Gradient-based beam peak searching method and device

Abstract

The invention discloses a gradient-based beam peak searching method, which relates to the technical field of antenna testing, and comprises the following steps of: firstly, setting the sizes of measurement grids in a spherical coordinate system along a horizontal angle direction and a pitch angle direction according to the half-power bandwidth of an antenna to be tested; and interpolating the measured radiation power value on the test grid to obtain a radiation pattern in a grid coverage range. And then, searching a local maximum point of the radiation pattern, taking the position of the local maximum point as the center of a region where a wave beam peak value may appear, and performing secondary measurement around the local maximum point to obtain radiation power. And finally, taking the maximum value of the radiation power in the secondary measurement as an initial position, iteratively updating a test position by using the gradient of the current position, and outputting a peak position and the size of the radiation power at the peak. The invention further discloses a gradient-based wave beam peak value searching device, and the required testing times can be remarkably reduced under the condition that the searching result precision is ensured.

Description

Technical Field

[0001] The invention relates to the technical field of antenna testing, and in particular to a gradient-based beam peak search method and device. Background Art

[0002] In modern wireless communication systems, antenna beamforming technology plays a key role in massive multiple-input multiple-output (MIMO) and millimeter-wave applications. With the development of the fifth-generation mobile communication system (5G) and the future sixth-generation mobile communication system (6G), beamforming antenna systems are widely used in base stations and user equipment to achieve enhanced signal coverage, improved data transmission rate and spectrum efficiency. In order to optimize the beamforming performance of the antenna, beam peak search has become an important test item to quickly locate the direction with the largest signal gain in the antenna radiation pattern, that is, the main beam peak position. The beam peak position not only affects the design of the antenna, but is also directly related to the system signal quality and user experience.

[0003] Traditional beam peak search methods usually determine the peak direction by traversing the radiation power in the three-dimensional (3D) test space. In the standards of 3GPP and CTIA, the specified beam peak test methods include the equal step method and the equal density method. The equal step method is to measure at a certain angle interval in the azimuth and elevation dimensions of the spherical coordinate system. After discussion, to ensure the accuracy of the test results, at least the test step size needs to be set to 7.5° for a total of 1106 measurements. The equal density method divides the sphere into a specified number of regions of equal area based on the Voronoi Region. To ensure the accuracy of the test results, at least 800 regions need to be set, that is, 800 measurements.

[0004] However, as the size of antenna arrays increases, the efficiency of traditional methods can no longer meet the demand. Massive MIMO antenna systems may contain dozens or even hundreds of antenna units, and point-by-point measurement is extremely time-consuming and computationally intensive. Especially in high-frequency millimeter-wave communications, the beams are narrower and the patterns are more complex. Measuring each angle is not only time-consuming, but also prone to introducing measurement errors. Therefore, in order to improve test efficiency and accuracy, the fast beam peak search method came into being.

[0005] The core idea of ​​fast beam peak search technology is to avoid traversal measurement of beam direction through optimization methods, and locate the peak position of the main beam with fewer tests. This technology also plays an important role in beam tracking in dynamic environments. For example, in a scenario where user equipment or base station antennas are moving in real time, fast beam peak search can quickly adjust the beam direction to ensure that the signal is always aimed at the target, thereby improving user experience and network reliability. As the complexity of wireless communication scenarios continues to increase, fast beam peak search technology will continue to evolve in the future to support a wider range of applications and higher performance requirements.

[0006] Common optimization methods include genetic algorithms and particle swarm algorithms. In fast beam peak search, these methods have global search capabilities during the optimization process, and are particularly effective in applications with large-scale antenna arrays and complex directional patterns. However, genetic algorithms simulate biological evolution processes for selection, crossover, and mutation. This process requires multiple generations of iterations to gradually approach the optimal solution, so the computational overhead is large. The particle swarm algorithm relies on the collaboration of a group of particles to find the optimal solution by updating the speed and position of each particle. This also requires a large number of iterations and a large number of candidate points to achieve convergence. Therefore, there is an urgent need for a method that can converge quickly with fewer test points and iterations, so as to significantly improve the computational efficiency while ensuring the optimization effect. Summary of the invention

[0007] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a gradient-based beam peak search method and device, which can perform rapid and high-precision beam peak search and positioning on any antenna to be tested.

[0008] The present invention adopts the following technical solutions to solve the above technical problems:

[0009] A gradient-based beam peak search method proposed in the present invention includes:

[0010] First, fix the device under test on a turntable, and make the geometric center of the device under test located at the center of the test area. Take the geometric center of the device under test as the origin of the spherical coordinate system, take the distance between the signal receiving device and the device under test as the radial distance of the spherical coordinate system, set the test interval in the elevation angle and azimuth angle dimensions of the spherical coordinate system according to the half-power bandwidth on the electric and magnetic planes of the antenna in the device under test, and form a test grid with the azimuth angle and elevation angle at a fixed radial distance as coordinates;

[0011] According to the angle position indicated by each grid point in the test grid, the turntable is used to control the direction of the device under test, and the angle indicated by the test grid is traversed. After the radiation power at the angle indicated by each grid point in the test grid is measured, the test grid is interpolated by the interpolation method to estimate the complete radiation pattern in the test area;

[0012] Next, the positions of the local maximum points are obtained from the estimated complete radiation pattern, and the positions of these local maximum points are used as the center positions of the secondary measurement areas; around the center position of each secondary measurement area, the positions of the required multiple secondary measurements are determined at a distance less than the grid spacing in the test grid, and the radiation power sets are measured at these secondary measurement positions;

[0013] Finally, the maximum radiation power is obtained from the set of radiation powers measured twice. The position corresponding to the maximum radiation power in the secondary measurement is used as the initial position. According to the gradient search algorithm, the test position is updated each time. When the gradient search algorithm converges to a certain position and does not change, the position and the radiation power measured at the position are output.

[0014] As a further optimization scheme of the gradient-based beam peak search method described in the present invention, around the center position of each secondary measurement area, including in the orthogonal direction and diagonal direction of the center position of each secondary measurement area.

[0015] As a further optimization scheme of the gradient-based beam peak search method described in the present invention, the specific steps are as follows:

[0016] Step 1: In the spherical coordinate system with the geometric center of the device under test as the origin The range of the pitch angle θ is -90°≤θ≤90°, and the azimuth angle The range is r is the radial distance; within the range of elevation angle variation, the half-power beam width θ of the antenna in the device under test on the θ plane 3dB Select N angles at intervals, and take the antenna to be tested as Half-power beamwidth on the plane Select M angles at intervals to form a test grid with dimensions of N×M; change the orientation of the device under test by using the turntable n=1...N,m=1...M,θ n is the selected nth pitch angle, For the selected mth horizontal angle, use the signal receiving device to The radiated power measured at to receive;

[0017] Step 2: Use the interpolation method to interpolate the test grid to obtain a complete radiation pattern after interpolation. There are V local maximum points in the radiation pattern. These local maximum points are the center points of the secondary measurement area. The position of the vth local maximum point is recorded as θ vis the pitch angle corresponding to the location of the vth local maximum point, is the horizontal angle corresponding to the location of the vth local maximum point, and the radiation power of these local maximum points is recorded as E(x v );

[0018] Step 3: At the vth local maximum point At Δθ=θ n+1 -θ n , is the interval, θ n+1 is the θ value corresponding to the n+1th row of the test grid, Δθ is the difference between the θ value corresponding to the n+1th row and the θ value corresponding to the nth row, The corresponding value for the m+1th row The value corresponding to the mth row The difference in value, The m+1th row of the test grid corresponds to Value, in The radiation power is measured again at q' positions in the orthogonal direction and the diagonal direction. The radiation power measured at each q position is recorded as

[0019] The radiation power E(x v ) and the value measured in step 3 Merge into a set E v , get the location of the maximum radiation power in the set θ ini is the pitch angle corresponding to the position of the maximum value in the set, is the horizontal angle corresponding to the position of the maximum value in the set;

[0020] Step 4: x ini As the starting position of the gradient search, the gradient g(x ini );

[0021] Step 5: Assume that it takes K position updates to determine the peak position. At k is the pitch angle of the kth test position, is the horizontal angle of the kth test position, according to the gradient g(x k ) Solve to obtain the search direction ρ from the kth test position to the k+1th test position required to update the position k and the search step length s k , solve for the k+1th test position for

[0022] x k+1 =xk +s k ρ k ;

[0023] Among them, θ k+1 is the pitch angle of the k+1th test position, is the horizontal angle of the k+1th test position;

[0024] Step 6: Set the maximum number of iterations Max_iter and repeat step 5 until any termination condition is met: (1) The number of iterations reaches Max_iter; (2) The result of three consecutive iterations is x k+1 =x k ;

[0025] After the iteration is terminated, the output converges to the beam peak position x peak And the corresponding radiation power E(x peak ).

[0026] As a further optimization scheme of the gradient-based beam peak search method described in the present invention, q' is 8.

[0027] As a further optimization scheme of the gradient-based beam peak search method described in the present invention, g(x k ) is calculated as follows:

[0028]

[0029] Where, Δg θ and are respectively the pitch angle θ and the horizontal angle in the spherical coordinate system The change in direction, for The radiation power at for The radiation power at for The magnitude of the radiated power at the location.

[0030] As a further optimization scheme of the gradient-based beam peak search method described in the present invention, ρ k The calculation formula is as follows:

[0031]

[0032] Among them, |·| represents g(x k ) model.

[0033] As a further optimization scheme of the gradient-based beam peak search method described in the present invention, s k Includes k,1 、s k,2and k,3 ,s k,1 、s k,2 and k,3 The calculation formula is as follows:

[0034] Based on the current k-th position x k And the search direction from the kth position to the k+1th position ρ k , according to the sufficient descent condition, set the initial step size s k,1 for

[0035]

[0036] Will s k,1 Substitute into the position update formula to get the k+1th position x calculated for the first time k+1,1 , compare x k+1,1 The radiation power E(x k+1,1 ) and x k The radiation power E(x k ), if E(x k+1,1 )>E(x k ), then increase ρ k step length in direction;

[0037] Let ξ(s k )=E(x k +s k ρ k ), E(x k +s k ρ k ) is x k +s k ρ k The radiated power at k ) is in x k Chu Chao k The radiation power in the direction with the step size as the independent variable, then the quadratic step size s k,2 for

[0038]

[0039] Where ξ(0) is s k When x is 0, k The radiation power at x is k The discrete derivative of the radiated power at k,1 ) is x k+1,1 The radiated power at x k+1,1 At k,2 Substitute into the position update formula to get the k+1th position x calculated for the second time k+1,2 ;

[0040] If E(x k+1,2 )>E(x k+1,1 ), E(x k+1,2 ) is x k+1,2 The radiated power at k+1,1 ) is x k+1,1 The radiation power at the point is calculated with a step size of s k,3 Increase ρ k The step length in the direction, let A and B represent complex polynomials, and simplify them to s k,3 The calculation formula is as follows:

[0041]

[0042] The calculation formulas for A and B are:

[0043]

[0044] Among them, ξ(s k,2 ) is x k +s k,2 ρ k The radiated power at .

[0045] As a further optimization scheme of the gradient-based beam peak search method of the present invention, at the kth position x k At , the search direction ρ required to update the position is calculated based on the gradient k After that, through N s Sub-radiation power measurement adjustment search step s k , the calculation formula is

[0046] s k =s k,1 +(N s -1)s k,2 +max(N s -2,0)s k,3 .

[0047] A gradient-based beam peak search device, comprising:

[0048] The device under test and the turntable, the device under test is fixed on the turntable, and the geometric center of the device under test is located at the center of the test area;

[0049] A test grid forming module is used to take the geometric center of the device under test as the origin of the spherical coordinate system, take the distance between the signal receiving device and the device under test as the radial distance of the spherical coordinate system, set the test interval in the elevation angle and azimuth angle dimensions of the spherical coordinate system according to the half-power bandwidth on the electric plane and the magnetic plane of the antenna in the device under test, and form a test grid with the azimuth angle and the elevation angle at a fixed radial distance as coordinates;

[0050] An estimation module is used to control the orientation of the device under test by using a turntable according to the angle position indicated by each grid point in the test grid, traverse the angles indicated by the test grid, measure the radiation power at the angle indicated by each grid point in the test grid, and then interpolate the test grid by an interpolation method to estimate the complete radiation pattern in the test area;

[0051] The secondary measurement module is used to obtain the positions of local maximum points from the estimated complete radiation pattern, and use the positions of these local maximum points as the center positions of the secondary measurement areas; around the center position of each secondary measurement area, multiple required secondary measurement positions are determined at a distance less than the grid spacing in the test grid, and the radiation power sets are measured at these secondary measurement positions;

[0052] The position search module is used to obtain the maximum radiation power from the set of radiation powers measured twice, take the position corresponding to the maximum radiation power in the secondary measurement as the initial position, update the test position each time according to the gradient search algorithm, and output the position and the radiation power measured at the position when the gradient search algorithm converges to a certain position without changing.

[0053] As a further optimization solution of the gradient-based beam peak search device described in the present invention, the turntable is rotated to an orientation consisting of a horizontal angle and a pitch angle with a resolution of 1° through software control.

[0054] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0055] (1) This scheme has no special requirements for the antenna to be tested, does not require any instruments other than conventional testing, and does not require any additional calibration steps. It can be directly tested with existing equipment, which is highly feasible.

[0056] (2) The amount of test data required by this solution is much smaller than that of the traversal solution, so it can effectively reduce the demand for human and material resources during the test process, and reduce power consumption and equipment load;

[0057] (3) This scheme can quickly locate the beam peak along the gradient path in the gain direction, significantly reducing the number of test points and avoiding the inefficient process of point-by-point testing in the traditional traversal scheme. The deviation of the search results relative to the traversal scheme is less than 0.5 dB, and the statistical average error relative to the true value is about 0.2 dB, which ensures the high precision and accuracy of the search results, indicating that this scheme has a very high application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a diagram of the test system device architecture of an implementation case of the present invention;

[0059] Figure 2is a flow chart of a peak search method in an implementation case of the present invention;

[0060] Figure 3 is a three-dimensional radiation pattern in an implementation case of the present invention;

[0061] Figure 4 is a radiation pattern estimated by interpolation from data measured by a test grid in an embodiment of the present invention;

[0062] Figure 5 It is a schematic diagram of searching for the beam peak position using a gradient algorithm in an implementation case of the present invention. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] A gradient-based beam peak search device comprises an antenna to be tested, a rotating platform and a probe antenna.

[0065] The antenna to be tested is used to radiate electromagnetic waves. In actual deployment scenarios, it can be any single-beam antenna such as a phased array or a flat-panel antenna.

[0066] The turntable is used to fix the antenna under test and realize 3D rotation with a resolution of 1° during the test, ensuring that the antenna's radiated power can be accurately measured in all directions.

[0067] The probe antenna is used as a signal receiving device to receive the electromagnetic waves radiated by the antenna to be tested and report the effective radiated power (EIRP) to the control device.

[0068] The device specifically includes:

[0069] The device under test and the turntable, the device under test is fixed on the turntable, and the geometric center of the device under test is located at the center of the test area;

[0070] A test grid forming module is used to take the geometric center of the device under test as the origin of the spherical coordinate system, take the distance between the signal receiving device and the device under test as the radial distance of the spherical coordinate system, set the test interval in the elevation angle and azimuth angle dimensions of the spherical coordinate system according to the half-power bandwidth on the electric plane and the magnetic plane of the antenna in the device under test, and form a test grid with the azimuth angle and the elevation angle at a fixed radial distance as coordinates;

[0071] An estimation module is used to control the orientation of the device under test by using a turntable according to the angle position indicated by each grid point in the test grid, traverse the angles indicated by the test grid, measure the radiation power at the angle indicated by each grid point in the test grid, and then interpolate the test grid by an interpolation method to estimate the complete radiation pattern in the test area;

[0072] The secondary measurement module is used to obtain the positions of local maximum points from the estimated complete radiation pattern, and use the positions of these local maximum points as the center positions of the secondary measurement areas; around the center position of each secondary measurement area, multiple required secondary measurement positions are determined at a distance less than the grid spacing in the test grid, and the radiation power sets are measured at these secondary measurement positions;

[0073] The position search module is used to obtain the maximum radiation power from the set of radiation powers measured twice, take the position corresponding to the maximum radiation power in the secondary measurement as the initial position, update the test position each time according to the gradient search algorithm, and output the position and the radiation power measured at the position when the gradient search algorithm converges to a certain position without changing.

[0074] Based on the above beam peak search device structure, the present invention also includes a beam peak search method:

[0075] First, the measurement grid size in the spherical coordinate system along the horizontal angle direction and the pitch angle direction is set according to the half-power bandwidth of the antenna to be tested, and the measured sparse grid data is interpolated to obtain the radiation pattern within the grid coverage range. Next, the vertex of the radiation pattern is retrieved and the vertex position is determined as the center of the area where the beam peak may appear. Secondary measurements are performed around the vertex to further narrow the area where the beam peak appears. Finally, the maximum value in the secondary measurement is used as the initial position, and the test position is updated using the gradient iteration of the current position. After the iterative result converges, the search program is terminated, and the peak position and the magnitude of the radiation power at the peak are output.

[0076] In one implementation example, the topology of the test device is as shown in the attached figure. Figure 1 As shown in the figure, it includes the antenna to be tested, the turntable and the probe antenna. Fix the antenna to be tested on the turntable and set the horizontal angle The test range of the test angle is [0°, 180°], and the test range of the pitch angle θ is [-90°, 90°]. During the test, the antenna to be tested is fixed at the center of the turntable and rotates 3D with the turntable at a resolution of 1°. The probe antenna is fixed and kept aligned with the center of the antenna to be tested. The radiation pattern of the antenna to be tested pointing to [0°, 0°] is shown in the attached figure. Figure 3 As shown. In the kth measurement, the direction of the probe antenna receiving the antenna to be measured is The signal power emitted at x is reported to the spectrum analyzer, power meter or other power measurement equipment. k The EIRP in the direction can be expressed as E(x k ).

[0077] For the half-power beamwidth of the electric plane, θ 3dB , the half-power beamwidth of the magnetic surface is The antenna under test is A test grid with dimensions of N×M covering the entire test area is formed for the grid cell size. The EIRP at each grid point is recorded, and the complete radiation pattern with a resolution of 1° in the test area is estimated based on the interpolation method. Figure 4 As shown, find the local maximum point As an alternative position for the beam peak. v The EIRP set E(x v ), find the position corresponding to the maximum EIRP in the set, and use this position as the starting point for the gradient search

[0078] x ini is the starting position, at the kth test position At x k Δg θ and To test in orthogonal directions, according to the gradient calculation formula

[0079]

[0080] Get x k The gradient at g(x k ) Estimate the next test position x k+1

[0081] x k+1 =x k +s k ρ k

[0082] where ρ k For x k+1 Relative x k Direction, s k To reach x k+1 The step length that needs to be crossed. After multiple iterations, the algorithm finally converges to the position x peak That is the beam peak position, the corresponding EIRP size E(x peak ) is the peak size, which enables the positioning and testing of the beam peak.

[0083] In order to further illustrate the present invention, a specific testing method is introduced below based on the above-mentioned testing device case.

[0084] like Figure 2 As shown, the specific steps are as follows:

[0085] 1. Traverse the θ plane [-90°, 90°], The half-power beam width θ of the antenna to be tested is respectively 3dB and The EIRP at each position on the test grid is calculated by taking into account the boundaries to form a matrix of dimension N×M. During the traversal process, the antenna under test keeps sending uplink signals and completes 3D rotation through the turntable. The probe antenna is connected to power measurement equipment such as spectrum analyzers and power meters to record the received signal power.

[0086] 2. Interpolate the test grid data to obtain the radiation power pattern matrix within the test range with a resolution of 1° and a size of 181×181. Get the positions x of all local maxima in it v ,v=1,2,...,V. At each local maximum x v There are 4 positions in the orthogonal direction, as well as the vectors of the two directions with larger EIRP measured in the 4 positions in the orthogonal direction and the diagonal direction pointing to a total of 5 positions x v,i ,v=1,2,...,V,i=1,2,...,5Test again, x v and x v,i Joint Second Test Position And the corresponding EIRP set

[0087] 3. Select The maximum value corresponding to As the starting position x of the gradient search algorithm ini , through K iterations, find the beam peak position x peak And the corresponding EIRP size E(x peak ).

[0088] 4. At the kth position x k When measuring at, the gradient at is calculated according to the gradient calculation formula

[0089]

[0090] Where, Δg θ and They are respectively along the θ direction and The desired increment in direction.

[0091] 5. According to the gradient information at, based on the position update formula, the k+1th test position x can be obtained k+1 for

[0092] x k+1 =x k +s k ρ k

[0093] where ρ k For x k+1 Relative x k Direction, s k To reach x k+1 The step size to be crossed. k and k are calculated based on the gradient at the position, ρ k is the normalized gradient vector, that is

[0094]

[0095] Among them, |·| represents g(x k ) model.

[0096] In order to fully search in the same direction with the largest possible step length, according to the sufficient rise criterion, the step length is gradually increased by the primary, secondary and tertiary calculation formulas under the condition of ensuring the increase of radiation power. The primary calculation formula of the step length is:

[0097]

[0098] Will s k,1 Substitute into the position update formula to calculate x k+1,1 , compare E(x k+1,1 ) and E(x k ), if E(x k+1,1 )>E(x k ), then increase the step size in that direction and set the second-order step size s k,2 for

[0099]

[0100] Among them, ξ(s k )=E(x k +s k ρ k ). k+1,1 At k,2 Substitute into the position update formula to calculate x k+1,2 . Similarly, if E(x k+1,2 )>E(x k+1,1 ), then with a third-order step size s k,3 Increase the step size in this direction, the calculation formula is as follows:

[0101]

[0102] The calculation formulas for A and B are as follows:

[0103]

[0104] As a further optimization scheme of the gradient-based beam peak search method of the present invention, at the kth position x k At, the direction ρ of the next test position is calculated according to the gradient k After that, through N s The total step size s that can be spanned by the sub-radiated power measurement adjustment k for

[0105] s k =s k,1 +(N s -1)s k,2 +max(N s -2,0)s k,3 .

[0106] 6. As attached Figure 5 As shown in the figure, after multiple iterations, until any termination condition is reached: (1) the number of iterations reaches the set maximum number; (2) the result of three consecutive iterations is x k =x k+1 After the iteration is terminated, the output algorithm converges to the beam peak position x peak .

[0107] The above description is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a technician familiar with the technical field within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A gradient-based beam peak search method, characterized in that: include: First, fix the device under test on a turntable, and make the geometric center of the device under test located at the center of the test area. Take the geometric center of the device under test as the origin of the spherical coordinate system, take the distance between the signal receiving device and the device under test as the radial distance of the spherical coordinate system, set the test interval in the elevation angle and azimuth angle dimensions of the spherical coordinate system according to the half-power bandwidth on the electric and magnetic planes of the antenna in the device under test, and form a test grid with the azimuth angle and elevation angle at a fixed radial distance as coordinates; According to the angle position indicated by each grid point in the test grid, the turntable is used to control the direction of the device under test, and the angle indicated by the test grid is traversed. After the radiation power at the angle indicated by each grid point in the test grid is measured, the test grid is interpolated by the interpolation method to estimate the complete radiation pattern in the test area; Next, the positions of the local maximum points are obtained from the estimated complete radiation pattern, and the positions of these local maximum points are used as the center positions of the secondary measurement areas; around the center position of each secondary measurement area, the positions of the required multiple secondary measurements are determined at a distance less than the grid spacing in the test grid, and the radiation power sets are measured at these secondary measurement positions; Finally, the maximum radiation power is obtained from the set of radiation powers measured twice. The position corresponding to the maximum radiation power in the secondary measurement is used as the initial position. According to the gradient search algorithm, the test position is updated each time. When the gradient search algorithm converges to a certain position and does not change, the position and the radiation power measured at the position are output.

2. A gradient-based beam peak search method according to claim 1, characterized in that: Around the center position of each secondary measurement area, including in the orthogonal direction and the diagonal direction of the center position of each secondary measurement area.

3. A gradient-based beam peak search method according to claim 1, characterized in that: The specific steps are as follows: Step 1: In the spherical coordinate system with the geometric center of the device under test as the origin The range of the pitch angle θ is -90°≤θ≤90°, and the azimuth angle The range is r is the radial distance; within the range of elevation angle variation, the half-power beam width θ of the antenna in the device under test on the θ plane 3dB N angles are selected as intervals, and the antenna to be tested is within the horizontal angle variation range. Half-power beamwidth on the plane Select M angles at intervals to form a test grid with dimensions of N×M; change the orientation of the device under test by using the turntable n=1...N,m=1...M,θ n is the selected nth pitch angle, For the selected mth horizontal angle, use the signal receiving device to The radiated power measured at to receive; Step 2: Use the interpolation method to interpolate the test grid to obtain a complete radiation pattern after interpolation. There are V local maximum points in the radiation pattern. These local maximum points are the center points of the secondary measurement area. The position of the vth local maximum point is recorded as v=1...V,θ v is the pitch angle corresponding to the location of the vth local maximum point, is the horizontal angle corresponding to the location of the vth local maximum point, and the radiation power of these local maximum points is recorded as E(x v ); Step 3: At the vth local maximum point At Δθ=θ n+1 -θ n , is the interval, θ n+1 is the θ value corresponding to the n+1th row of the test grid, Δθ is the difference between the θ value corresponding to the n+1th row and the θ value corresponding to the nth row, The corresponding value for the m+1th row The value corresponding to the mth row The difference in value, The m+1th row of the test grid corresponds to Value, in The radiation power is measured again at q' positions in the orthogonal direction and the diagonal direction. The radiation power measured at each q position is recorded as q=1,2,...,q′; The radiation power E(x v ) and the value measured in step 3 Merge into a set E v , get the location of the maximum radiation power in the set θ ini is the pitch angle corresponding to the position of the maximum value in the set, is the horizontal angle corresponding to the position of the maximum value in the set; Step 4: x ini As the starting position of the gradient search, the gradient g(x ini ); Step 5: Assume that it takes K position updates to determine the peak position. At k is the pitch angle of the kth test position, is the horizontal angle of the kth test position, according to the gradient g(x k ) Solve to obtain the search direction ρ from the kth test position to the k+1th test position required to update the position k and the search step length s k , solve for the k+1th test position for x k+1 =x k +s k r k ; Among them, θ k+1 is the pitch angle of the k+1th test position, is the horizontal angle of the k+1th test position; Step 6: Set the maximum number of iterations Max_iter and repeat step 5 until any termination condition is met: (1) The number of iterations reaches Max_iter; (2) The result of three consecutive iterations is x k+1 =x k ; After the iteration is terminated, the output converges to the beam peak position x peak And the corresponding radiation power E(x peak ).

4. The gradient-based beam peak search method according to claim 1, characterized in that: q' is 8.

5. The gradient-based beam peak search method according to claim 3, characterized in that: g(x k ) is calculated as follows: Where, Δg θ and are respectively the pitch angle θ and the horizontal angle in the spherical coordinate system The change in direction, for The radiation power at for The radiation power at for The magnitude of the radiated power at .

6. A gradient-based beam peak search method according to claim 3, characterized in that: ρ k The calculation formula is as follows: Among them, |·| represents g(x k ) model.

7. The gradient-based beam peak search method according to claim 3, characterized in that: s k Includes k,1 、s k,2 and k,3 ,s k,1 、s k,2 and k,3 The calculation formula is as follows: Based on the current k-th position x k And the search direction from the kth position to the k+1th position ρ k , according to the sufficient descent condition, set the initial step size s k,1 for Will s k,1 Substitute into the position update formula to get the k+1th position x calculated for the first time k+1,1 , compare x k+1,1 The radiation power E(x k+1,1 ) and x k The radiation power E(x k ), if E(x k+1,1 )>E(x k ), then increase ρ k step length in direction; Let ξ(s k )=E(x k +s k ρ k ), E(x k +s k ρ k ) is x k +s k ρ k The radiated power at k ) is in x k Chu Chao k The radiation power in the direction with the step size as the independent variable, then the quadratic step size s k,2 for Where ξ(0) is s k When x is 0, k The radiation power at the point where ξ′(0) is x k The discrete derivative of the radiated power at k,1 ) is x k+1,1 The radiation power at x k+1,1 At k,2 Substitute into the position update formula to get the k+1th position x calculated for the second time k+1,2 ; If E(x k+1,2 )>E(x k+1,1 ), E(x k+1,2 ) is x k+1,2 The radiated power at k+1,1 ) is x k+1,1 The radiation power at the point is calculated with a step size of s k,3 Increase ρ k The step length in the direction, let A and B represent complex polynomials, and simplify them to s k,3 The calculation formula is as follows: The calculation formulas for A and B are: Among them, ξ(s k,2 ) is x k +s k,2 ρ k The radiated power at .

8. A gradient-based beam peak search method according to claim 7, characterized in that: At the kth position x k At , the search direction ρ required to update the position is calculated based on the gradient k After that, through N s Sub-radiation power measurement adjustment search step s k , the calculation formula is s k =s k,1 +(N s -1)s k,2 +max(N s -2,0)s k,3 .

9. A gradient-based beam peak search device, characterized in that: include: The device under test and the turntable, the device under test is fixed on the turntable, and the geometric center of the device under test is located at the center of the test area; A test grid forming module is used to take the geometric center of the device under test as the origin of the spherical coordinate system, take the distance between the signal receiving device and the device under test as the radial distance of the spherical coordinate system, set the test interval in the elevation angle and azimuth angle dimensions of the spherical coordinate system according to the half-power bandwidth on the electric plane and the magnetic plane of the antenna in the device under test, and form a test grid with the azimuth angle and the elevation angle at a fixed radial distance as coordinates; An estimation module is used to control the orientation of the device under test by using a turntable according to the angle position indicated by each grid point in the test grid, traverse the angles indicated by the test grid, measure the radiation power at the angle indicated by each grid point in the test grid, and then interpolate the test grid by an interpolation method to estimate the complete radiation pattern in the test area; The secondary measurement module is used to obtain the positions of local maximum points from the estimated complete radiation pattern, and use the positions of these local maximum points as the center positions of the secondary measurement areas; around the center position of each secondary measurement area, multiple required secondary measurement positions are determined at a distance less than the grid spacing in the test grid, and the radiation power sets are measured at these secondary measurement positions; The position search module is used to obtain the maximum radiation power from the set of radiation powers measured twice, take the position corresponding to the maximum radiation power in the secondary measurement as the initial position, update the test position each time according to the gradient search algorithm, and output the position and the radiation power measured at the position when the gradient search algorithm converges to a certain position without changing.

10. The gradient-based beam peak search device according to claim 9, characterized in that: The turntable is controlled by software to rotate to an orientation consisting of horizontal and elevation angles with a resolution of 1°.

Images