrected directional map obtained in the i-th iteration:

[0086] F [i+1]=R(w+Δw [i] )

[0087] If F [i+1] Satisfy the target pattern condition, that is, |F [i+1] |≤αM strictly satisfies, then Δw [i] Record it as a feasible solution and go to step S304; otherwise, return to step S302, the number of iterations i←i+1, and start a new round of iterations.

[0088] Step S304: Due to F [i+1] The target pattern conditions are met and the alternate projections are used to achieve convergence. In order to avoid falling into the local optimum, the present invention proposes an innovative iterative termination condition. [i] The non-zero element with the smallest absolute value in is set to 0, and the modified compensation excitation vector Δw′ is obtained. [i] According to Δw′ [i] Recalculate the correction pattern for the i-th iteration:

[0089] F [i+1] =R(w+Δw′ [i] )

[0090] If any of the following iteration termination conditions is met, the algorithm stops iterating and enters step S305; otherwise, it returns to step S302, the number of iterations i←i+1, and a new round of iterations begins. The iteration termination conditions are: 1) the number of iterations reaches the maximum value; 2) the algorithm is stuck and a sparser (i.e., a smaller number of non-zero elements) compensation excitation vector Δw cannot be obtained in several consecutive iterations.

[0091] In this step, after recalculating the correction direction map, entering a new iteration can avoid falling into the local optimum and help find the true global optimal solution.

[0092] Step S305: Select a solution with the lowest complexity (i.e., the least number of non-zero elements) and the best corresponding correction pattern radiation performance from all feasible solutions of Δw as the global optimal solution. Its corresponding pattern is output as the final correction pattern. The solution with the best radiation performance is determined based on the performance indicators that the user specifically cares about, such as the pattern with the narrowest beam width, the largest gain, etc.

[0093] In order to verify the accuracy of the method of the present invention, fault repair is performed on an ideal 32-unit Chebyshev linear array, where the number of faulty units is 2, the sequence numbers are {2, 5}, and the sidelobe level is -35 dB. Figure 3The correction pattern obtained by the method of the present invention is compared with the original array pattern and the damaged array pattern. Among them, the sidelobe level of the damaged array pattern deteriorates and rises to -27.11dB. After the damaged array pattern is repaired by the method of the present invention, the sidelobe level of the correction array is reduced to -35.01dB, which is even slightly better than the sidelobe level of the original array pattern, proving the high-quality repair effect of the method of the present invention. Figure 4 The results of the corrected array element excitation amplitude and the original array element excitation and the damaged array element excitation are compared. It can be seen that the method of the present invention achieves the best repair effect by using only 16 reconstruction units (i.e., compensating for the excitation of 16 intact units), reducing the correction cost and complexity.

[0094] In another embodiment,

[0095] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for correcting array antenna faults based on alternating projection, characterized in that: The following steps are involved: Determine the location and number of faulty antenna units in the antenna array, and set the faulty antenna units to a shutdown state; Obtaining an original radiation pattern of the antenna array before a fault occurs and a radiation pattern after the antenna array is damaged, setting correction target parameters, and establishing a target pattern mask according to the target parameters; Based on the target pattern mask, an improved alternating projection method combining Bayesian compressed sensing technology is used to iteratively optimize the excitation compensation of the remaining non-faulty array elements to achieve fault correction.

2. The array antenna fault correction method based on alternating projection according to claim 1, characterized in that: The correction target parameters include a target sidelobe level and / or a mainlobe beamwidth.

3. The array antenna fault correction method based on alternating projection according to claim 1, characterized in that: When performing iterative optimization of excitation compensation on the remaining non-faulty array elements, the objective function expression used is: minimize‖Δw‖0s.tR(w+Δw)≤M Among them, ‖Δw‖0 represents the L0 norm of Δw, that is, the number of non-zero elements in the excitation compensation vector Δw; w is the original excitation vector; M is the discrete representation of the target pattern mask; the matrix R is Discretization representation of x n represents the position of the nth unit, represents the imaginary unit, k is the wave number of free space, u=sin(θ), θ represents the pitch angle in space.

4. The array antenna fault correction method based on alternating projection according to claim 3, characterized in that: The improved alternating projection method comprises the following steps: Step S301: Initialize the number of iterations i = 0 and set the initial directional pattern F [0] is the original radiation pattern before the fault occurs; Step S302: The initial directional pattern F of the i-th iteration [i] Project it onto the target pattern mask to get the forward projection Step S303: Based on the forward projection The Bayesian compressed sensing method is used to solve the compensation excitation vector of the i-th iteration, and then the correction direction map F obtained by the i-th iteration is calculated. [i+1] , determine the correction direction diagram F [i+1] Whether the target pattern condition is met, if so, proceed to step S304, if not, set the number of iterations i=i+1, and return to step S302; Step S304: Set the non-zero element with the smallest absolute value in the current compensation excitation vector to 0, recalculate the correction direction map, and determine whether the iteration termination condition is met. If so, proceed to step S305. If not, set the number of iterations i=i+1 and return to step S302. Step S305: Select the compensation excitation vector with the least number of non-zero elements and the best corresponding correction pattern radiation performance from all compensation excitation vectors as the global optimal solution, and perform excitation compensation on the remaining non-faulty array elements.

5. The array antenna fault correction method based on alternating projection according to claim 4, characterized in that: In step S302, the forward projection The calculation formula is: in, Represents F [i] The forward projection of ; |.| represents taking the absolute value of the complex number; α is the scaling factor to adaptively adjust the absolute size of the directional pattern mask; δ dB It is a penalty factor with dB as the dimension; ∠ represents the angle calculation of complex numbers.

6. The array antenna fault correction method based on alternating projection according to claim 4, characterized in that: In step S303, the compensation excitation vector is calculated by solving the following matrix equation: Among them, Δw [i] represents the compensation excitation vector of the i-th iteration.

7. The array antenna fault correction method based on alternating projection according to claim 6, characterized in that: In step S303, when the Bayesian compressed sensing method is used to solve the compensation excitation vector, the matrix equation is first transformed into: Among them, Re and Im represent the real part and imaginary part operations respectively; Then, the Bayesian compressed sensing method is used to solve Re(Δw [i] ) and Im(Δw [i] ); Finally, the compensation excitation vector Δw of the i-th iteration is obtained [i] =Re(Δw [i] )+jIm(Δw [i] ).

8. The array antenna fault correction method based on alternating projection according to claim 4, characterized in that: In step S304, the iteration termination condition includes any one of the following: The number of iterations reaches the preset maximum value; A compensation excitation vector with a smaller number of non-zero elements cannot be obtained in several consecutive iterations.

9. An electronic device, characterized in that: include: one or more processors; Memory; and One or more programs stored in a memory, wherein the one or more programs include instructions for executing the array antenna fault correction method as claimed in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: It comprises one or more programs for execution by one or more processors of an electronic device, wherein the one or more programs comprise instructions for executing the array antenna fault correction method as claimed in any one of claims 1 to 8.

Images