Coupling error calibration optimization method and system for radiation pattern of phased-array antenna
By abstracting the desired radiation pattern into a gate function and combining FFT and LS algorithms to optimize the coupling error of the phased array antenna, the problem of radiation pattern distortion caused by inter-element coupling is solved, and efficient radiation pattern calibration and specific beamforming are achieved.
Patent Information
- Application Number
- CN202511385281.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-26
- Publication Date
- 2026-01-20
AI Technical Summary
In phased array systems, the coupling effect between array elements causes the beamforming result to deviate from the expected value, resulting in pattern distortion. It is difficult to effectively reduce the mutual coupling effect by changing the antenna structure or polarization method.
By abstracting the desired radiation pattern into a gate function, the coupling error between array elements is optimized using FFT and LS algorithms, the radiation pattern is adjusted by combining a window function, and the excitation vector is solved by the least squares algorithm, thus realizing radiation pattern calibration and specific beamforming.
The phased array radiation pattern was effectively calibrated, reducing the cost of coupling error calibration, shortening the test time, and enabling flexible beam generation.
Smart Images
Figure CN121367522A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of wireless communication, in particular to a coupling error calibration optimization method and system for a phased array antenna radiation pattern. BACKGROUND
[0002] A phased array system is composed of a large number of antenna elements, and by adjusting the attenuator and phase shifter of each phased array antenna element, the amplitude and phase of its output can be changed to control the antenna pattern. However, in the array environment, due to the coupling effect between the elements, the beamforming result deviates from the expected value, causing pattern distortion. Therefore, these coupling errors must be properly compensated. Due to the limited spacing between the phased array antenna elements, it is difficult to integrate more complex antenna designs, so it is difficult to apply methods such as changing the structure and polarization of the antenna to reduce the mutual coupling effect in the phased array system. SUMMARY
[0003] In order to solve the problems existing in the prior art, the present application provides a coupling error calibration optimization method and system for a phased array antenna radiation pattern, which realizes pattern calibration and specific beamforming by optimizing the coupling error between the elements.
[0004] In a first aspect, the embodiments of the present application provide a coupling error calibration optimization method for a phased array antenna radiation pattern, comprising the following steps:
[0005] S1, setting an expected pattern, abstracting it as a gate function; defining a main lobe region of the expected pattern, and presetting an upper limit for a sidelobe region of the expected pattern; obtaining the active pattern of each element of the phased array through testing or simulation, and extracting the gain curve and phase curve of the active pattern of each element of the phased array;
[0006] S2, initializing the excitation of the phased array, calculating the initial pattern of the antenna; adjusting the sidelobe level amplitude of the initial pattern by a window function, so that the sidelobe level amplitude of the initial pattern is less than the preset upper limit of the sidelobe region, and adjusting the beam width to the defined main lobe region to obtain an adjusted pattern;
[0007] S3, performing FFT fast Fourier transform on the adjusted pattern, and applying a least squares algorithm to solve the excitation vector; calculating the first fitting pattern result of the solved excitation vector;
[0008] S4, setting a performance function to judge the performance, and setting a threshold value for the performance function, if the performance function is less than the threshold value, output the corresponding excitation vector, otherwise return to step S2, and assign the initial pattern in step S2 to the pattern function obtained by the inverse excitation vector excitation until the performance function is less than the threshold value or the allowed number of iterations is reached.
[0009] Preferably, the main lobe region of the desired pattern is defined as the region in the pattern function that is more than half of the maximum gain linear value, and is the convex width of the gate function; the maximum side lobe level amplitude of the desired pattern corresponds to the non-convex region of the gate function; if the desired pattern has a null region, the concave region of the gate function corresponds.
[0010] Preferably, step S2 comprises:
[0011] S21, in initializing the phased array excitation, finding the maximum value of the pattern pointing direction, performing a circular shift operation on the pattern, so that the angle corresponding to the maximum value of the pattern function is the preset beam direction θ0;
[0012] S22, using a window function T(θ) consistent with the length of the pattern, adjusting the pattern function to obtain an adjusted pattern function f'(θ).
[0013] Further, step S1 also pre-processes the gain curve g n (θ) and the phase curve φ n (θ) of the active pattern of each array element of the phased array to obtain a pre-processed active pattern h n (θ);
[0014] Step S21, in initializing the phased array excitation, the initial phase excitation is configured to only satisfy the alignment condition at the preset beam direction θ0, the amplitude excitation is each unit weight is 1, and the excitation vector corresponding to the nth array element is:
[0015]
[0016] Where d is the uniform spacing between the N antenna elements of the phased array antenna array; further calculation obtains the initial excitation vector ω0=[ω 10 ,ω 20 ,...,ω n0 ,...,ω N0 ];
[0017] Then, according to the pre-processed active pattern h n (θ) and the initial excitation vector ω0, the initial pattern f(θ) is calculated.
[0018] In a second aspect, the embodiments of the present application also provide a coupling error calibration optimization system for a phased array antenna radiation pattern, which is implemented by using the coupling error calibration optimization method, and the system comprises the following modules:
[0019] A preset module is expected to set a desired pattern, which is abstracted as a gate function, to define a main lobe region of the desired pattern, and to preset an upper limit for a side lobe region of the desired pattern; an active pattern of each array element of the phased array is obtained through testing or simulation, and a gain curve and a phase curve of the active pattern of each array element of the phased array are extracted;
[0020] A pattern adjustment module is configured to initialize excitation of the phased array, to calculate an initial pattern of the antenna, to adjust a side lobe level amplitude of the initial pattern through a window function, so that the side lobe level amplitude of the initial pattern is less than the preset upper limit of the side lobe region, and to adjust a beam width to the defined main lobe region to obtain an adjusted pattern;
[0021] A pattern fitting module is configured to perform FFT fast Fourier transform on the adjusted pattern, and to solve an excitation vector by using a least square algorithm; and a first fitting pattern result is calculated based on the solved excitation vector.
[0022] An iterative calculation module is configured to set a performance function to judge performance, and to set a threshold for the performance function; if the performance function is less than the threshold, a corresponding excitation vector is output, otherwise, the initial pattern in step S2 is assigned as a pattern function obtained by excitation of the inverse excitation vector, until the performance function is less than the threshold or a permitted number of iterations is reached.
[0023] Compared with the prior art, the technical scheme of the present application has the following advantages:
[0024] 1) The target pattern is abstracted as a gate function, and the pattern characteristics are extracted, which is beneficial to setting the beam width and the side lobe level indicators;
[0025] 2) DFT and LS algorithms are used to realize mutual conversion between the pattern and the excitation vector, and FFT is used to speed up the operation and reduce the algorithm running time;
[0026] 3) When the pattern is measured, only the active pattern of each antenna array element needs to be tested, and the remaining operations are performed in the computer, thereby reducing the test time and the coupling error calibration cost;
[0027] 4) The window function is introduced to realize pattern adjustment, and the beam can be flexibly generated while being calibrated. BRIEF DESCRIPTION OF DRAWINGS
[0028] Figure 1 Fig. 1 is a flowchart of a coupling error calibration optimization method for a phased array antenna radiation pattern in an embodiment of the present application;
[0029] Figure 2 Fig. 3 is a schematic diagram of a gate function used in an embodiment of the present application, wherein (a) is a gate function abstracted from a flat-top low side lobe beam, and (b) is a gate function abstracted from a wide-scan nulling beam;
[0030] Figure 3 The gain curve and phase diagram of the active radiation pattern of the phased array antenna element in the embodiment of the present invention;
[0031] Figure 4 The diagrams provided in this embodiment of the invention are the results of pattern coupling error calibration optimization, where (a) is the result of flat-top low sidelobe beam calibration optimization, and (b) is the result of wide-scan null beam calibration optimization. Detailed Implementation
[0032] The technical solution of the present invention will be further described clearly and completely below with reference to the embodiments and accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0033] Examples
[0034] It should be noted that the step numbers in this document are only for the convenience of explaining the specific embodiments and are not intended to limit the order in which the steps are executed. The method provided in this embodiment can be executed by a relevant computer, and the MATLAB simulation platform is used as the execution subject for illustration.
[0035] like Figure 1 As shown, this embodiment provides a coupling error calibration and optimization method for the radiation pattern of a phased array antenna. It achieves efficient calibration and optimization of the radiation pattern based on the Active Element Pattern (AEP). The active radiation patterns of the phased array antenna elements are tested, and the test results incorporate the coupling error of the phased array. The desired radiation pattern is abstracted as a gate function, and a window function is defined through the gate function to correct the radiation pattern with coupling error to the desired pattern. The Fast Fourier Transform (FFT) and Least Squares (LS) algorithm are used to inversely solve the excitation vectors of each element, and the phased array radiation pattern is fitted to the desired pattern through multiple iterations. This embodiment can achieve efficient calibration and optimization of the phased array radiation pattern while considering the coupling error between elements, and achieve specific beamforming. The coupling error calibration and optimization method of this embodiment specifically includes steps S1 to S4:
[0036] S1. Set the desired radiation pattern and abstract it as a gate function; define the main lobe region of the desired radiation pattern and preset the upper limit of the side lobe region of the desired radiation pattern; obtain the active radiation pattern of each element of the phased array through testing or simulation, and extract the gain curve and phase curve of the active radiation pattern of each element of the phased array.
[0037] In the present embodiment, the directional diagram is a normalized directional diagram, i.e. the maximum value of the directional diagram is 0. The main lobe region of the expected directional diagram is defined as the region in the directional diagram function that is more than half of the maximum gain linear value, i.e. the 3dB beam width, and is taken as the convex width of the gate function; the maximum side lobe level amplitude of the expected directional diagram corresponds to the non-convex region of the gate function; if there is a null region in the expected directional diagram, the concave region of the gate function corresponds to the null region.
[0038] As an example, in step S1, the gate function of the target beam is abstracted from the flat-top low side lobe beam and the wide-scan null beam, as shown in Figure 2 Specifically, for the flat-top low side lobe beam, the preset beam direction θ0 is set to 0°, the beam width is -20° to 20°, and the maximum side lobe level is -20dB; for the wide-scan null beam, the preset beam direction θ0 is set to 50°, the null region is -20° to -10°, the null level is -30dB, and the maximum side lobe level is -20dB. The optimization calibration of the phased array flat-top low side lobe radiation directional diagram is based on a 1x4 phased array to realize directional diagram coupling error calibration and flat-top low side lobe beam generation, and the active directional diagram data of the 1x4 phased array to realize directional diagram coupling error calibration and flat-top low side lobe beam generation is obtained by microwave darkroom testing; the optimization calibration of the phased array wide-scan null radiation directional diagram is based on a 1x8 phased array to realize directional diagram coupling error calibration and wide-scan null beam generation, and the active directional diagram data of the 1x8 phased array to realize directional diagram coupling error calibration and wide-scan null beam generation is obtained by simulation with commercial electromagnetic simulation software ANSYS HFSS, wherein the active directional diagram gain curve and phase diagram of the antenna array element are as shown in Figure 3 .
[0039] Further, a load is connected behind the other array elements, and then the active directional diagram of each array element is obtained by darkroom testing or simulation testing. In the darkroom testing, the test method of the active directional diagram of each array element is to connect a load behind the other array elements, to excite only the tested array element, and to measure the directional diagram as the active directional diagram of the tested array element. In the simulation testing, the test method of the active directional diagram of each array element is to set the amplitude excitation of the other array elements to 0w, to set the excitation amplitude of the tested array element to 1w, and to simulate the directional diagram as the active directional diagram of the tested array element.
[0040] In the present embodiment, the gain curve g n (θ) and the phase curve φ n (θ) of the active directional diagram of the satellite-borne phased array antenna element are extracted in the commercial electromagnetic simulation software ANSYS HFSS, as shown in Figure 3 .
[0041] The corresponding phased array antenna array is composed of N antenna array elements arranged at a uniform interval d along the x-axis, and the far-field radiation directional diagram can be written as:
[0042]
[0043] In the formula, In free space, the beam k = 2π / λ, where λ is the wavelength and ω is the wavelength. n =[ω0,ω1,...,ω N-1 ] represents the complex excitation vector; g n (θ) is the gain curve of the active pattern of the nth antenna obtained through testing or simulation.
[0044] Extract the phase curve φ from the active radiation pattern obtained through testing or simulation. n (θ), the gain curve and phase curve of the active radiation pattern are preprocessed using equation (2) to obtain the preprocessed active radiation pattern h. n (θ):
[0045] h n (θ)=g n (θ)cosφ n (θ)+jg n (θ)sinφ n (θ) (2)
[0046] Therefore, the far-field radiation pattern can be simplified as follows:
[0047]
[0048] S2. Initialize the phased array excitation and calculate the initial radiation pattern of the antenna using the AEP algorithm. Adjust the sidelobe level amplitude of the initial radiation pattern through a window function so that the sidelobe level amplitude of the initial radiation pattern is less than the upper limit of the preset sidelobe region, and adjust the beamwidth to the defined main lobe region to obtain the adjusted radiation pattern.
[0049] This step involves signal processing of the radiation pattern function to obtain the adjusted radiation pattern; specifically:
[0050] S21. First, during the initialization of the phased array excitation, the maximum value of the radiation pattern pointing direction is found. Due to the influence of errors, the radiation pattern pointing value of the initial excitation vector ω0 may deviate. Therefore, a cyclic shift operation is performed on the radiation pattern so that the angle corresponding to the maximum value of the radiation pattern function is the preset beam direction θ0.
[0051] More specifically, in the initial phased array excitation, the initial phase excitation is configured to satisfy the alignment condition only in the preset beam direction θ0, and the unit weight of the amplitude excitation is 1. The excitation vector corresponding to the nth array element is shown in equation (4):
[0052]
[0053] Where d is the uniform spacing between the N antenna elements of the phased array antenna array. According to equation (4), the initial excitation vector ω0=[ω 10 ,ω 20 ,...,ω n0 ,…,ω N0 ].
[0054] Substituting equation (4) into equation (3), and then based on the preprocessed active radiation pattern h... n Given the initial excitation vector ω0, the initial radiation pattern f(θ) is calculated using the AEP algorithm. The AEP algorithm is a method for synthesizing the full radiation pattern of a phased array based on the active radiation patterns of each array element.
[0055] Due to the influence of errors, the orientation of the initial excitation vector ω0 may deviate. Therefore, it is necessary to further perform a cyclic shift operation on the radiation pattern so that the angle corresponding to the maximum value of the radiation pattern function is the preset beam direction θ0.
[0056] S22. Next, the pattern function is adjusted by using a window function T(θ) that is consistent with the pattern length, resulting in the adjusted pattern function f'(θ).
[0057] Find the peak values of each sidelobe in the beam pattern function and their corresponding indices, then directly adjust the sidelobe levels to preset values. Simultaneously, based on the required beamwidth, normalize the gain value at the 3dB width of the main lobe (i.e., the main lobe region) to 0dB. This adjustment is achieved using the window function T(θ).
[0058] f'(θ)=f(θ)·T(θ)(5)
[0059] The adjusted pattern function f'(θ) is the desired pattern function, possessing the desired low sidelobe level and beamwidth.
[0060] In other words, the adjusted radiation pattern is obtained by multiplying the initial radiation pattern by the window function T(θ); the adjusted radiation pattern satisfies the preset conditions of the desired radiation pattern. In this embodiment, the window function T(θ) is generated as follows:
[0061] The initial radiation pattern is subjected to feature analysis and normalization to obtain the main lobe region and the maximum level of each side lobe of the initial radiation pattern. The maximum level of each side lobe is divided by the gate function abstracted from the preset conditions of the desired radiation pattern, interpolated and smoothed to obtain the window function T(θ).
[0062] In summary, in this step, the window function T(θ) is defined by the gate function abstracted from the flat-top low sidelobe beam and the wide-sweep null beam, and the pattern function is adjusted by Equation (5).
[0063] In this embodiment, the derivation process of the window function T(θ) expression is as follows:
[0064]
[0065] Where T′(θ) represents the discrete window function, p SLL For the desired maximum sidelobe level, Φ = [θ] left ,θ right ,θ i ,θ 0-left ,θ 0-right ], θ left Let θ be the index of the leftmost point of the pattern function f(θ). right Let θ be the index of the rightmost point in f(θ). i Let θ be the index of the maximum level of each sidelobe of f(θ). 0-left Let θ be the index of the first zero to the left of f(θ). 0-right Let θ0 be the index of the first zero point on the right side of f(θ), and θ0 be the preset beam direction. The resulting T′(θ) is a discrete window function composed of some discrete points, which is then subjected to cubic spline interpolation and smoothed to make its length consistent with that of the pattern function f(θ).
[0066] For flat-top low-sidelobe beams, index the flat-top region θ in the window function. flat The corresponding value is set to 0, thus obtaining the desired window function T(θ):
[0067]
[0068] More specifically, for flat-top low-sidelobe beams, the leftmost point f(θ) of the pattern function left ) and the rightmost point f(θ) right The value is set to -20dB, which is the expected maximum sidelobe level. The corresponding value at this index of the window function is -20 / f(θ). left ) and -20 / f(θ right Then, the sidelobe levels f(θ) of the pattern function are... i The value is set to -20dB, corresponding to the window function T'(θ). i )=-20 / f(θ i The first zero point f(θ) on the left side of the pattern function. 0-left ) and the first zero point on the right f(θ) 0-right When the value is set to -20dB, the corresponding window function value is -20 / f(θ). 0-left ) and -20 / f(θ) 0-right The preset beam direction θ0 remains unchanged, and the corresponding index value of the window function at that position is 1. The discrete window function T'(θ) is interpolated to match the length of the pattern function and smoothed. Then, the value corresponding to the flat-top region index in the window function is set to 0, thus obtaining the desired window function T(θ).
[0069] For wide-sweep null beams, the null regions in the window function are indexed by θ. trap The corresponding value is set to p trap / f(θ), p trap To achieve a zero-travel level, the desired window function T(θ) is obtained:
[0070]
[0071] More specifically, for wide-sweep null-notch beams, the leftmost point f(θ) of the pattern function... left ) and the rightmost point f(θ) right The value is set to -20dB, which is the expected maximum sidelobe level. The corresponding value at this index of the window function is -20 / f(θ). left ) and -20 / f(θ) right Then, the sidelobe levels f(θ) of the pattern function are... i The value is set to -20dB, corresponding to the window function T'(θ). i )=-20 / f(θ i The first zero point f(θ) on the left side of the pattern function. 0-left ) and the first zero point on the right f(θ) 0-right When the value is set to -20dB, the corresponding window function value is -20 / f(θ). 0-left ) and -20 / f(θ) 0-right The preset beam direction θ0 remains unchanged, and the corresponding index value of the window function at that position is 1. Interpolation is performed based on the obtained index and corresponding value to match the length of the pattern function, and then smoothed. Then, the null region p of the pattern function is... trap The value corresponding to the index is set to -30, which corresponds to the value of -30 / f(θ) at that index position in the window function. trap This yields the desired window function T(θ).
[0072] S3. Perform an N-point FFT (Fast Fourier Transform) on the adjusted radiation pattern and apply the LS (least squares) algorithm to solve for the excitation vector; use the AEP algorithm to calculate the first fitted radiation pattern result on the solved excitation vector.
[0073] Where N is the data length of the pattern function; the LS algorithm obtains the corresponding excitation vector by fitting the actual pattern function calculated by the excitation vector using the AEP algorithm.
[0074] In step S3, the mapping relationship between the pattern function and the excitation vector in the frequency domain is established using DFT according to equation (3). Constructing a linear system Hω in the frequency domain n =F, where The Fourier transform is represented and solved quickly using the FFT algorithm.
[0075] Using complex least square method, since g n (θ) is a fixed value, thus realizing the mutual conversion of f(θ) and ω n Solve by least square method:
[0076]
[0077] The least square solution ω n = (H H H) -1 H H F, the first inverse solution excitation vector value is obtained Also called excitation vector, realizing the mutual conversion of the pattern function and the excitation vector.
[0078] Further, the obtained excitation vector value ω The first fitting pattern result is calculated by using AEP algorithm, specifically:
[0079] According to the pre-processed active pattern function, the excitation vector value ω The first fitting calibration optimized pattern function f”(θ) is obtained.
[0080] S4, set performance function to judge performance, and set threshold value for performance function, if performance function is less than threshold value, output corresponding excitation vector, otherwise return to step S2, let f(θ) = f”(θ), that is, assign the initial pattern in step S2 to the pattern function excited by the inverse solution excitation vector, until the performance function is less than the threshold value or the allowed number of iterations is reached.
[0081] Due to the nonlinearity of the adjustment of the pattern function f(θ), the obtained excitation vector value ω n The pattern function f”(θ) excited by the excitation vector and the adjusted pattern function f'(θ) will have differences. Therefore, an iterative method is used to give a performance function L as an error function to judge the pattern calibration optimization result. The performance function L has two indexes: beam width index and sidelobe level index; the beam width index is the absolute value of the difference between the pattern beam bandwidth (i.e. the maximum value of the main lobe of the pattern-3dB range) and the expected beam width; the sidelobe level index is the cumulative value of the absolute value of the difference between each sidelobe peak value of the pattern function and the expected sidelobe level.
[0082] For the beam width index, let the main lobe maximum value-3dB range of the pattern be the pattern beam bandwidth α, and the expected beam width be α0; if α and α0 are inconsistent, then the beam width index loss1 = |α-α0|.
[0083] For the sidelobe level index, find each sidelobe peak value p i, directly compare it with the desired side lobe level p0, obtain and accumulate the absolute value of each difference, and obtain the side lobe level index M is the number of side lobes.
[0084] Therefore, the performance function L (i.e. error function) can be written as:
[0085] L = loss1 + loss2 (10)
[0086] The threshold of the performance function is set, and if the performance function L is less than the threshold, the corresponding excitation vector value ω is output n , otherwise f(θ) = f"(θ) is set, and the pattern function adjustment is performed again. The threshold R = 3 can be set, and the maximum number of iterations iter = 5. After multiple iterations, until the performance function is less than the threshold or the allowed number of iterations is reached.
[0087] In step S4, the performance function is set according to the gate function abstracted from the flat-top low side lobe beam and the wide-scan nulling beam, as shown in equation (10). After multiple iterations, until the performance function is less than the threshold or the allowed number of iterations is reached.
[0088] Based on the same inventive concept, the embodiment also provides a coupling error calibration optimization system for a phased array antenna radiation pattern, which is implemented by using the coupling error calibration optimization method. The coupling error calibration optimization system of the embodiment includes the following modules:
[0089] An expected preset module is configured to abstract an expected pattern into a gate function, define a main lobe region of the expected pattern, preset an upper limit for a side lobe region of the expected pattern, obtain an active pattern of each array element of the phased array through testing or simulation, and extract a gain curve and a phase curve of the active pattern of each array element of the phased array.
[0090] A pattern adjustment module is configured to initialize excitation of the phased array, calculate an initial pattern of the antenna, adjust a side lobe level amplitude of the initial pattern through a window function so that the side lobe level amplitude of the initial pattern is less than the preset upper limit of the side lobe region, and adjust a beam width to the defined main lobe region to obtain an adjusted pattern.
[0091] A pattern fitting module is configured to perform FFT (Fast Fourier Transform) on the adjusted pattern, and solve an excitation vector by using a least square algorithm.
[0092] An iteration calculation module is configured to set a performance function to judge performance, set a threshold for the performance function, output a corresponding excitation vector if the performance function is less than the threshold, or return to step S2, assign the initial pattern in step S2 to a pattern function excited by the inverse excitation vector, until the performance function is less than the threshold or the allowed number of iterations is reached.
[0093] The above modules are respectively implemented by steps S1-S4, and the detailed implementation process can be referred to the description of the above steps.
[0094] The embodiment provides results of realizing a directional diagram coupling error calibration and a flat-top low-sidelobe beam based on a 1*4 phased array and results of realizing a directional diagram coupling error calibration and a wide-scan nulling beam based on a 1*8 phased array, as shown in the following table. Figure 4 The satellite-borne phased array antenna radiation directional diagram coupling error calibration optimization method provided by the embodiment effectively realizes a target directional diagram function index, and completes phased array beam calibration and specific beam forming under the consideration of the coupling error.
[0095] Those skilled in the art can understand that all or part of the processes in the above embodiment can be completed by completing the related work and computer program instructions.
[0096] The above is the preferred embodiment of the present application. It should be noted that those skilled in the art can make several improvements and refinements without departing from the principles of the present application, and these improvements and refinements are also considered within the protection scope of the present application.
Claims
1. A method for calibrating and optimizing coupling errors of a phased array antenna radiation pattern, characterized in that, The method comprises the following steps: S1, setting an expected directional diagram, abstracting it as a gate function; defining a main lobe region of the expected directional diagram, and presetting an upper limit for a side lobe region of the expected directional diagram; Obtaining an active directional diagram of each array element of the phased array through testing or simulation, and extracting a gain curve and a phase curve of the active directional diagram of each array element of the phased array; S2, initializing the excitation of the phased array, and calculating an initial directional diagram of the antenna; Adjusting a side lobe level amplitude of the initial directional diagram through a window function, so that the side lobe level amplitude of the initial directional diagram is less than the preset upper limit of the side lobe region, and adjusting a beam width to the defined main lobe region to obtain an adjusted directional diagram; S3, performing FFT fast Fourier transform on the adjusted directional diagram, and solving an excitation vector by using a least square algorithm; and calculating a first fitting directional diagram result of the solved excitation vector; S4, setting a performance function to judge performance, and setting a threshold value for the performance function; if the performance function is less than the threshold value, outputting a corresponding excitation vector, otherwise returning to step S2, assigning the initial directional diagram in step S2 to a directional diagram function obtained by excitation of the inverse excitation vector, until the performance function is less than the threshold value or the allowed number of iterations is reached.
2. The coupling error calibration optimization method of claim 1, wherein, The main lobe region of the expected directional diagram is a region in the directional diagram function that is more than half of the maximum gain linear value, and is used as the convex width of the gate function; the maximum side lobe level amplitude of the expected directional diagram corresponds to the non-convex region of the gate function; if there is a null region in the expected directional diagram, the null region corresponds to the concave region of the gate function.
3. The coupling error calibration optimization method of claim 1, wherein, Step S2 comprises: S21, in the initialization of the excitation of the phased array, finding a maximum value of a directional diagram pointing direction, and performing a circular shift operation on the directional diagram, so that an angle corresponding to a maximum value of the directional diagram function is a preset beam direction θ0; S22, using a window function T(θ) consistent with the length of the directional diagram to adjust the directional diagram function, to obtain an adjusted directional diagram function f'(θ).
4. The coupling error calibration optimization method of claim 3, wherein, Step S1 also pre-processes the gain curve g n (θ) of the active pattern of each array element of the phased array to obtain a pre-processed active pattern h n (θ) of each array element of the phased array. n (θ). In step S21, in the initialization of the excitation of the phased array, the initial phase excitation is configured to only satisfy the alignment condition at the preset beam direction θ0, the amplitude excitation has each unit weight being 1, and the excitation vector corresponding to the nth array element is: wherein d is the uniform interval between the N antenna elements of the phased array antenna array; an initial excitation vector ω0=[ω0,...,ω0,...,ω0] is further calculated. 10 20 n0 N0 Further, the active directivity pattern h n (θ) with the initial excitation vector ω0, the initial directivity pattern f(θ) is calculated.
5. The coupling error calibration optimization method of claim 3, wherein, The adjustment method of the directional diagram function in step S22 is: Finding each side lobe peak value of the directional diagram function and the corresponding index, directly adjusting the side lobe level of the directional diagram to a preset value; and simultaneously normalizing the gain value corresponding to the main lobe region of the directional diagram function to 0 dB according to the required beam width.
6. The coupling error calibration optimization method of claim 3, wherein, The generation method of the window function in step S22 is: Performing characteristic analysis on the initial directional diagram, obtaining the main lobe region of the initial directional diagram and the maximum level of each side lobe of the initial directional diagram through normalization processing; dividing each side lobe maximum level by the gate function abstracted through the preset condition of the expected directional diagram, interpolating and smoothing to obtain the window function T(θ).
7. The coupling error calibration optimization method of claim 1, wherein, The performance function of step S4 is provided with a beam width index and a side lobe level index; the beam width index is an absolute value of a difference between a directional diagram beam bandwidth and an expected beam width; and the side lobe level index is a cumulative value of absolute values of differences between each side lobe peak value of the directional diagram function and an expected side lobe level.
8. The coupling error calibration optimization method of claim 7, wherein, In the performance function, for the beam width index, let the maximum value of the main lobe of the directional diagram-3dB range be the directional diagram beam bandwidth α, and the expected beam width be α0; if α is not consistent with α0, then there is a beam width index loss1=|α-α0|; For the sidelobe level indicator, find each sidelobe peak value p of the calibrated optimized directional pattern function f"(θ) i directly compare it with the expected sidelobe level p0, calculate and accumulate the absolute value of each difference, and obtain the sidelobe level indicator M is the number of sidelobes; The performance function L is: L=loss1+loss2.
9. The coupling error calibration optimization method of claim 1, wherein, In step S4, the threshold R=3 is set, and the maximum iteration number iter=5.
10. A coupling error calibration and optimization system for the radiation pattern of a phased array antenna, characterized in that, The coupling error calibration optimization method is implemented by using the system of any one of claims 1-9, and the system comprises the following modules: An expected preset module is configured to set an expected directional diagram, abstracted as a gate function; define a main lobe region of the expected directional diagram, and preset an upper limit for a side lobe region of the expected directional diagram; An active directional diagram of each array element of the phased array is obtained through testing or simulation, and a gain curve and a phase curve of the active directional diagram of each array element of the phased array are extracted; A directional diagram adjustment module is configured to initialize excitation of the phased array, and calculate an initial directional diagram of the antenna; A window function is used to adjust a side lobe level amplitude of the initial directional diagram, so that the side lobe level amplitude of the initial directional diagram is less than the preset upper limit of the side lobe region, and the beam width is adjusted to the defined main lobe region, to obtain an adjusted directional diagram; A directional diagram fitting module is configured to perform FFT fast Fourier transform on the adjusted directional diagram, and solve an excitation vector by using a least square algorithm; and calculate a first fitting directional diagram result based on the solved excitation vector. An iteration calculation module is configured to set a performance function to judge performance, and set a threshold for the performance function; if the performance function is less than the threshold, the corresponding excitation vector is output, otherwise, step S2 is returned, the initial directional diagram in step S2 is assigned as a directional diagram function obtained by excitation of the inverse excitation vector, until the performance function is less than the threshold or the allowed iteration number is reached.