Antenna pattern calibration method, apparatus and device for dual-polarized radar and medium

Abstract

The application relates to a kind of antenna pattern calibration method, device, equipment and medium of dual-polarized radar.The method comprises: the antenna pattern data of original three-dimensional dual-polarized radar is divided into multiple groups of two-dimensional data by data dimensionality reduction slice, then each group of two-dimensional data is fitted to obtain fitting parameter, the two-dimensional fitting result obtained using fitting parameter is de-duplicated, so that the fitting curve is more smooth and coherent, to obtain the high-quality antenna pattern with more stable curve and smaller error, then using the refined two-dimensional data after de-duplication reconstructs three-dimensional pattern, thereby significantly reducing the amount of calculation, reducing the calculation complexity and improving the fitting accuracy, thus greatly improving the efficiency and precision of obtaining dual-polarized radar antenna pattern, providing an important basis for the calibration of dual-polarized radar, and laying a foundation for accurate measurement of dual-polarized radar.

Description

Technical Field

[0001] This application relates to the field of radar technology, and in particular to an antenna pattern calibration method, apparatus, device, and medium for dual-polarized radar. Background Technology

[0002] Dual-polarization radar is a type of radar that uses two orthogonal polarization channels to simultaneously acquire echo signals from two polarization directions. By analyzing the echo signals from both polarization directions, the polarization characteristics of the target can be obtained. Compared to single-polarization radar, it can provide richer target information, thereby improving target detection and identification capabilities. Dual-polarization radar has a wide range of applications, such as meteorological monitoring, target detection and identification, and environmental monitoring.

[0003] With the advancement of radar technology, higher measurement accuracy requirements have been placed on dual-polarization radar. To meet the need for clearly and accurately acquiring the polarization characteristics of targets, precise system calibration is necessary. The first step in radar system calibration is obtaining the parameters of the antenna pattern, which is the foundation and prerequisite for subsequent calibrations. Only by calibrating based on the antenna pattern can the dual-polarization radar clearly and accurately obtain the polarization parameters of the target being measured, thereby improving measurement accuracy.

[0004] Obtaining a continuous antenna pattern from finite and discrete measurement data points requires fitting or interpolation. Researchers have proposed a surface fitting method for irregular three-dimensional data. This method constructs a nonlinear function using an artificial neural network model for fitting.

[0005] However, the above-mentioned method for obtaining the radiation pattern of a dual-polarized radar antenna ignores existing prior information and relies too much on a large amount of measurement data for multiple iterative training, resulting in low antenna radiation pattern fitting accuracy and problems of low efficiency and large error. Summary of the Invention

[0006] Therefore, it is necessary to provide a method for calibrating the antenna pattern of a dual-polarized radar, a device for calibrating the antenna pattern of a dual-polarized radar, a computer device, and a computer-readable storage medium to address the aforementioned technical problems.

[0007] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0008] On one hand, the present invention provides an antenna pattern calibration method for a dual-polarization radar, comprising:

[0009] The measurement data from the dual-polarization radar is preprocessed; the measurement data includes the azimuth, elevation, and echo amplitude of the calibration target obtained by the dual-polarization radar.

[0010] The azimuth angle, elevation angle, and echo amplitude are combined to form three-dimensional data. The three-dimensional data is sliced ​​to obtain multiple sets of two-dimensional data of echo amplitude corresponding to elevation angle i as the azimuth angle changes; i = 1, 2, ..., N, where N is the total number of elevation angles.

[0011] The nonlinear least squares method is used to fit the two-dimensional data corresponding to the N pitch angles with Gaussian functions to obtain the corresponding N sets of fitting parameters.

[0012] Based on a set of fitted azimuth angles and N sets of fitted parameters selected within a specific azimuth angle range, N sets of fitted echo amplitudes are generated respectively; a set of fitted azimuth angles includes multiple fitted azimuth angles, and the specific azimuth angle range is the azimuth angle range of the echo amplitude set.

[0013] N sets of fitted echo amplitudes are each paired with a set of fitted azimuth angles to form N sets of fitted two-dimensional data. Duplicate data are removed from the N sets of fitted two-dimensional data to obtain each set of refined two-dimensional data.

[0014] Each set of refined two-dimensional data and corresponding pitch angles are merged into refined three-dimensional data. The refined three-dimensional data is then smoothed by biharmonic spline interpolation to obtain reconstructed three-dimensional data.

[0015] The calibration factor of the measured reflectivity factor is calculated based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar.

[0016] The antenna pattern corresponding to the reconstructed 3D data is calibrated using a calibration factor to obtain the calibrated antenna pattern.

[0017] On the other hand, an antenna pattern calibration device for dual-polarized radar is also provided, comprising:

[0018] The data preprocessing module is used to preprocess the measurement data of the dual-polarization radar; the measurement data includes the azimuth, elevation and echo amplitude of the calibration target obtained by the dual-polarization radar.

[0019] The data slicing module is used to combine azimuth, elevation, and echo amplitude into three-dimensional data. It slices the three-dimensional data to obtain multiple sets of two-dimensional data of echo amplitude changing with azimuth for elevation angle i; i = 1, 2, ..., N, where N is the total number of elevation angles.

[0020] The data fitting module is used to fit each set of two-dimensional data corresponding to N pitch angles using a nonlinear least squares method and a Gaussian function to obtain the corresponding N sets of fitting parameters.

[0021] The new data generation module is used to generate N sets of fitted echo amplitudes based on a set of fitted azimuth angles selected within a specific azimuth angle range and N sets of fitted parameters. A set of fitted azimuth angles includes multiple fitted azimuth angles, and the specific azimuth angle range is the azimuth angle range of the echo amplitude set.

[0022] The data deduplication module is used to combine N sets of fitted echo amplitudes with a set of fitted azimuth angles to form N sets of fitted two-dimensional data, and then deduplicatize the N sets of fitted two-dimensional data to obtain each set of refined two-dimensional data.

[0023] The data interpolation module is used to merge each set of refined two-dimensional data and the corresponding pitch angle into refined three-dimensional data, and to use the biharmonic spline interpolation method to smooth the refined three-dimensional data to obtain the reconstructed three-dimensional data.

[0024] The data calibration module is used to calculate the calibration factor of the measured reflectivity factor based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar. The calibration factor is then used to calibrate the antenna pattern corresponding to the reconstructed three-dimensional data to obtain the calibrated antenna pattern.

[0025] In another aspect, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described dual-polarized radar antenna pattern calibration method.

[0026] Furthermore, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the above-described dual-polarized radar antenna pattern calibration method.

[0027] One of the above technical solutions has the following advantages and beneficial effects:

[0028] The aforementioned method, apparatus, equipment, and medium for calibrating the antenna pattern of dual-polarized radar employ a dimensionality reduction fitting approach. This method divides the original three-dimensional antenna pattern data of the dual-polarized radar into multiple sets of two-dimensional data through dimensionality reduction slicing. Then, fitting parameters are obtained by fitting each set of two-dimensional data. Finally, the three-dimensional pattern is reconstructed using the two-dimensional fitting results obtained from the fitting parameters. This significantly reduces the computational load, lowers computational complexity, and improves fitting accuracy. Therefore, it greatly improves the efficiency and accuracy of acquiring the antenna pattern of dual-polarized radar, providing an important basis for the calibration of dual-polarized radar and laying the foundation for accurate measurement of dual-polarized radar.

[0029] By employing a deduplication method, multiple sets of two-dimensional fitting data with similar elevation angles are deduplicated, resulting in smoother and more coherent fitting curves. This leads to a higher-quality antenna pattern with a more stable curve and smaller error, providing an important basis for the calibration of dual-polarized radar and laying the foundation for accurate measurement of dual-polarized radar. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this application or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart illustrating an antenna pattern calibration method for a dual-polarized radar in one embodiment.

[0032] Figure 2 This is a flowchart illustrating the measurement data preprocessing steps in one embodiment;

[0033] Figure 3 This is a flowchart illustrating the three-dimensional data slicing process in one embodiment;

[0034] Figure 4 This is a flowchart illustrating the process of deduplicating two-dimensional data in one embodiment;

[0035] Figure 5 This is a schematic diagram of the original measurement data in one embodiment;

[0036] Figure 6 This is a schematic diagram of preprocessed measurement data in one embodiment;

[0037] Figure 7 This is a schematic diagram of the antenna pattern corresponding to the reconstructed three-dimensional data in one embodiment;

[0038] Figure 8 This is a schematic diagram of the antenna pattern after the radar's pointing error has been calibrated in one embodiment.

[0039] Figure 9 This is a schematic diagram of the H-channel calibration factor with respect to angular position in one embodiment;

[0040] Figure 10 This is a schematic diagram of the module structure of an antenna pattern calibration device for a dual-polarized radar in one embodiment. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this application 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 and not intended to limit the scope of this application.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0043] It should be noted that, in this document, the reference to "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The presentation of this phrase in various locations throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments.

[0044] In the process of developing this application, the inventors discovered that during radar calibration using UAV-based point-to-point tracking, when there are repeated values ​​at measurement data points, the antenna pattern obtained after interpolation using a function exhibits overfitting, resulting in low fitting accuracy. Simultaneously, the inventors believe that the computational efficiency of two-dimensional data is far higher than that of three-dimensional data during fitting.

[0045] Based on this, the present invention provides a method, apparatus, device, and medium for calibrating the antenna pattern of a dual-polarized radar. It employs a dimensionality reduction fitting method, dividing the original three-dimensional dual-polarized radar antenna pattern data into multiple sets of two-dimensional data through data dimensionality reduction slicing. Then, fitting parameters are obtained by fitting each set of two-dimensional data. The three-dimensional pattern is then reconstructed using the two-dimensional fitting results obtained from the fitting parameters. This significantly reduces the computational load, lowers computational complexity, and improves fitting accuracy, thus greatly improving the efficiency and accuracy of obtaining dual-polarized radar antenna patterns. This provides an important basis for the calibration of dual-polarized radar and lays the foundation for accurate measurement of dual-polarized radar. A deduplication method is also employed, removing duplicates from multiple sets of two-dimensional fitting data with similar elevation angles. This makes the fitting curve smoother and more consistent, resulting in a higher-quality antenna pattern with a more stable curve and smaller error. This provides an important basis for the calibration of dual-polarized radar and lays the foundation for accurate measurement of dual-polarized radar.

[0046] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0047] Please see Figure 1 In one embodiment, this application provides an antenna pattern calibration method for a dual-polarized radar, including the following processing steps S11-S18:

[0048] S11 preprocesses the measurement data from the dual-polarization radar; the measurement data includes the azimuth, elevation, and echo amplitude of the calibration target obtained by the dual-polarization radar.

[0049] It is understandable that calibration targets can efficiently reflect the radiation signals of dual-polarized radar and generate strong echo signals. Corner reflectors, spherical reflectors, and plate reflectors can be used as calibration targets, but are not limited to these.

[0050] When acquiring measurement data, the first step is to use dual-polarized radar to locate a clutter-free area as the measurement region. After measurement, initial measurement data is obtained. To improve the quality and validity of the data, preprocessing of the initial measurement data is necessary. Preprocessing steps may include filtering interference data, deleting invalid data from the initial measurement stage, and replacing intermediate invalid data.

[0051] S12, the azimuth angle, pitch angle and echo amplitude are combined into three-dimensional data, the three-dimensional data are sliced ​​to obtain multiple sets of two-dimensional data of the echo amplitude corresponding to the pitch angle i as the azimuth angle changes; i = 1, 2, ..., N, N is the total number of pitch angles.

[0052] It is understandable that fitting data from horizontal or vertical motion separately yields better results than fitting data from a combination of both. To achieve a better fit, the three-dimensional data, consisting of azimuth, pitch, and echo amplitude, needs to be sliced. Each slice corresponds to a pitch angle, meaning each slice represents the data of the calibration target's horizontal motion (i.e., a fixed pitch angle with only a changing azimuth). The data from each slice provides the foundation for subsequent fitting. The total number of pitch angles is N, and each pitch angle corresponds to multiple sets of two-dimensional data. Each set of two-dimensional data consists of the azimuth and echo amplitude of the calibration target at a specific measurement point.

[0053] Dimensionality reduction slicing of the three-dimensional data composed of measurement data can reduce the amount of computation required for subsequent fitting, reduce computational complexity, and lay the foundation for obtaining better fitting results.

[0054] S13. The nonlinear least squares method is used to fit the two-dimensional data corresponding to the N pitch angles to the Gaussian function to obtain the corresponding N sets of fitting parameters.

[0055] As can be understood, nonlinear least squares is a widely used optimization algorithm for curve fitting. It achieves a good fit between the data and a nonlinear model by solving for the optimal parameters to minimize the sum of squared residuals. The Gaussian function is a commonly used nonlinear function model. The expression for a one-dimensional Gaussian function is: Where x represents the azimuth angle, f(x) represents the echo amplitude, A represents the peak amplitude, u represents the peak center position, and σ represents the standard deviation.

[0056] Different azimuth angles at the same elevation angle can be denoted as az1, az2, ..., az n And their corresponding echo amplitudes, for example, can be denoted as amp1, amp2, ..., amp n The objective function for nonlinear least squares fitting is:

[0057]

[0058] Where A, u, and σ are the parameters to be fitted, where A represents the peak amplitude, u represents the peak center location, and σ represents the standard deviation. The goal of the fitting is to minimize the Q value.

[0059] For N pitch angles, obtain the corresponding N sets of fitting parameters, such as A1, u1, σ1; A2, u2, σ2; ...; A N u N , σ N .

[0060] S14. Based on a set of fitted azimuth angles and N sets of fitted parameters selected within a specific azimuth angle range, generate corresponding N sets of fitted echo amplitudes respectively; a set of fitted azimuth angles includes multiple fitted azimuth angles, and the specific azimuth angle range is the azimuth angle range of the echo amplitude set.

[0061] It is understandable that in the measurement data of dual-polarized radar, the echo amplitude of the calibrated target usually exhibits a Gaussian distribution-like variation trend with respect to the azimuth angle. That is, within a certain azimuth angle range, the echo amplitude is larger, while outside this range, the echo amplitude is smaller. This certain azimuth angle range is the azimuth angle range where the echo amplitude is concentrated, and is called the specific azimuth angle range.

[0062] Within a specific azimuth range, a set of fitted azimuth angles can be selected. This can be done by selecting the fitted azimuth angles according to an arithmetic sequence or by randomly selecting them. For example... For the first set of fitted parameters, such as A1, u1, and σ1, substitute each fitted azimuth angle and the first set of fitted parameters into the above one-dimensional Gaussian function expression to generate the corresponding first set of fitted echo amplitudes, for example... Similarly, for the second set of fitted parameters, such as A2, u2, and σ2, the corresponding second set of fitted echo amplitudes is generated, for example... Similarly, by continuously calculating the amplitudes of N sets of fitted echoes, the final N sets of fitted echo amplitudes are obtained, for example...

[0063] S15. Combine the N sets of fitted echo amplitudes with a set of fitted azimuth angles to form N sets of fitted two-dimensional data. Remove duplicates from the N sets of fitted two-dimensional data to obtain each set of refined two-dimensional data.

[0064] It is understandable that overfitting can easily occur during interpolation when multiple sets of data points with the same or similar pitch angles but different echo amplitudes are used. Therefore, it is necessary to first deduplicate the N sets of fitted two-dimensional data. Among the multiple sets of fitted two-dimensional data with the same or similar pitch angles, only one set with the highest correlation is retained, and the remaining identical or similar fitted two-dimensional data are deleted, thus obtaining refined two-dimensional data sets for each set. For example...

[0065] By deduplicating multiple sets of two-dimensional fitting data with similar elevation angles, the fitting curves become smoother and more coherent, resulting in a higher-quality antenna pattern with a more stable curve and smaller error.

[0066] S16. The refined two-dimensional data and corresponding pitch angles of each group are merged into refined three-dimensional data. The refined three-dimensional data are then smoothed by biharmonic spline interpolation to obtain reconstructed three-dimensional data.

[0067] It's understandable that the number of refined 2D data sets obtained after deduplication is less than N, for example, 23 sets of refined 2D data, each corresponding to a pitch angle. After assembling each set of refined 2D data and its corresponding pitch angle into separate 3D data sets, these sets are further merged into a single refined 3D data set. For example, after assembling 23 sets of refined 2D data and their corresponding pitch angles into 23 sets of 3D data, these 23 sets of 3D data are then further merged into a single refined 3D data set. Since this refined 3D data set is obtained by directly merging 3D data sets with different pitch angles, the data is not smooth and continuous. To obtain smooth and continuous 3D data, further smoothing interpolation processing is needed on this merged refined 3D data set. After interpolation processing, a smooth and continuous reconstructed 3D data set can be obtained.

[0068] The interpolation surface is a linear combination of Green's functions centered at each newly generated data point. For example, the fitted surface can be represented as:

[0069]

[0070] Among them, w j It is the weight of the j-th data point. It is the position vector of the unknown data point. It is the position constraint of the j-th data. It is the Green's function, and the specific calculation is as follows:

[0071]

[0072] weight w j By requiring surface It is determined precisely through all fitted data points, and the specific calculation is as follows:

[0073]

[0074] For example, the above equation generates a system of linear equations with 11,523 unknowns. Solving this system of equations will yield the weights w. j The value can be calculated using the bitohmic spline interpolation function in Matlab.

[0075] Using biharmonic spline interpolation can further improve the smoothness and continuity of the data, increase the number of sampling points, improve the data density, and improve the effect of 3D reconstruction.

[0076] S17. Calculate the calibration factor of the measured reflectivity factor based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar.

[0077] It is understandable that the theoretical reflectivity factor of the calibration target can be calculated from the theoretical echo power and radar constant.

[0078] dBZ target =C+p r +20lgL+L×L at

[0079] Where C is the radar constant, p r L represents the theoretical echo power, and L is the distance between the radar and the calibration target. at This represents the loss per kilometer of electromagnetic wave propagation through the atmosphere.

[0080] In the above expressions, the theoretical echo power and radar constant are calculated as follows:

[0081]

[0082] Among them, P t Where λ is the transmitted pulse power, G is the antenna gain, λ is the wavelength, L is the distance between the radar and the calibration target, σ is the scattering cross-section, and θ is the horizontal beamwidth. L is the vertical beamwidth. in This represents the total system loss excluding atmospheric losses.

[0083] For example, if the calibration target is a corner reflector, the scattering cross-section is calculated in the above expression as follows:

[0084]

[0085] Where a is the side length of the corner reflector and λ is the wavelength.

[0086] By quantizing the echo power of the H-channel and V-channel of a dual-polarization radar, the reflectivity factor (dBZ) of the H-channel, as measured by the radar system, can be obtained. H and the reflectivity factor dBZ of the V channel V .

[0087] Compare the theoretical reflectivity factor dBZ of the calibration target target The calibration factors for the reflectance factors of both channels can be obtained. The calibration factor α for the H channel... H and the calibration factor α of the V channel V for:

[0088] α H = dBZ target -dBZ H

[0089] α V = dBZ target -dBZ V

[0090] S18. The antenna pattern corresponding to the reconstructed three-dimensional data is calibrated using a calibration factor to obtain the calibrated antenna pattern.

[0091] It is understandable that the calculated calibration factor is used to calculate the echo amplitude of each data point in the reconstructed 3D data, compensating for any amplitude errors, and ultimately obtaining an accurate dual-channel 3D antenna pattern. The specific process of completing the calibration using the calibration factor can be understood by referring to the existing antenna pattern calibration procedures in this field.

[0092] In the aforementioned method for calibrating the antenna pattern of a dual-polarized radar, a dimensionality reduction fitting approach is employed. The original three-dimensional antenna pattern data of the dual-polarized radar is divided into multiple sets of two-dimensional data through dimensionality reduction slicing. Then, fitting parameters are obtained by fitting each set of two-dimensional data. Finally, the three-dimensional pattern is reconstructed using the two-dimensional fitting results obtained from the fitting parameters. This significantly reduces the amount of computation, lowers the computational complexity, and improves the fitting accuracy. Therefore, it greatly improves the efficiency and accuracy of obtaining the antenna pattern of the dual-polarized radar, providing an important basis for the calibration of the dual-polarized radar and laying the foundation for accurate measurement of the dual-polarized radar.

[0093] By employing a deduplication method, multiple sets of two-dimensional fitting data with similar elevation angles are deduplicated, resulting in smoother and more coherent fitting curves. This leads to a higher-quality antenna pattern with a more stable curve and smaller error, providing an important basis for the calibration of dual-polarized radar and laying the foundation for accurate measurement of dual-polarized radar.

[0094] Please see Figure 2 In one embodiment, the step of preprocessing the measurement data of the dual-polarization radar in the above method of the present invention may include the following sub-steps S111-S112:

[0095] S111: Based on the scanning data of the calibration target by the dual-polarized radar and the data provided by the calibration system, a range Doppler image is plotted. After filtering out the interference points caused by the carrier of the calibration target in the range Doppler image, the measurement data corresponding to the calibration target is extracted.

[0096] It is understandable that the calibration system can be an existing part of the dual-polarization radar, used for system calibration and providing position parameters. Alternatively, the calibration system can be an independent existing position system connected to the dual-polarization radar to provide position information, as long as the position information of the calibration target can be obtained. Based on the scan data and the position information provided by the calibration system, a range-Doppler map can be plotted. The range-Doppler map is an important graphical representation method in radar signal processing; the coordinate values ​​of a point on the Doppler frequency axis can be converted into the radial velocity of the target.

[0097] Since the radial velocity of the calibration target relative to the radar is 0, the point corresponding to the zero velocity in the range-Doppler image is the echo of the calibration target. During the measurement process, the carrier of the calibration target will introduce interference. For example, when the carrier is a UAV, the rotation of the UAV rotor will generate a Doppler frequency shift component, which will cause interference. This will appear as a non-zero velocity echo in the range-Doppler image. By filtering out the points corresponding to these non-zero velocities, the rotor echo is suppressed, and finally the echo measurement data of the calibration target is extracted.

[0098] S112, Initial invalid data is deleted from the measurement data and outlier values ​​of each echo amplitude are replaced using default values ​​to obtain preprocessed measurement data.

[0099] It is understandable that in the initial stage of measurement, the calibration target is not yet within the effective measurement range, and the data obtained at this time is initially invalid data. In order to improve the quality and validity of the data, it is necessary to delete this initial invalid data. In the measurement data, changing the pitch angle will cause abnormal echo amplitude, thus generating abnormal echo amplitude values. Replace each abnormal echo amplitude value with a default value, such as 0 or other easily distinguishable and usable settings, to provide boundaries for subsequent 3D data tiling.

[0100] Preprocessing of dual-polarization radar measurement data improves the quality and validity of the data, providing an important foundation for subsequent slicing and calibration.

[0101] Please see Figure 3 In one embodiment, the process of slicing three-dimensional data in the method of the present invention may include the following sub-steps S121-S126:

[0102] S121, store the positions of each default value into the first array respectively.

[0103] It is understandable that in step S112, changing the pitch angle produces anomalies in the echo amplitude. The default values ​​are used to replace the anomalies in the echo amplitude. That is, the pitch angles corresponding to the data before and after the default values ​​are different. Therefore, the default values ​​are used for slicing. First, the position corresponding to each default value is stored in the first array to prepare for subsequent slicing.

[0104] S122: Iterate through each position in the first array. If the position is not adjacent to the next position, pair the position with the next position and store each pair in the second array.

[0105] It's understandable that for each position currently being processed, a position pair is formed as long as that position is not adjacent to its next position. This process is repeated for each position, resulting in a total of position pairs after the traversal is complete. These position pairs are then stored in a second array, where each pair represents the start and end points of a data slice.

[0106] S123, combine the position of the first echo amplitude and the position before the first default value to form a head position pair, and store the head position pair in the second array.

[0107] It is understandable that in step S112, the initial invalid data of the measurement data has been deleted, so the value of the first echo amplitude is not the default value. There is a data slice before the first default value. The starting point of this data slice is the position of the first echo amplitude, and the ending point of this data slice is the position before the first default value. Therefore, the position of the first echo amplitude and the position before the first default value need to be combined to form the beginning position pair and stored in the second array.

[0108] S124, determine the end position pair based on the value of the last echo amplitude and store it in the second array.

[0109] In some implementations, step S124 above may specifically include the following step S124A: if the value of the last echo amplitude is the default value, then the position pair corresponding to the value of the last echo amplitude is determined as the end position pair.

[0110] It is understandable that if the value of the last echo amplitude is the default value, then the position pair corresponding to the value of the last echo amplitude is the last data slice, and its position pair is the end position pair.

[0111] In some implementations, step S124 above may specifically include step S124B, where if the value of the last echo amplitude is not the default value, the position after the last default value and the position of the last echo amplitude are combined to form an end position pair.

[0112] It is understandable that if the value of the last echo amplitude is not the default value, then there is another data slice after the default value. The starting point of this data slice is the position after the last default value, and the ending point of this data slice is the position of the last echo amplitude. Therefore, the position after the last default value and the position of the last echo amplitude need to be combined to form an end position pair.

[0113] S125, delete the position pairs in the second array that are not within the specific azimuth range corresponding to the azimuth angle.

[0114] It is understandable that, in order to improve detection accuracy, processing and analysis are limited to a specified effective azimuth angle range, and positions outside the specified azimuth angle range are deleted.

[0115] S126, through difference filtering, remove overfitted position pairs in the second array.

[0116] It is understandable that, based on previous observations and measurements, if a position pair contains too little data, it indicates that anomalies in echo amplitude appear in the measurement data for the same pitch angle. During segmentation, the data corresponding to that pitch angle is divided into multiple groups. These multiple groups of data have the same pitch angle, but the fitting results are different, leading to overfitting during surface fitting. To avoid dividing a set of data corresponding to the same pitch angle into multiple groups during segmentation, which would introduce overfitting again later, position pairs where the absolute value of the difference between the two values ​​is less than a specific value are deleted.

[0117] In one embodiment, the antenna pattern calibration method for a dual-polarized radar provided by the present invention provides that the fitted azimuth angle values ​​selected within a specific azimuth angle range form an arithmetic sequence.

[0118] It can be understood that by starting with the minimum value of a specific azimuth range and ending with the maximum value of a specific azimuth range, the tolerance of the selected fitting azimuth can be determined, thus determining a set of fitting azimuths that form an arithmetic sequence.

[0119] Arithmetic sequences ensure that the fitted azimuth angles are uniformly distributed within a specified range, without clustering or missing values, thus facilitating accurate fitting and analysis of echoes across the entire range. Equally spaced fitted azimuth angles maximize the use of correlation information between each azimuth angle sampling point, ensuring optimal curve division between two sampling points and reducing overall fitting error.

[0120] Please see Figure 4 In one embodiment, the process of removing duplicates from N sets of fitted two-dimensional data to obtain refined two-dimensional data in the above method of the present invention includes the following sub-steps S151-S153:

[0121] S151, iterate through N groups of fitted two-dimensional data, select fitted two-dimensional data with the same pitch angle or a pitch angle difference less than a set difference between groups, and obtain each similar fitted two-dimensional dataset.

[0122] It can be understood that, for the currently processed set of fitted 2D data, as long as the pitch angle of this set is the same as the pitch angle of the next set, or the difference between the two is less than a set difference, the next set of fitted 2D data is included in the same similar fitted 2D dataset along with the current set of fitted 2D data. Each set of fitted 2D data is processed separately in this way, thus obtaining various similar fitted 2D datasets after traversal. All fitted 2D data are divided into multiple individual 2D fitted data sets and multiple similar fitted 2D datasets. Within a similar fitted 2D dataset, there are two or more sets of fitted 2D data sets.

[0123] S152, For each similar fitting two-dimensional dataset, calculate the correlation coefficient between each set of fitted two-dimensional data in the similar fitting two-dimensional dataset and the two adjacent sets of fitted two-dimensional data.

[0124] It is understandable that in a similarly fitted two-dimensional dataset, there exist more than two sets of fitted two-dimensional data. For example, if the pitch angle difference is less than or equal to 0.05°, it is considered the same (difference value is 0) or similar pitch angle. The pitch angles of groups 8, 9, and 10 are 5.05°, and the pitch angle of group 11 is 5.0427°. These four sets of data are considered to have the same pitch angle and form a similarly fitted two-dimensional dataset. The correlation coefficient between the fitted echo amplitude of group 8 and the fitted echo amplitudes of groups 7 and 12 is calculated. Similarly, the same operation is performed on the remaining three groups to obtain four correlation coefficients, such as ρ1, ρ2, ρ3, and ρ4. The calculation of the correlation coefficients is as follows:

[0125]

[0126] Among them, R(y) k i ), R(y l i ) represent the rank of the i-th data in the k-th and l-th groups, respectively. are the average rank of the k-th and l-th data groups, respectively, and n is the number of data points in each group.

[0127] S153: Retain the set of fitted two-dimensional data with the highest correlation coefficient in each fitted two-dimensional dataset to obtain refined two-dimensional data.

[0128] It's understandable that for a similarly fitted two-dimensional dataset, the number of fitted two-dimensional data sets will result in the number of correlation coefficients calculated. For example, if four correlation coefficients are obtained: ρ1, ρ2, ρ3, and ρ4, the magnitudes of these four correlation coefficients are compared. If ρ2 is the largest, then the fitted two-dimensional data set corresponding to ρ2 is retained, while the fitted two-dimensional data sets corresponding to ρ1, ρ3, and ρ4 are deleted. This process is repeated for each similarly fitted two-dimensional dataset, and all the retained fitted two-dimensional data sets form the refined two-dimensional dataset.

[0129] In some embodiments, to more intuitively and comprehensively illustrate the antenna pattern calibration method for dual-polarized radar described above, the following are application examples of this method. It should be noted that the implementation examples given in this specification are merely illustrative and not the only limitation on specific implementation examples of the present invention. Those skilled in the art can use the antenna pattern calibration method for dual-polarized radar provided above, based on the illustrative examples provided by the present invention, to achieve antenna pattern calibration for different application scenarios.

[0130] For raw measurement data (such as Figure 5 The following processing is performed on the data (as shown): filtering out vibrations from the drone rotor; cleaning the data; identifying and marking invalid data to obtain preprocessed measurement data (such as...). Figure 6 (As shown). During the calibration experiment to acquire measurement data, the UAV flew in a horizontal zigzag pattern. When acquiring target echo data, the target was stationary relative to the radar. The two raw data streams obtained by rotating the radar beam clockwise and counterclockwise are stored in dataCalibrationUp (radar clockwise rotation) and dataCalibrationDown (radar counterclockwise rotation), respectively.

[0131] Plot a range-Doppler map based on the original data and identify the points where the velocity is zero. Keep these points and record the resulting new data as `new_dataCalibrationUp` and `new_dataCalibrationDown`. Remove consecutive invalid data from rows 1-4 of `new_dataCalibrationUp` and rows 1-3 of `new_dataCalibrationDown`, keeping only the remaining data. Columns 1-3 of `new_dataCalibrationUp` represent the azimuth, elevation, and H-channel echo amplitude values ​​when the radar rotates clockwise, denoted as `azUp`, `elUp`, and `amplitudeH_Up`, respectively. Columns 1-3 of `new_dataCalibrationDown` represent the azimuth, elevation, and H-channel echo amplitude values ​​when the radar rotates counter-clockwise, denoted as `azDown`, `elDown`, and `amplitudeH_Down`, respectively. Convert invalid data in `amplitudeH_Up` and `amplitudeH_Down` to 0.

[0132] Store the positions of the zero elements in amplitudeH_Up into array A, where A = [A1, A2, ..., A...]. 3378 ] T .

[0133] Traverse array A and store the positions of two non-adjacent zero elements in A into f_double. The condition for determining whether to store in f_double is: A(i+1)! = A(i)+1.

[0134] Since the index of the first zero value in amplitudeH_Up is not 1, and there is a set of data with the same pitch angle before it, [1,A(1)-1] needs to be added to the first line of f_double.

[0135] Determine if the last element in amplitudeH_Up is zero. If it is not zero, store A(3378)+1 and the position of the last element in amplitudeH_Up into f_double. The final f_double is a 26×2 matrix.

[0136] The center of the antenna pattern is located between 303° and 308° azimuth. Values ​​of f_double that do not contain data within this range are discarded. To avoid overfitting caused by dividing a set of data corresponding to the same elevation angle into multiple sets during the initial partitioning, rows in the f_double row where the absolute value of the difference between the two values ​​is less than 1000 are deleted.

[0137] The fitting parameters are obtained by fitting the two-dimensional data with a Gaussian function under a fixed pitch angle using the nonlinear least squares method.

[0138] For the experimentally measured azimuth data at the same elevation angle: az1, az2, ..., az n and their corresponding echo amplitudes: amp1, amp2, ..., amp n The objective function for nonlinear least squares fitting is:

[0139]

[0140] The goal of the fitting is to minimize the Q-value. Here, A, u, and σ are the parameters to be fitted, representing the peak amplitude, peak center location, and standard deviation, respectively. The fitted parameters are stored in one-dimensional arrays a, b, and c.

[0141] New data on echo amplitude for each group under a fixed pitch angle and a changing azimuth angle are generated using the fitted parameters.

[0142] Since the echoes are concentrated between 303° and 308° azimuth angles, to ensure fitting accuracy while reducing data footprint and computation time, 501 azimuth angle data points were generated at equal intervals of 0.01, resulting in 26 sets of identical data (f_double has 26 different elevation angle values), denoted as x. i (i = 1, 2, ..., 26).

[0143] Create a length with x i The same, both are one-dimensional arrays z with values ​​of elUp(f_double(i)). i As new pitch angle data.

[0144] Based on the fitted parameters, the corresponding pitch angle z is calculated. i The fitted echo amplitude is denoted as y. i .

[0145] For multiple sets of data with similar elevation angles, each set of data is correlated with its two adjacent non-repeating sets of data, and the set with the highest correlation is retained. In the data obtained by the radar rotating clockwise, three sets of data have an elevation angle of 5.05°, and one set has an elevation angle of 5.0427°. These four sets of data are considered to have the same elevation angle. The Spearman correlation coefficients between the four sets of data and the fitted data at adjacent elevation angles are calculated using the `corr` function in MATLAB: 0.9994, 0.9949, 0.9975, and 0.9986, respectively. The set with a correlation coefficient of 0.9994 is retained. In the data obtained by the radar rotating counterclockwise, three sets of data correspond to an elevation angle of 5.05°. Similarly, their correlation coefficients are calculated: 0.9932, 0.9964, and 0.9969, respectively. The set with a correlation coefficient of 0.9969 is retained.

[0146] The fitted multiple sets of two-dimensional data: (x i j ,y i j (i = 1, 2, ..., 23; j = 1, 2, ..., 501) and the pitch angle are combined to form three-dimensional data: (x j ,y j ,z j (j=1,2,...,23×501), perform bitone spline interpolation using the spline interpolation method.

[0147] The interpolation surface is a linear combination of Green's functions centered at each newly generated data point. The fitted surface is represented as:

[0148]

[0149] Where n is the total number of data points, such as n = 23 × 501, w j It is the weight of the j-th data point. It is the position vector of the unknown data point. It is the position constraint of the j-th data.

[0150] It's the Green's function. The weight w j By requiring surface It is determined precisely through all the fitted data points, that is:

[0151]

[0152] The above equation generates a system of linear equations with 11,523 unknowns. Solving this system of equations will yield the weight w. j The value can be calculated using the bitohmic spline interpolation function in Matlab.

[0153] like Figure 7 As shown, the antenna pattern obtained by fitting the echo amplitude values ​​measured by the radar clockwise and counterclockwise rotation has a fixed angular deviation. Figure 8 As shown, the pointing error of the radar itself is calibrated by bringing the maximum values ​​of the two antenna patterns to the same angle. In this embodiment, the azimuth angle of the antenna pattern obtained by clockwise rotation of the radar is increased by 0.5°, and the elevation angle is decreased by 0.095°; the azimuth angle of the antenna pattern obtained by counterclockwise rotation of the radar is decreased by 0.5°, and the elevation angle is increased by 0.095°, thus moving the maximum values ​​of the two antenna patterns to the same angle.

[0154] Assuming the side length of the trihedral angle is 'a', then when the radar illumination direction is parallel to the axis of symmetry, its radar backscattering cross-section is:

[0155]

[0156] Where λ is the wavelength. According to the radar equation: P r =P t G 2 σλ 2 / ((4π) 3 L 4 The theoretical echo power of the trihedral reflection can be calculated. Where P t Let G be the transmitted pulse power, λ be the antenna gain, λ be the wavelength, and L be the distance between the radar and the trihedral reflector. The theoretical echo power P of the trihedral reflector is calculated. r Taking the logarithm:

[0157] p r =10lg(P r ×10 3 )

[0158] p r The unit is dBm. The radar constant C can be expressed as:

[0159]

[0160] Where θ is the horizontal beamwidth; L is the vertical beamwidth. in P represents the total system loss excluding atmospheric losses. t This represents the radar transmit power. (From p) r The theoretical reflectivity factor of the trihedral reflector can be calculated from the radar constant C as follows:

[0161] dBZ target =C+p r +20lgL+L×L at

[0162] Where L at This represents the loss per kilometer of electromagnetic wave propagation through the atmosphere.

[0163] By quantizing the echo power of the H-channel and V-channel, the reflectivity factor (dBZ) of the H-channel measured by the radar system can be obtained. H and the reflectivity factor dBZ of the V channel V Compare the theoretical reflectivity factor dBZ of the trihedral reflector. target The calibration factors for the reflectance factors of both channels can be obtained. The calibration factor α for the H channel... H and the calibration factor α of the V channel V for:

[0164] α H = dBZ target -dBZ H

[0165] α V = dBZ target -dBZ V

[0166] like Figure 9 The image shown is an image of the calibration factor at each location calculated according to the above formula.

[0167] Using the calibration factor α of the H channel H and the calibration factor α of the V channel V The antenna pattern corresponding to the three-dimensional data is calibrated to obtain the calibrated antenna pattern.

[0168] It should be understood that, although Figure 1-4 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1-4 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0169] In one embodiment, such as Figure 10As shown, an antenna pattern calibration 200 for a dual-polarized radar is provided, including a data preprocessing module 21, a data slicing module 22, a data fitting module 23, a new data generation module 24, a data deduplication module 25, a data interpolation module 26, and a data calibration module 27. The system includes: a data preprocessing module 21 for preprocessing measurement data from the dual-polarization radar; the measurement data includes the azimuth, elevation, and echo amplitude of the calibration target obtained by the dual-polarization radar; a data slicing module 22 for composing three-dimensional data from the azimuth, elevation, and echo amplitude, slicing the three-dimensional data to obtain multiple sets of two-dimensional data showing the echo amplitude changing with the azimuth at elevation angle i; i = 1, 2, ..., N, where N is the total number of elevation angles; a data fitting module 23 for fitting the two-dimensional data corresponding to each of the N elevation angles using a Gaussian function via a nonlinear least squares method to obtain the corresponding N sets of fitting parameters; and a new data generation module 24 for generating the corresponding N sets of fitted echo amplitudes based on a set of fitted azimuth angles selected within a specific azimuth range and the N sets of fitting parameters. A set of fitted azimuth angles includes multiple fitted azimuth angles, and the specific azimuth angle range is the azimuth angle range where the echo amplitude is concentrated; the data deduplication module 25 is used to form N sets of fitted two-dimensional data by correspondingly matching N sets of fitted echo amplitudes with a set of fitted azimuth angles, and to remove duplicates from the N sets of fitted two-dimensional data to obtain each set of refined two-dimensional data; the data interpolation module 26 is used to merge each set of refined two-dimensional data and the corresponding elevation angle into refined three-dimensional data, and to perform smooth interpolation on the refined three-dimensional data using the biharmonic spline interpolation method to obtain reconstructed three-dimensional data; the data calibration module 27 is used to calculate the calibration factor of the measured reflectivity factor based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar; and to calibrate the antenna pattern corresponding to the reconstructed three-dimensional data using the calibration factor to obtain the calibrated antenna pattern.

[0170] The antenna pattern calibration device 200 for the aforementioned dual-polarized radar employs a dimensionality reduction fitting method. It divides the original three-dimensional dual-polarized radar antenna pattern data into multiple sets of two-dimensional data through dimensionality reduction slicing. Then, it fits each set of two-dimensional data to obtain fitting parameters, and reconstructs the three-dimensional pattern using the two-dimensional fitting results obtained from the fitting parameters. This significantly reduces the computational load, lowers computational complexity, and improves fitting accuracy. Therefore, it greatly improves the efficiency and accuracy of acquiring dual-polarized radar antenna patterns, providing an important basis for dual-polarized radar calibration and laying the foundation for accurate dual-polarized radar measurements. Furthermore, by employing a deduplication method, multiple sets of two-dimensional fitting data with similar elevation angles are deduplicated, resulting in smoother and more consistent fitting curves. This leads to a higher-quality antenna pattern with a more stable curve and smaller error, providing an important basis for dual-polarized radar calibration and laying the foundation for accurate dual-polarized radar measurements.

[0171] In one embodiment, the step of preprocessing the measurement data of the dual-polarized radar in the above-described apparatus 200 of the present invention includes: drawing a range Doppler image based on the scanning data of the dual-polarized radar on the calibration target and the data provided by the calibration system; filtering out the interference points caused by the carrier of the calibration target in the range Doppler image and extracting the measurement data corresponding to the calibration target; deleting the initial invalid data of the measurement data and replacing the abnormal values ​​of each echo amplitude using the default value to obtain the preprocessed measurement data.

[0172] In one embodiment, the process of slicing the three-dimensional data in the above-described apparatus 200 of the present invention includes: storing the positions of each default value into a first array; traversing each position in the first array, and if the position is not adjacent to the next position, forming a position pair with the next position, and storing each position pair into a second array; forming a first-end position pair with the position of the first echo amplitude and the position preceding the first default value, and storing the first-end position pair into the second array; determining the last position pair based on the value of the last echo amplitude and storing it into the second array; deleting position pairs in the second array that correspond to azimuth angles not within a specific azimuth angle range; and deleting overfitted position pairs in the second array through difference filtering.

[0173] In one embodiment, the process of determining the end position pair based on the value of the last echo amplitude in the above-described apparatus 200 of the present invention includes: if the value of the last echo amplitude is a default value, then determining the position pair corresponding to the value of the last echo amplitude as the end position pair.

[0174] In one embodiment, the process of determining the end position pair based on the value of the last echo amplitude in the above-described apparatus 200 of the present invention includes: if the value of the last echo amplitude is not a default value, then forming an end position pair by combining the position after the last default value and the position of the last echo amplitude.

[0175] In one embodiment, when selecting the fitting azimuth angle in the above-described device 200 of the present invention, the values ​​of each fitting azimuth angle selected within a specific azimuth angle range form an arithmetic sequence.

[0176] In one embodiment, the process of deduplicating N sets of fitted two-dimensional data and obtaining each set of refined two-dimensional data in the above-described apparatus 200 of the present invention includes: traversing the N sets of fitted two-dimensional data, selecting fitted two-dimensional data with the same pitch angle or a pitch angle difference less than a set difference between the groups to obtain each similar fitted two-dimensional dataset; for each similar fitted two-dimensional dataset, calculating the correlation coefficient between each set of fitted two-dimensional data in the similar fitted two-dimensional dataset and the two adjacent sets of fitted two-dimensional data; and retaining the set of fitted two-dimensional data with the largest correlation coefficient in each fitted two-dimensional dataset to obtain refined two-dimensional data.

[0177] Specific limitations regarding the antenna pattern calibration device 200 for dual-polarized radar can be found in the corresponding limitations of the antenna pattern calibration method for dual-polarized radar mentioned above, and will not be repeated here. Each module in the aforementioned antenna pattern calibration device 200 for dual-polarized radar can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware or independently of the processor in a computer device, or stored in software in the memory of a computer device, so that the processor can call and execute the operations corresponding to each module.

[0178] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: preprocessing measurement data from a dual-polarized radar; the measurement data includes the azimuth, elevation, and echo amplitude of a calibration target obtained by the dual-polarized radar; assembling the azimuth, elevation, and echo amplitude into three-dimensional data; slicing the three-dimensional data to obtain multiple sets of two-dimensional data showing the echo amplitude changing with the azimuth at elevation angle i; i = 1, 2, ..., N, where N is the total number of elevation angles; using a nonlinear least squares method to fit each set of two-dimensional data corresponding to the N elevation angles using a Gaussian function to obtain the corresponding N sets of fitting parameters; and selecting a set of fitted azimuth angles within a specific azimuth range... N sets of fitting parameters are used to generate corresponding N sets of fitted echo amplitudes. Each set of fitted azimuth angles includes multiple fitted azimuth angles, with a specific azimuth angle range being the azimuth angle range of the echo amplitude set. The N sets of fitted echo amplitudes are combined with each set of fitted azimuth angles to form N sets of fitted two-dimensional data. The N sets of fitted two-dimensional data are deduplicated to obtain each set of refined two-dimensional data. Each set of refined two-dimensional data and the corresponding elevation angle are combined to form refined three-dimensional data. The refined three-dimensional data are then smoothed using biharmonic spline interpolation to obtain reconstructed three-dimensional data. Based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar, the calibration factor of the measured reflectivity factor is calculated. The antenna pattern corresponding to the reconstructed three-dimensional data is calibrated using the calibration factor to obtain the calibrated antenna pattern.

[0179] It is understood that, in addition to the memory and processor mentioned above, the computer equipment described above also includes other hardware and software components not listed in this specification. The specific components can be determined according to the model of the computer equipment in different application scenarios, and will not be listed and described in detail in this specification.

[0180] In one embodiment, when the processor executes the computer program, it can also implement the steps or sub-steps added to the various embodiments of the antenna pattern calibration method for the dual-polarized radar described above.

[0181] In one embodiment, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, it performs the following steps: preprocessing measurement data from a dual-polarized radar; the measurement data includes the azimuth, elevation, and echo amplitude of the calibration target obtained by the dual-polarized radar; assembling the azimuth, elevation, and echo amplitude into three-dimensional data; slicing the three-dimensional data to obtain multiple sets of two-dimensional data corresponding to the echo amplitude i as a function of the azimuth; i = 1, 2, ..., N, where N is the total number of elevation angles; using a nonlinear least squares method to fit each set of two-dimensional data corresponding to the N elevation angles with a Gaussian function to obtain the corresponding N sets of fitting parameters; and based on a set of fitted azimuth angles selected within a specific azimuth range and the N sets of fitted parameters... The parameters are combined to generate N sets of fitted echo amplitudes. Each set of fitted azimuth angles includes multiple fitted azimuth angles, with a specific azimuth angle range being the azimuth angle range of the echo amplitude set. The N sets of fitted echo amplitudes are combined with a set of fitted azimuth angles to form N sets of fitted two-dimensional data. The N sets of fitted two-dimensional data are deduplicated to obtain each set of refined two-dimensional data. Each set of refined two-dimensional data and the corresponding elevation angle are combined to form refined three-dimensional data. The refined three-dimensional data is smoothed by biharmonic spline interpolation to obtain reconstructed three-dimensional data. The calibration factor of the measured reflectivity factor is calculated based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar. The antenna pattern corresponding to the reconstructed three-dimensional data is calibrated using the calibration factor to obtain the calibrated antenna pattern.

[0182] In one embodiment, when the computer program is executed by the processor, it can also implement the steps or sub-steps added to the various embodiments of the antenna pattern calibration method for the dual-polarized radar described above.

[0183] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0184] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0185] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for calibrating the antenna pattern of a dual-polarization radar, characterized in that, Including the following steps: The measurement data from the dual-polarization radar is preprocessed; the measurement data includes the azimuth, elevation, and echo amplitude of the calibration target obtained by the dual-polarization radar. The azimuth angle, the pitch angle, and the echo amplitude are combined to form three-dimensional data. The three-dimensional data is sliced ​​to obtain multiple sets of two-dimensional data of the echo amplitude corresponding to the pitch angle i as the azimuth angle changes; i = 1, 2, ..., N, where N is the total number of pitch angles. The nonlinear least squares method is used to fit the two-dimensional data corresponding to the N pitch angles using a Gaussian function to obtain the corresponding N sets of fitting parameters. Based on a set of fitted azimuth angles selected within a specific azimuth angle range and N sets of fitted parameters, N sets of fitted echo amplitudes are generated respectively; a set of fitted azimuth angles includes multiple fitted azimuth angles, and the specific azimuth angle range is the azimuth angle range of the echo amplitude set; The N sets of fitted echo amplitudes are respectively matched with one set of fitted azimuth angles to form N sets of fitted two-dimensional data. The N sets of fitted two-dimensional data are deduplicated to obtain each set of refined two-dimensional data. Each set of refined two-dimensional data and the corresponding pitch angle are combined into refined three-dimensional data. The refined three-dimensional data are then smoothed by biharmonic spline interpolation to obtain reconstructed three-dimensional data. The calibration factor of the measured reflectivity factor is calculated based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar. The antenna pattern corresponding to the reconstructed three-dimensional data is calibrated using the calibration factor to obtain the calibrated antenna pattern.

2. The antenna pattern calibration method for dual-polarization radar according to claim 1, characterized in that, The steps for preprocessing measurement data from dual-polarization radar include: Based on the scanning data of the calibration target by the dual-polarized radar and the data provided by the calibration system, a range Doppler image is plotted. After filtering out the interference points caused by the carrier of the calibration target in the range Doppler image, the measurement data corresponding to the calibration target is extracted. The measurement data is initially invalidated and replaced with default values ​​for each echo amplitude outlier to obtain the preprocessed measurement data.

3. The antenna pattern calibration method for dual-polarization radar according to claim 2, characterized in that, The process of slicing the three-dimensional data includes: Store the positions of each of the default values ​​into the first array; Iterate through each position in the first array. If the position is not adjacent to the next position, form a position pair with the next position and store each position pair in the second array. The position of the first echo amplitude and the position preceding the first default value are combined to form a first-end position pair, and the first-end position pair is stored in the second array; The end position pair is determined based on the value of the last echo amplitude and stored in the second array; Delete the position pairs in the second array that correspond to the azimuth angles that are not within the specified azimuth angle range; By using difference filtering, the overfitted position pairs in the second array are removed.

4. The antenna pattern calibration method for dual-polarization radar according to claim 3, characterized in that, The process of determining the end position pair based on the last echo amplitude value includes: If the last echo amplitude value is the default value, then the position pair corresponding to the last echo amplitude value is determined as the end position pair.

5. The antenna pattern calibration method for dual-polarization radar according to claim 3, characterized in that, The process of determining the end position pair based on the last echo amplitude value includes: If the value of the last echo amplitude is not the default value, then the position after the last default value and the position of the last echo amplitude are combined to form the end position pair.

6. The antenna pattern calibration method for a dual-polarization radar according to claim 4 or 5, characterized in that, The fitted azimuth angle values ​​selected within the specific azimuth angle range form an arithmetic sequence.

7. The antenna pattern calibration method for dual-polarization radar according to claim 6, characterized in that, The process of removing duplicates from the N sets of fitted two-dimensional data to obtain the refined two-dimensional data for each set includes: Traverse the N groups of fitted two-dimensional data, and select the fitted two-dimensional data with the same pitch angle or a pitch angle difference less than a set difference between the groups to obtain each similar fitted two-dimensional dataset. For each of the similar fitting two-dimensional datasets, calculate the correlation coefficient between each set of fitted two-dimensional data in the similar fitting two-dimensional dataset and the two adjacent sets of fitted two-dimensional data; Each fitted two-dimensional dataset retains the set of fitted two-dimensional data with the highest correlation coefficient to obtain refined two-dimensional data.

8. An antenna pattern calibration device for a dual-polarization radar, characterized in that, include: The data preprocessing module is used to preprocess the measurement data of the dual-polarization radar; the measurement data includes the azimuth angle, elevation angle and echo amplitude of the calibration target obtained by the dual-polarization radar. The data slicing module is used to combine the azimuth angle, the pitch angle and the echo amplitude into three-dimensional data, slice the three-dimensional data to obtain multiple sets of two-dimensional data of the echo amplitude corresponding to the pitch angle i as the azimuth angle changes; i = 1, 2, ..., N, where N is the total number of pitch angles; The data fitting module is used to fit the two-dimensional data corresponding to the N pitch angles respectively using the nonlinear least squares method and the Gaussian function to obtain the corresponding N sets of fitting parameters. The new data generation module is used to generate N sets of fitted echo amplitudes based on a set of fitted azimuth angles selected within a specific azimuth angle range and N sets of fitted parameters; a set of fitted azimuth angles includes multiple fitted azimuth angles, and the specific azimuth angle range is the azimuth angle range of the echo amplitude set. The data deduplication module is used to combine the N sets of fitted echo amplitudes with a set of fitted azimuth angles to form N sets of fitted two-dimensional data, and to deduplicat the N sets of fitted two-dimensional data to obtain each set of refined two-dimensional data. The data interpolation module is used to merge each set of refined two-dimensional data and the corresponding pitch angle into refined three-dimensional data, and to perform smooth interpolation on the refined three-dimensional data using the biharmonic spline interpolation method to obtain reconstructed three-dimensional data. The data calibration module is used to calculate the calibration factor of the measured reflectivity factor based on the theoretical reflectivity factor of the calibration target and the measured reflectivity factor of the dual-polarized radar, and to calibrate the antenna pattern corresponding to the reconstructed three-dimensional data using the calibration factor to obtain the calibrated antenna pattern.

9. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the antenna pattern calibration method for dual-polarized radar according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the antenna pattern calibration method for dual-polarized radar according to any one of claims 1 to 7.

Images