A high-precision antenna plane near-field measurement method based on data assimilation

Abstract

This invention belongs to the field of planar near-field testing technology, and relates to a high-precision planar near-field measurement method for antennas based on data assimilation. The method includes the following steps: acquiring simulated data of the antenna under test (ATT) and measured data of the ATTT obtained from a planar near-field testing system; inputting the simulated and measured data of the ATTT into a data assimilation algorithm, fusing them using the algorithm to obtain an estimated omnidirectional radiation pattern, thus obtaining a high-precision antenna radiation pattern. This invention's measurement method, through data assimilation theory, rationally combines simulated data with actual test data, reducing the truncation error caused by finite-plane testing in traditional methods, especially for test data outside the confidence interval; under consistent sampling settings, the method described in this invention achieves higher testing accuracy than traditional planar near-field testing methods.

Description

Technical Field

[0001] This invention belongs to the field of planar near-field testing technology and relates to a high-precision planar near-field measurement method for antennas based on data assimilation. Background Technology

[0002] Near-field antenna testing technology is widely used in antenna pattern testing due to its small testing space. With the rapid development of wireless devices, manufacturers have increasingly higher demands for antenna testing accuracy, and the need for efficient testing is also becoming more urgent. Among near-field testing techniques, planar near-field testing is one of the most prevalent. This method is performed in an anechoic chamber, sampling on a plane within the near-field region of the antenna under test, and then using a near-field to far-field conversion algorithm to obtain the test results for the antenna in the far-field region.

[0003] An anechoic chamber uses absorbing materials to eliminate electromagnetic wave reflections, simulating a free-space testing environment, and a shielded chamber to eliminate interference from external electromagnetic signals. Planar near-field testing systems have become one of the most common and universal antenna testing methods. For example... Figure 1 The conventional planar near-field testing system shown consists of four main parts: absorbing material, a robotic arm, a probe, and a platform for placing the antenna under test. The absorbing material is installed on the walls of the anechoic chamber and the surface of the robotic arm, or other locations that may generate irrelevant electromagnetic reflections, creating a free-space testing environment. The robotic arm is mounted next to the testing platform, and the test probe is attached to the robotic arm, which controls the probe to reach different sampling positions. The testing platform is used to place the antenna under test.

[0004] During testing, the antenna under test (DUT) is placed on the test platform and connected to a vector network analyzer along with the probe. Generally, to ensure accurate test results, the main radiation direction of the DUT is aligned with the robotic arm. The robotic arm is then controlled by a program to sample data on a pre-designed sampling plane. After sampling at all preset locations, the sampled near-field data is converted into the required far-field data, such as a radiation pattern, using the test system's built-in near-far-field conversion algorithm.

[0005] The principle of planar near-field testing is based on the plane wave expansion of the electromagnetic field. That is, in a passive time-harmonic field, the electromagnetic field can be expressed as an integral of a plane wave. By sampling on an infinitely large plane within the near-field distance, the tangential component of the electromagnetic field is obtained. Then, based on the sampled values, the electromagnetic field data on that plane can be obtained. Based on the inversion of the plane wave expansion, the complex amplitude vector of the electromagnetic field can be obtained, and the field at any point can be calculated from the relevant data. For example... Figure 2As shown, in actual testing, it is obviously impossible to achieve an infinitely large sampling plane; sampling only a finite plane introduces truncation error. Theoretically, the data error outside the truncated area between the sampling plane and the antenna aperture is large, and such data is generally considered unreliable. However, the data error within the truncated area is smaller and meets the testing requirements for testers primarily concerned with the antenna's main lobe radiation performance; this truncated area is usually referred to as the confidence interval.

[0006] Due to limitations in robotic arm reach and testing space, the confidence interval of data obtained from planar near-field testing is typically not large. This means that planar near-field testing methods can only provide relatively accurate radiation information for a portion of the antenna under test. To obtain accurate information within the elevation coordinate range [-90°, 90°], testers will choose other testing methods. Employing other testing methods, such as multi-probe anechoic chambers, requires building new testing systems, which incurs additional construction costs. Therefore, wireless equipment manufacturers have a pressing need to reduce the truncation error of planar near-field testing systems under the same testing conditions without increasing costs. Summary of the Invention

[0007] The purpose of this invention is to provide a high-precision antenna planar near-field measurement method based on data assimilation, which solves the problems of high testing cost and large truncation error in existing methods.

[0008] This invention is achieved through the following technical solution:

[0009] This invention discloses a high-precision antenna pattern testing method based on data assimilation, comprising the following steps:

[0010] Acquire simulation data of the antenna under test and measured data of the antenna under test obtained from testing based on a planar near-field testing system;

[0011] The simulated data and measured data of the antenna under test are input into the data assimilation algorithm, and then fused using the data assimilation algorithm to obtain the estimated omnidirectional radiation pattern, thus obtaining the high-precision antenna radiation pattern.

[0012] Furthermore, the process of obtaining the simulation data of the antenna under test is as follows:

[0013] The antenna under test is modeled, and the far-field radiation pattern data of the corresponding simplified model is obtained in the simulation software, which is used as the simulation data of the antenna under test.

[0014] Furthermore, the testing process for the measured data of the antenna under test is as follows:

[0015] Position the radiating aperture of the antenna under test at the calibration point of the planar near-field test system, and align the main radiation direction with the robotic arm; then connect the antenna under test and the probe mounted on the robotic arm to the vector network analyzer.

[0016] According to the experimental requirements, set the test frequency band, sampling interval, side length of the antenna aperture, distance between the sampling plane and the antenna aperture, and side length of the sampling plane, and generate all preset sampling point information;

[0017] The controller controls the robotic arm to sequentially collect and record the electric fields in the two polarization directions at the preset sampling points, and store them in the data memory.

[0018] After the robotic arm controls the probe to complete sampling at all sampling points, it uploads the obtained near-field data to the control terminal. The near-field conversion algorithm built into the control terminal then converts the near-field data of the antenna under test into far-field data, thus obtaining the measured data of the antenna under test.

[0019] Furthermore, if the test requirement is to obtain the omnidirectional radiation information of the entire antenna, the antenna under test is rotated 180° horizontally during the planar near-field test, and the test process of the measured data of the antenna under test is repeated again to obtain the data in the elevation coordinate range [-180°, 180°] again. The omnidirectional radiation pattern is estimated again using the data assimilation method.

[0020] By stitching together the first and second omnidirectional radiation patterns, a high-precision omnidirectional radiation pattern is obtained.

[0021] Furthermore, if the test requirement is to obtain the omnidirectional radiation information of the entire antenna, the antenna under test is rotated horizontally by 180° during the planar near-field test, and the test process of the measured data of the antenna under test is repeated again to obtain the data in the elevation coordinate range [-180°, 180°] again, which is the second measured data of the antenna under test.

[0022] After stitching together the measured data of the antenna under test obtained from the two tests, the omnidirectional radiation pattern is estimated using the data assimilation method to obtain a high-precision omnidirectional radiation pattern.

[0023] Furthermore, the fusion process of the data assimilation algorithm specifically includes the following steps:

[0024] 2.1 Based on the trigonometric series model, using the simulation data of the antenna under test, a set of coefficients of the trigonometric series model are first solved using the orthogonality of trigonometric functions as prior parameters p. prior ;

[0025] 2.2. A set of antenna pattern data obtained from actual testing is called the observed value, denoted as z; and the confidence interval is calculated based on the planar near-field test configuration parameters of the planar near-field test system.

[0026] 2.3. Determine the model prediction matrix H based on the sampling point distribution of the measured data of the antenna under test;

[0027] The observation weight matrix R is determined based on the sampling point distribution and confidence interval of the actual planar near-field test data;

[0028] A priori information weight matrix W is established based on the number of sampling points in the simulation data of the antenna under test.

[0029] 2.4. Based on the data assimilation calculation formula, and using the prior parameter p... prior The updated estimated state value is calculated from the model prediction matrix H, the observation weight matrix R, and the prior information weight matrix W, thus realizing the fusion of the observation value z and the prior information.

[0030] 2.5 Using the estimated state values ​​as parameters of the trigonometric series model, the antenna pattern at the required location is calculated.

[0031] Furthermore, 2.1 specifically refers to:

[0032] The trigonometric series model represents the relationship between the elevation angle coordinate θ and the gain g(θ):

[0033]

[0034] a0 represents the constant term, a n The coefficients of the nth-order cosine function, b n The coefficients represent the nth-order sine function; given 2N+1 samples, the orthogonality of trigonometric functions can be used to obtain 2N+1 coefficients {a0, a1, ..., a...}. N ,b1,...,b N The values ​​of these coefficients are taken as prior parameters, and the column vector of prior parameters is shown in the following form:

[0035]

[0036] Furthermore, in section 2.2, the confidence interval is θ. max The calculation formula is:

[0037]

[0038] Where L is the length of the sampling plane, D is the length of the antenna aperture, and z is the distance between the sampling plane and the antenna aperture.

[0039] Furthermore, in 2.3,

[0040]

[0041]

[0042]

[0043] in, The coordinates are θj The weights corresponding to the sampling points, j = 1, 2, ..., N test ;

[0044] The prior information weight matrix W is: W = diag(1,...,1).

[0045] Furthermore, in section 2.4, the updated state estimate... The formula for calculation is:

[0046]

[0047] The estimated state value is in the following form:

[0048]

[0049] 2.5 Specifically: Using the estimated state values ​​as parameters of the trigonometric series model, and substituting them into the relationship between the elevation coordinate θ and the gain g(θ), the antenna gain at different positions is calculated, and the expression is updated as follows:

[0050]

[0051] The antenna pattern at the desired location is obtained by calculating the antenna gain at different locations.

[0052] Compared with the prior art, the present invention has the following beneficial technical effects:

[0053] This invention discloses a high-precision planar near-field measurement method for antennas based on data assimilation. By using data assimilation theory, it reasonably combines simulation data with actual test data, reducing the truncation error caused by finite plane testing in traditional methods, especially for test data outside the confidence interval. Under consistent sampling settings, the method of this invention can achieve higher test accuracy than traditional planar near-field testing methods.

[0054] The measurement method of this invention utilizes the simulation data of the antenna under test, which solves the problem that the overall accuracy of the simulation data is not high and cannot be directly applied to actual testing. The corresponding data is simple and easy to obtain, and no additional testing system needs to be built. The measurement method of this invention can be applied to any conventional planar near-field testing system, which has strong applicability and low construction cost. The measurement method of this invention adopts program control, and the testing process is simple and efficient, requiring no manual intervention, which greatly improves the efficiency of air interface testing. Attached Figure Description

[0055] Figure 1 It uses a planar near-field testing system;

[0056] Figure 2 This is a schematic diagram illustrating the confidence interval calculation method used in an embodiment of the present invention;

[0057] Figure 3 This is a schematic diagram of the dimensions of the antenna under test used in an embodiment of the present invention;

[0058] Figure 4 The comparison between the test error calculated by the improved method used in the embodiments of the present invention and the error of the standard planar near-field test results and the simulation data error is shown in (a) [-90°, 90°] and (b) [-90°, -38°].

[0059] Figure 5 The comparison of the antenna pattern estimation results using reference values, simulation data, planar near-field test data, and improved methods in the embodiments of the present invention is shown in (a) E-plane and (b) H-plane. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention; that is, the described embodiments are only a part of the embodiments of the present invention, and not all of them.

[0061] The components described and illustrated in the accompanying drawings and embodiments of this invention can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the invention provided in the following drawings is not intended to limit the scope of the claimed invention, but merely to illustrate one selected embodiment of the invention. All other embodiments obtained by those skilled in the art based on the accompanying drawings and embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0062] This invention discloses a high-precision antenna pattern testing method based on data assimilation, comprising the following steps:

[0063] Acquire simulation data of the antenna under test and measured data of the antenna under test obtained from testing based on a planar near-field testing system;

[0064] The simulated data and measured data of the antenna under test are input into the data assimilation algorithm, and then fused using the data assimilation algorithm to obtain the estimated omnidirectional radiation pattern, thus obtaining the high-precision antenna radiation pattern.

[0065] The process of obtaining the simulation data of the antenna under test is as follows: model the antenna under test, obtain the far-field radiation pattern data of the corresponding simplified model in the simulation software, and use it as the simulation data of the antenna under test.

[0066] The specific testing process for the measured data of the antenna under test is as follows:

[0067] Position the radiating aperture of the antenna under test at the calibration point of the planar near-field test system, and align the main radiation direction with the robotic arm; then connect the antenna under test and the probe mounted on the robotic arm to the vector network analyzer.

[0068] According to the experimental requirements, set the test frequency band, sampling interval, side length of the antenna aperture, distance between the sampling plane and the antenna aperture, and side length of the sampling plane, and generate all preset sampling point information;

[0069] The controller controls the robotic arm to sequentially collect and record the electric fields in the two polarization directions at the preset sampling points, and store them in the data memory.

[0070] After the robotic arm controls the probe to complete sampling at all sampling points, it uploads the obtained near-field data to the control terminal. The near-field conversion algorithm built into the control terminal then converts the near-field data of the antenna under test into far-field data, thus obtaining the measured data of the antenna under test.

[0071] The fusion process of data assimilation algorithms specifically includes the following steps:

[0072] 2.1 Based on the trigonometric series model, using the simulation data of the antenna under test, a set of coefficients of the trigonometric series model are first solved using the orthogonality of trigonometric functions as prior parameters p. prior ;

[0073] 2.2. A set of antenna pattern data is obtained based on the actual test, which is the observed value, denoted as z; and the confidence interval is calculated based on the planar near-field test configuration parameters of the planar near-field test system.

[0074] 2.3. Determine the model prediction matrix H based on the sampling point distribution of the measured data of the antenna under test;

[0075] The observation weight matrix R is determined based on the sampling point distribution and confidence interval of the actual planar near-field test data;

[0076] A priori information weight matrix W is established based on the number of sampling points in the simulation data of the antenna under test.

[0077] 2.4. Based on the data assimilation calculation formula, the updated estimated state value is calculated based on the prior parameters, model prediction matrix H, observation weight matrix R, and prior information weight matrix W, thereby realizing the fusion of observation values ​​and prior information.

[0078] 2.5 Using the estimated state values ​​as parameters of the trigonometric series model, the antenna pattern at the required location is calculated.

[0079] Specifically, the confidence interval is calculated based on the planar near-field test configuration parameters of the planar near-field test system. The theoretical derivation and principle are introduced below:

[0080] 1) Planar near-field testing

[0081] In a linear, homogeneous, and isotropic source-free region, the time-harmonic field Satisfies the Helmholtz equation:

[0082]

[0083] in, Let ω be the electromagnetic constant of the propagation medium, and μ and ε be the electromagnetic constants of the medium. In a rectangular coordinate system, equation (1) has a general solution:

[0084]

[0085] Among them, the complex amplitude vector Sum of wave vectors They are respectively:

[0086]

[0087] Since the field divergence in the source-free region is 0, according to formulas (2) and (3), we have:

[0088]

[0089] If we assume and If they are independent variables, then

[0090]

[0091] For all regions of k x k y Integrating, we can obtain the source-free time-harmonic field. The general solution is the plane wave expansion:

[0092]

[0093] Assume the tangential field components on the z = d plane are known. Its expression is:

[0094]

[0095] Furthermore, according to formula (6), we have:

[0096]

[0097] in, Inverting formula (8) yields:

[0098]

[0099] In the standard planar near-field testing method, the tangential field components on a finite plane are obtained through sampling. You can calculate using formulas 3, 5, and 9. The antenna gain at the required far-field location can be obtained using Equation 2. Typically, the sampling plane is a square, as shown in the diagram below. Figure 2 As shown, L is the length of the sampling plane, D is the length of the antenna aperture, and z is the distance between the sampling plane and the antenna aperture. The finite sampling plane leads to truncation error. Test results outside the truncated area have larger errors, while test results within the truncated area are more accurate. Therefore, we call the reliable test data range the confidence interval θ. max The calculation formula is as follows:

[0100]

[0101] The improved method proposed in this invention is based on planar near-field test data and simulation data within confidence intervals. Test results within the confidence interval of planar near-field tests are relatively reliable, while test results outside the confidence interval have larger errors. Simulation data does not have truncation errors, but the simplification of the simulation model leads to low overall accuracy, making it unsuitable as a direct substitute for actual test results. Therefore, this invention uses a data assimilation method to obtain more accurate test results by combining simulation data with actual planar near-field test data.

[0102] The data assimilation problem addresses how to combine observations and the model to obtain the optimal combination when neither the model nor the observations accurately describe the system state. Data assimilation employs various mathematical methods to reliably fuse data from different sources. In other words, it continuously updates the model's input parameters based on data from other sources, ensuring that the model's predictions remain at a high level of accuracy. The essence of this problem is minimizing the mismatch between model predictions and observations, as well as the mismatch between the model and prior information, and providing a new, combined prediction. This is also the foundation for applying data assimilation to the field of antenna testing.

[0103] The specific calculation method for data assimilation is as follows:

[0104] First, the antenna pattern is represented using a finite series of trigonometric functions, i.e., the elevation coordinate θ and the gain g(θ) have the following relationship:

[0105]

[0106] Clearly, with just 2N+1 samples, the orthogonality of trigonometric functions can be used to obtain 2N+1 coefficients {a0, a1, ..., a...}. N ,b1,...,b N The value of}. The corresponding coefficients are calculated using simulation data as prior parameters:

[0107]

[0108] Then, a set of observations was obtained based on the planar near-field test:

[0109]

[0110] The method proposed in this invention does not have any special requirements on the number of observations or their coordinate distribution.

[0111] The prediction model for this method is:

[0112] g = Hp (14)

[0113] Where H is the prediction matrix and p is the state value.

[0114] The prediction matrix H and the state value p are as follows:

[0115]

[0116] in, This represents the sampling point number, corresponding to the measured data in Formula 13.

[0117]

[0118] Multiplying the first row of the H matrix by the state value p yields the predicted sampled value at the first sampling point location in the observations. Multiplying the second row of the H matrix by the state value p yields the predicted sampled value at the second sampling point location in the observations. This process continues, forming the prediction model based on Equation 11. This model can be directly solved using the observed value z to obtain a set of state values ​​p, but using these state values ​​to predict results at other locations is not accurate enough, thus requiring further processing.

[0119] Let the cost function be:

[0120] Θ = (Hp - z) T R -1 (Hp-z)+(pp prior ) T W -1 (pp prior (17)

[0121] Formula (17) is used to explain the physical meaning of matrices R and W. When this function value reaches its minimum, the corresponding state value p is what this invention aims for (i.e., as stated later). ).

[0122] Wherein, the prior information weight matrix W is:

[0123] W = diag(1,...,1) (18)

[0124] The observation weight coefficient matrix R is:

[0125]

[0126] For θ at different positions j have:

[0127]

[0128] Where, θ sampling The range of acceptable planar near-field test data is typically represented by θ. sampling =θ max C R To represent an arbitrary constant, it is generally much greater than 1, for example, 10. 4 Based on the actual test data, the updated estimated state values ​​are obtained. The formula for calculation is:

[0129]

[0130] The estimated state value is in the following form:

[0131]

[0132] Based on the estimated parameters, the antenna gain at different locations can be calculated:

[0133]

[0134] An antenna radiation pattern is drawn based on the antenna gain at different locations. The horizontal axis represents the elevation angle θ, and the vertical axis represents the antenna gain g(θ) at θ. The resulting graph is the antenna radiation pattern.

[0135] The present invention will be further described in detail below with reference to the embodiments.

[0136] The test frequency in this embodiment of the invention is 2.9 GHz, and a horn antenna is used as the antenna under test. To verify the effectiveness of the improved planar near-field testing method, this embodiment of the invention tests the antenna under test in both a multi-probe anechoic chamber and a planar near-field test:

[0137] 1. Based on the dimensions of the antenna under test, a model was built in the simulation software FEKO, and the antenna radiation pattern of the horn antenna at a frequency of 2.9 GHz was obtained through simulation. Figure 5 The sampling interval for the exported data is set to 1°.

[0138] 2. The device under test (DUT) is placed normally in a multi-probe anechoic chamber. According to the standard multi-probe anechoic chamber testing method, reference test data with an elevation angle coordinate θ∈[-90°, 90°] and an azimuth angle coordinate φ∈[0°, 360°] is obtained with a sampling interval of 1°. The multi-probe anechoic chamber testing system mainly consists of a vertical ring and a turntable. During the test, the antenna under test is placed on the turntable, and then the turntable is controlled to rotate at equal intervals. Combined with the probes on the vertical ring, sampling of the entire test sphere is completed. Theoretically, the test sphere in the multi-probe anechoic chamber completely encloses the antenna under test; however, in practice, due to the need for a support structure, no probe is installed at the bottom of the vertical ring. Nevertheless, the test sphere still surrounds all upper hemisphere radiation, so the test data within the elevation angle coordinate range of -90° to 90° is still relatively accurate and can be used as a reference value. The test frequency is 2.9GHz, and the results are as follows: Figure 5 The antenna pattern shown has a sampling interval of 1°.

[0139] 3. In a planar near-field anechoic chamber system, a set of tests were conducted using the same antenna under test. The experimental parameters were set as follows: the sampling plane was a square with sides of 1m, the sampling interval was 45mm, and the aperture dimensions of the antenna under test were as follows. Figure 3 As shown, the sampling plane is 100mm away from the antenna aperture. The confidence interval is calculated to be approximately 74° according to formula (10). The test frequency is set to 2.9GHz, and the antenna pattern with a sampling interval of 1° is obtained after testing.

[0140] 4. To facilitate calculation and explanation, and considering the case of a multi-probe anechoic chamber with a missing probe facing away from the target, we only use data within the elevation angle range of [-90°, 90°] for processing and comparison. The improved method calculates prior information from simulation data, and then compares it with the planar near-field test data (assuming θ) used as the observed values. sampling =θ max C R =10 4 Substitute these parameters into formula (21) to obtain the estimated parameters. Then, use the estimated parameters to calculate the gain of the antenna at the same location according to formula (23).

[0141] Finally, using the data obtained from the multi-probe anechoic chamber test as a reference value, the errors between the planar near-field test data, simulation data, and data obtained from the improved method based on data assimilation and the reference value were calculated respectively. The errors were calculated at different azimuth angles, and the average error at each elevation angle was obtained. The results were then plotted. Figure 4 Then, the radiation pattern test results for the E-plane and H-plane were plotted on... Figure 5 middle.

[0142] Figure 5 and Figure 4 Their purposes are the same; both aim to demonstrate the superiority of the improved method. The difference lies in... Figure 4 It represents the average error, indicating the overall performance of the method. And... Figure 5 Two sets of data were selected and plotted to specifically demonstrate the direction graph output by the improved method.

[0143] Figure 4 Figure a shows the error results in the range of [-90 degrees to -90 degrees], reflecting the overall improvement effect and demonstrating that the new measurement method of this invention has the smallest error. Figure 4 b is a magnification of figure a, which is the main improvement of this method for a clearer display. Figure 4 The improvement in accuracy of the improved method is mainly concentrated outside the confidence interval, while the impact within the confidence interval is minimal because the differences between the data groups are not significant. Figure 4 b then magnifies Figure 4 The situation outside the confidence interval in a makes the performance of the improved method clearer.

[0144] To quantitatively demonstrate the effectiveness of the improved method, from Figure 4 In section a, near-field test errors, simulation errors, and improved method errors were selected at 15° intervals under different coordinates, and the data are recorded in Table 1. According to... Figure 5 a and Figure 5 b. On the E-plane and H-plane of the antenna under test, at 15° intervals, the normalized radiation pattern data under different coordinates were estimated using planar near-field test methods, simulation methods, and improved methods. These data were recorded in Tables 2 and 3.

[0145] Table 1. Estimation Errors of the Three Methods

[0146]

[0147]

[0148] Table 2 shows the antenna's normalized gain in the E-plane measured using three methods.

[0149]

[0150]

[0151] Table 3 shows the antenna's normalized gain in the H-plane measured using three methods.

[0152]

[0153]

[0154] More preferably, if the test requires obtaining the omnidirectional radiation information of the entire antenna, there are two processing methods:

[0155] The first method involves rotating the antenna under test horizontally by 180° during the near-field planar test and repeating the test process for the measured data of the antenna under test. This yields data within the elevation coordinate range of [-180°, 180°], and the omnidirectional radiation pattern is estimated again using the data assimilation method. The omnidirectional radiation pattern obtained in the first and second tests is then stitched together to obtain a high-precision omnidirectional radiation pattern.

[0156] The second method involves rotating the antenna under test horizontally by 180° during the planar near-field test and repeating the test process to obtain the elevation coordinates within the range of [-180°, 180°]. This is the second set of measured data for the antenna under test. After stitching the two sets of measured data together, the omnidirectional radiation pattern is estimated using the data assimilation method to obtain a high-precision omnidirectional radiation pattern.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A high-precision antenna pattern testing method based on data assimilation, characterized in that, The process includes the following: Acquire simulation data of the antenna under test and measured data of the antenna under test obtained from testing based on a planar near-field testing system; The simulated data and measured data of the antenna under test are input into the data assimilation algorithm, and then fused using the data assimilation algorithm to obtain the estimated omnidirectional radiation pattern, thus obtaining the high-precision antenna radiation pattern. The data assimilation algorithm's fusion process specifically includes the following steps: Step 2.1: Based on the trigonometric series model, using the simulation data of the antenna under test, a set of coefficients of the trigonometric series model are first solved using the orthogonality of trigonometric functions as prior parameters p. prior ; Step 2.2: Obtain a set of antenna pattern data based on the actual test, which is the observed value, denoted as z; and calculate the confidence interval based on the planar near-field test configuration parameters of the planar near-field test system; Step 2.3: Determine the model prediction matrix H based on the sampling point distribution of the measured data of the antenna under test; The observation weight matrix R is determined based on the sampling point distribution and confidence interval of the actual planar near-field test data; A priori information weight matrix W is established based on the number of sampling points in the simulation data of the antenna under test. Step 2.4: Based on the data assimilation calculation formula and prior parameter p... prior The updated estimated state value is calculated from the model prediction matrix H, the observation weight matrix R, and the prior information weight matrix W, thus realizing the fusion of the observation value z and the prior information. Step 2.5: Using the estimated state values ​​as parameters of the trigonometric series model, calculate the antenna pattern at the required location.

2. The high-precision antenna pattern testing method based on data assimilation according to claim 1, characterized in that, The process of obtaining the simulation data of the antenna under test is as follows: The antenna under test is modeled, and the far-field radiation pattern data of the corresponding simplified model is obtained in the simulation software, which is used as the simulation data of the antenna under test.

3. The high-precision antenna pattern testing method based on data assimilation according to claim 1, characterized in that, The specific testing process for the measured data of the antenna under test is as follows: Position the radiating aperture of the antenna under test at the calibration point of the planar near-field test system, with the main radiation direction aligned with the robotic arm; then connect the antenna under test and the probe mounted on the robotic arm to the vector network analyzer. According to the experimental requirements, set the test frequency band, sampling interval, side length of the antenna aperture, distance between the sampling plane and the antenna aperture, and side length of the sampling plane, and generate all preset sampling point information; The controller controls the robotic arm to sequentially collect and record the electric fields in the two polarization directions at the preset sampling points, and store them in the data memory. After the robotic arm controls the probe to complete sampling at all sampling points, it uploads the obtained near-field data to the control terminal. The near-field conversion algorithm built into the control terminal then converts the near-field data of the antenna under test into far-field data, thus obtaining the measured data of the antenna under test.

4. The high-precision antenna pattern testing method based on data assimilation according to claim 3, characterized in that, If the test requirement is to obtain the omnidirectional radiation information of the entire antenna, the antenna under test is rotated 180° horizontally during the planar near-field test, and the test process of the measured data of the antenna under test is repeated again to obtain the data in the elevation coordinate range [-180°, 180°] again. The omnidirectional radiation pattern is estimated again using the data assimilation method. By stitching together the first and second omnidirectional radiation patterns, a high-precision omnidirectional radiation pattern is obtained.

5. The high-precision antenna pattern testing method based on data assimilation according to claim 3, characterized in that, If the test requirement is to obtain the omnidirectional radiation information of the entire antenna, then during the planar near-field test, rotate the antenna under test horizontally by 180° and place it, repeat the test process of the antenna under test to obtain the data in the elevation coordinate range [-180°, 180°] again, which is the second obtained actual measurement data of the antenna under test; After stitching together the measured data of the antenna under test obtained from the two tests, the omnidirectional radiation pattern is estimated using the data assimilation method to obtain a high-precision omnidirectional radiation pattern.

6. The high-precision antenna pattern testing method based on data assimilation according to claim 1, characterized in that, Step 2.1 specifically involves: The trigonometric series model uses elevation coordinates. With gain Relationship: ； Represents a constant term. The coefficients represent the nth-order cosine function. The coefficients represent the nth-order sine function; the matrix form of the trigonometric series model is: ； in: ； ； ； In the formula, , , for The elevation angle coordinates of each point; As long as there are 2 N +1 sample, using the orthogonality of trigonometric functions to obtain 2 N +1 coefficient The values ​​of these coefficients are used as prior parameters, and the column vector of prior parameters takes the form shown in the following equation: 。 7. The high-precision antenna pattern testing method based on data assimilation according to claim 1, characterized in that, In step 2.2, the confidence interval is... The calculation formula is: ； Where L is the side length of the sampling plane, and D is the side length of the antenna aperture. This is the distance between the sampling plane and the antenna aperture.

8. A high-precision antenna pattern testing method based on data assimilation according to claim 6, characterized in that, In step 2.3, ; in, Indicates coordinates as The weights corresponding to the sampling points, ; The prior information weight matrix W is: .

9. A high-precision antenna pattern testing method based on data assimilation according to claim 8, characterized in that, In step 2.4, the updated state estimate The formula for calculation is: The estimated state value is in the following form: ； 2.5 Specifically, the estimated state values ​​are used as parameters of the trigonometric series model and substituted into the elevation coordinates. With gain In the relational expression, the antenna gain at different locations is calculated, and the expression is updated as follows: The antenna pattern at the desired location is obtained by calculating the antenna gain at different locations.

Images