Phase-free array antenna fault diagnosis method based on Bayesian compressed sensing

Abstract

The invention relates to a Bayesian compressed sensing-based phase-free array antenna fault diagnosis method. The method comprises the following steps of: acquiring phase-free radiation field measurement data of a damaged array antenna; extracting an active array element directional diagram of the array antenna, and constructing a virtual differential array mathematical model based on the phase-free radiation field measurement data; the virtual differential array mathematical model is linearized through a phase-free mapping method, and a twinborn equation set is constructed; solving the twinborn equation set by using a multi-task Bayesian compressed sensing algorithm to obtain virtual differential array excitation information; and obtaining a fault diagnosis result according to the virtual differential array excitation information. Compared with the prior art, the method has the advantages that the fault unit in the antenna array can be quickly and accurately diagnosed without phase information, and the method is particularly suitable for a large-scale antenna array in a high-frequency and complex environment.

Description

Technical Field

[0001] The present invention belongs to the field of electronic technology and relates to an array antenna fault diagnosis technology, in particular to a phaseless array antenna fault diagnosis method based on Bayesian compressed sensing, which is suitable for array antenna system fault diagnosis in the fields of modern wireless communication systems, radar and remote sensing. Background Art

[0002] With the rapid development of modern communication technology, array antennas are widely used due to their high gain and directional control capabilities. However, the performance of antenna arrays is highly dependent on the correct excitation of each antenna unit. Due to environmental factors and equipment aging, antenna units may fail, which in turn affects the performance of the entire antenna array. Traditional fault diagnosis techniques rely on phase information, but it is challenging to obtain accurate phase information in high frequency bands. Therefore, it is of great practical value to develop a method for fault detection without phase information.

[0003] At present, there are some methods for diagnosing faults of phase-free array antennas. For example, patent application CN116735987A proposes an alternating iterative method based on a phase error matrix for array antenna diagnosis, but this method is only applicable to far-field measurements and cannot achieve effective diagnosis from more complex near-field measurement data. Patent application CN116992355A proposes a method for diagnosing faults of phase-free array antennas based on artificial neural networks and compressed sensing, but artificial neural networks require a pre-training process, which is time-consuming and not suitable for rapid diagnosis.

[0004] It can be seen that due to the limitations of existing methods, it is impossible to achieve efficient, accurate and applicable phased array antenna fault diagnosis effects. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a phaseless array antenna fault diagnosis method based on Bayesian compressed sensing, which can quickly and accurately diagnose the faulty unit in the antenna array without the need for phase information.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A method for fault diagnosis of phaseless array antenna based on Bayesian compressed sensing comprises the following steps:

[0008] Obtain the phase-free radiation field measurement data of the damaged array antenna;

[0009] Extracting the active element pattern of the array antenna, and constructing a virtual differential array mathematical model based on the phase-free radiation field measurement data;

[0010] The virtual differential array mathematical model is linearized by a phase-free mapping method, and a twin equation group is constructed;

[0011] Using a multi-task Bayesian compressed sensing algorithm to solve the twin equations to obtain virtual differential array excitation information;

[0012] A fault diagnosis result is obtained according to the virtual differential array excitation information.

[0013] Furthermore, the near-field scanning or far-field scanning is obtained by near-field scanning or far-field scanning.

[0014] Furthermore, when only the nth array element is excited and other array elements are connected to matching loads, the radiation field of the antenna array is simulated to extract and obtain the active array element directivity pattern.

[0015] Furthermore, the construction of the virtual differential array mathematical model specifically includes:

[0016] Based on the active array element pattern, the radiation field E of the damaged antenna array is obtained. F (r) and the corresponding radiation field E of the original lossless antenna array G (r):

[0017]

[0018]

[0019] Where N represents the number of array elements, and denote the unit excitations of the damaged array and the original array, respectively, ψ n (r) represents the active array element pattern, n is the number of the array element, and r is the position of the electric field sampling point;

[0020] The radiation field of the damaged array and the original array is differentiated to obtain a virtual differential array with sparse element excitation. Its radiation field E D (r) is expressed as:

[0021]

[0022] in, is the excitation of the differential array, and The sparsity of is S, where S is the number of faulty array elements;

[0023] Based on the measurement data of phase-free radiation field|E F (r)|, construct the mathematical model of virtual differential array:

[0024] |E G (r)| 2 -|EF (r)| 2 =2Re[E G (r)E D (r) * ]-|E D (r)| 2

[0025] Among them, Re represents the real part operation, and * is the conjugate operator.

[0026] Furthermore, after the virtual differential array mathematical model is linearized into a matrix equation, the twin equation group is constructed by exchanging the positions of the real and imaginary elements in the matrix equation.

[0027] Furthermore, the matrix equation is expressed as:

[0028] ΨI=E

[0029] in, I is the unknown vector to be found, Ψ is the measurement matrix, Represent the field values ​​corresponding to the mth sampling point of the original array and the damaged array respectively.

[0030] Furthermore, when the multi-task Bayesian compressed sensing algorithm is used to solve the twin equations, the statistical correlation between the real and imaginary information in the data is considered.

[0031] Furthermore, the fault diagnosis result includes the location of the faulty unit and the fault type.

[0032] Further, according to the virtual differential array excitation information, obtaining the fault diagnosis result is specifically:

[0033] In the virtual differential array excitation information, the position of the array element whose excitation value is greater than the set value is the position of the faulty unit;

[0034] The damaged array excitation is obtained based on the virtual differential array excitation information, and the amplitude and phase values ​​of the damaged array excitation and the original array excitation are compared to determine the fault type of the faulty unit.

[0035] 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 phaseless array antenna fault diagnosis method as described above.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. The present invention collects the electric field data of the antenna array through near-field or far-field measurement, does not rely on phase information, and reduces the technical requirements and difficulty of near-field or far-field measurement.

[0038] 2. The present invention establishes a linearized model based on Bayesian compressed sensing through a phase-free mapping method, transforming the nonlinear phase-free fault diagnosis problem into a linear problem. The model uses statistical correlation to process real and imaginary information to enhance the accuracy and robustness of fault detection. At the same time, the multi-task Bayesian compressed sensing (MT-BCS) algorithm is applied to solve the linear problem, and the faulty unit of the antenna array is identified based on a small amount of phase-free measurement data. It can effectively process highly coherent electric field data, quickly and accurately locate and identify faults, and improve the accuracy and efficiency of fault diagnosis.

[0039] 3. The present invention can quickly and accurately identify and locate faulty units in antenna arrays in complex environments, and is particularly suitable for real-time online monitoring and maintenance of large-scale antenna arrays in high-frequency and complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the process of the present invention;

[0041] Figure 2 The figure is a diagnostic result for an ideal 32-unit Chebyshev linear array in an embodiment of the present invention, specifically the amplitude distribution and phase distribution of virtual differential array excitation. The figure compares the results of the method of the present invention and the convex optimization method with the true value;

[0042] Figure 3 1 is the diagnostic result of an actual 5×5 dipole array antenna in an embodiment of the present invention, where (a) is the amplitude distribution of the virtual differential array excitation, and (b) is the phase distribution. The figure compares the results of the method of the present invention and the convex optimization method with the actual situation. DETAILED DESCRIPTION

[0043] 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.

[0044] refer to Figure 1 As shown, this embodiment provides a method for diagnosing faults of a phaseless array antenna based on Bayesian compressed sensing, and the specific implementation steps of the method include:

[0045] Step S1: Acquire the phase-free radiation field measurement data of the damaged array antenna.

[0046] In this embodiment, the radiated electric field amplitude data of the damaged array antenna is obtained by near-field scanning or far-field scanning, which does not include phase information, and the number of sampling points is M.

[0047] Step S2: Extract the active element pattern of the array antenna and construct a virtual difference array mathematical model based on the non-phase radiation field measurement data.

[0048] In this embodiment, the active element pattern ψ n (r) of the array antenna is extracted through full-wave simulation or other methods, where n is the serial number of the element and r is the position of the electric field sampling point.

[0049] The specific extraction method in this embodiment is: when only the nth element is excited and other elements are connected with matching loads, simulate the radiation field of the antenna array, denoted as ψ n (r). Using the active element pattern can consider the mutual coupling between the antenna array elements and improve the modeling accuracy. Furthermore, the radiation field E F (r) of the damaged antenna array can be expressed as:

[0050]

[0051] The radiation field E G (r) of its corresponding original undamaged antenna array is expressed as:

[0052]

[0053] Among them, N represents the number of array elements, and represent the element excitations of the damaged array and the original array respectively. Let the number of faulty elements be S. Generally, the number of faulty elements is much smaller than the total number of elements, that is, S << N. Taking the difference between the radiation fields of the above-mentioned damaged array and the original array, a virtual difference array with sparse characteristics of element excitation can be obtained, and its radiation field is expressed as:

[0054]

[0055] Among them, is the excitation of this difference array, and has a sparsity of S.

[0056] According to the non-phase radiation field measurement data of the damaged array antenna obtained in step S1, that is, the radiation field amplitude information |E F (r)|, perform a difference operation on the squares of |E G (r)| and |E F (r)| to obtain the following mathematical equation:

[0057] |E G (r)| 2 -|E F (r)| 2 =2Re[E G (r)E D(r) * -E D (r)| 2

[0058] Among them, Re represents the operation of taking the real part, and * is the conjugate operator. The mathematical equation describes a nonlinear problem regarding the excitation of a virtual difference array.

[0059] Step S3: Linearize the above nonlinear problem through a non-phase mapping method and construct a twin equation set.

[0060] In order to transform the nonlinear problem into a linear problem that is easy to solve, the present embodiment adopts the following non-phase mapping method, which specifically includes:

[0061] Expand the two terms 2Re[E G (r)E D (r) * and -|E D (r)| 2 on the right side of the nonlinear mathematical equation respectively as:

[0062]

[0063] and

[0064]

[0065] Among them, and ψ m,n respectively represent the field values corresponding to the m-th sampling point. Im represents the operation of taking the imaginary part. n1 and n2 represent the serial numbers of the array elements, and 1 ≤ n1 < n2 ≤ N. The above expansion shows that For the variables and it is linear, For the variables and it is linear. Therefore, the nonlinear problem is successfully transformed into a linear problem through the non-phase mapping method and can be expressed in the form of the following matrix equation:

[0066] ΨI = E

[0067] Among them, The unknown vector I to be solved = [I1; I2], specifically

[0068]

[0069] and

[0070]

[0071] Among them, the superscript T represents the transpose operator. I1 and I2 are 2N-dimensional and N respectively2 dimensional real column vector, I is the concatenation of I1 and I2. The elements of the measurement matrix Ψ are based on the and The expansion of the two terms is determined.

[0072] Then, the twin equation of the matrix equation is constructed to take into account the statistical correlation between the real and imaginary parts of the differential array excitation. The twin equation is specifically:

[0073] Ψ t I t =E

[0074] Among them, t is the twin matrix of Ψ, obtained by swapping the real and imaginary elements of the internal symmetry of Ψ, i.e., swapping and Location and and Location. t =[I 1t ;I 2t ], I 1t and I 2t are the twin vectors of I1 and I2 respectively. Specifically, I 1t and I 2t They are

[0075]

[0076] and

[0077]

[0078] Therefore, there are the following twin equations to be solved:

[0079]

[0080] Step S4: Solve the twin equations using the multi-task Bayesian compressed sensing (MT-BCS) method.

[0081] The twin equations are solved using the multi-task Bayesian compressed sensing (MT-BCS) algorithm to obtain a set of sparse and statistically related solutions I and I t This process is and The same sparsity constraints are imposed to enhance the accuracy and robustness of the solution.

[0082] Since I=[I1;I2] and I t =[I 1t ;I 2t ], solve for I and I t , that is, I1 and I 1tThe excitation of the virtual differential array can be based on I1 and I 1t The elements in are expressed as follows:

[0083]

[0084] in, is the imaginary number symbol, I1(n) and I 1t (n) represent I1 and I respectively 1t The nth element in represents the excitation of the nth array element in the virtual differential array, n = 1, 2, ...N.

[0085] Step S5: Obtain a fault diagnosis result according to the virtual differential array excitation information obtained by solving.

[0086] According to the excitation of the virtual differential array obtained Determine the location and fault type of the faulty element in the damaged array antenna. Specifically, in the virtual differential array, the location of the element with significant non-zero excitation is the location of the faulty element. The excitation of the original array is known, The excitation of the damaged array can be inferred by the excitation of the virtual differential array obtained. Comparing damaged array excitation With the original array excitation The amplitude and phase values ​​of the faulty unit can be used to determine the fault type, including switch fault, current amplitude and phase fault, position fault, failure fault, etc.

[0087] If the above method is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0088] In order to verify the accuracy of the present invention, the method of the present invention is used to diagnose an ideal 32-unit Chebyshev linear array, in which the number of faulty units is 3, the type is complete damage (ie, the excitation of the faulty array element is 0), and the signal-to-noise ratio is 30 dB. Figure 2 The amplitude and phase distribution of the virtual differential array excitation reconstructed using the method of the present invention are compared with the results obtained using the convex optimization method. It can be seen that the results obtained using the method of the present invention are highly consistent with the true value, which is better than the convex optimization method, proving the accuracy of the algorithm of the present invention.

[0089] Furthermore, the method of the present invention is used to diagnose an actual 5×5 dipole array antenna, wherein the number of faulty units is 3, the types are complete damage, amplitude fault and phase fault, and the signal-to-noise ratio is 30 dB. Figure 3 The amplitude and phase distribution of the virtual differential array excitation reconstructed by the method of the present invention are compared with the results obtained by the convex optimization method. It can be seen that the results obtained by the method of the present invention are highly consistent with the true value and are better than the convex optimization method, which proves the accuracy and effectiveness of the algorithm of the present invention in diagnosing actual antenna arrays.

[0090] In another embodiment, an electronic device is provided, including one or more processors, a memory, and one or more programs stored in the memory, wherein the one or more programs include instructions for executing the method described above.

[0091] 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 fault diagnosis of phaseless array antenna based on Bayesian compressed sensing. It is characterized in that The following steps are involved: Obtain the phase-free radiation field measurement data of the damaged array antenna; Extracting the active element pattern of the array antenna, and constructing a virtual differential array mathematical model based on the phase-free radiation field measurement data; The virtual differential array mathematical model is linearized by a phase-free mapping method, and a twin equation group is constructed; Using a multi-task Bayesian compressed sensing algorithm to solve the twin equations to obtain virtual differential array excitation information; A fault diagnosis result is obtained according to the virtual differential array excitation information.

2. According to the method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing in claim 1, It is characterized in that The near-field scanning or far-field scanning is obtained by a near-field scanning or a far-field scanning.

3. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 1, It is characterized in that When only the nth array element is excited and other array elements are connected to matching loads, the radiation field of the antenna array is simulated to extract and obtain the active array element directivity pattern.

4. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 1, It is characterized in that The construction of the virtual differential array mathematical model specifically includes: Based on the active array element pattern, the radiation field E of the damaged antenna array is obtained. F (r) and the corresponding radiation field E of the original lossless antenna array G (r): Where N represents the number of array elements, and denote the unit excitations of the damaged array and the original array, respectively, ψ n (r) represents the active array element pattern, n is the number of the array element, and r is the position of the electric field sampling point; The radiation field of the damaged array and the original array is differentiated to obtain a virtual differential array with sparse element excitation. Its radiation field E D (r) is expressed as: in, is the excitation of the differential array, and The sparsity of is S, where S is the number of faulty array elements; Based on the measurement data of phase-free radiation field|E F (r)|, construct the mathematical model of virtual differential array: |And G (r)| 2 -|And F (r)| 2 =2Re[E G (re D (r) * ]-|And D (r)| 2 Among them, Re represents the real part operation, and * is the conjugate operator.

5. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 1, It is characterized in that After linearizing the virtual differential array mathematical model into a matrix equation, the twin equation group is constructed by exchanging the positions of the real and imaginary elements in the matrix equation.

6. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 5, It is characterized in that The matrix equation is expressed as: ΨI=E in, I is the unknown vector to be found, Ψ is the measurement matrix, Represent the field values ​​corresponding to the mth sampling point of the original array and the damaged array respectively.

7. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 1, It is characterized in that When using the multi-task Bayesian compressed sensing algorithm to solve the twin equations, the statistical correlation between the real and imaginary information in the data is considered.

8. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 1, It is characterized in that The fault diagnosis result includes the location of the faulty unit and the fault type.

9. The method for fault diagnosis of non-phased array antenna based on Bayesian compressed sensing according to claim 8, It is characterized in that According to the virtual differential array excitation information, obtaining the fault diagnosis result is specifically: In the virtual differential array excitation information, the position of the array element whose excitation value is greater than the set value is the position of the faulty unit; The damaged array excitation is obtained based on the virtual differential array excitation information, and the amplitude and phase values ​​of the damaged array excitation and the original array excitation are compared to determine the fault type of the faulty unit.

10. A computer-readable storage medium, It is 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 method for diagnosing faults of a phaseless array antenna as claimed in any one of claims 1 to 9.

Images