Radar calibration method, device and system and storage medium

Abstract

The invention discloses a radar calibration method, device and system and a storage medium, and belongs to the field of radar antennas. The method comprises the following steps: acquiring first echo data received by each virtual channel of a radar system at each to-be-calibrated first angle; determining the current angle deviation of the turntable system; correcting each first angle based on the current angle deviation to obtain a plurality of second angles; and determining a first correction matrix based on the first steering vector and the first echo data of each virtual channel at each second angle, so as to correct the first echo data through the first correction matrix. According to the method, the current angle deviation of the turntable system can be determined, the to-be-calibrated angles can be corrected, and then the correction matrix is generated based on the corrected angles, so that the influence of correction effect deterioration caused by the angle deviation of the turntable system can be greatly relieved; and the calibration effect of the radar system can still be ensured when the angle of the turntable system has a precision problem.

Description

Technical Field

[0001] This application relates to the field of radar antenna technology, specifically to a radar calibration method, apparatus, system, and storage medium. Background Technology

[0002] Before a radar system is put into formal use, it is usually necessary to calibrate the initial phase and amplitude differences of the transceiver channels to improve the accuracy of subsequent radar angle measurements. The calibration of the radar system is mainly carried out by a calibration system, which mainly includes a microwave anechoic chamber, a turntable system, and a target simulation device. Its core is to detect and compensate for the initial phase and amplitude differences of the radar transceiver channels by creating an environment that simulates the target position and angle realistically and accurately.

[0003] In practice, as the usage time and frequency gradually increase, the accuracy of the calibration system will gradually decrease. For example, the turntable system will gradually develop angular deviations, making it impossible to accurately complete the radar calibration work, which in turn affects the calibration effect of the radar system. Summary of the Invention

[0004] Embodiments of this application provide a radar calibration method, apparatus, system, and storage medium to ensure the calibration effect of the radar system even when the calibration system has accuracy problems.

[0005] To address the aforementioned technical problems, embodiments of this application disclose the following technical solutions:

[0006] Firstly, a radar calibration method is provided, the method comprising:

[0007] Acquire the first echo data received by each virtual channel of the radar system at each first angle to be calibrated;

[0008] Determine the current angular deviation of the turntable system;

[0009] Based on the current angle deviation, each of the first angles is corrected to obtain multiple second angles;

[0010] Based on the first steering vector of each virtual channel at each second angle and the first echo data, a first correction matrix is ​​determined to correct the first echo data using the first correction matrix.

[0011] Secondly, a radar calibration device is provided, the device comprising:

[0012] The data acquisition unit is used to acquire the first echo data received by each virtual channel of the radar system at each first angle to be calibrated.

[0013] Deviation estimation unit is used to determine the current angular deviation of the turntable system;

[0014] An angle correction unit is used to correct each of the first angles based on the current angle deviation to obtain a plurality of second angles;

[0015] A matrix generation unit is configured to determine a first correction matrix based on the first steering vector of each virtual channel at each second angle and the first echo data, so as to correct the first echo data by means of the first correction matrix.

[0016] Thirdly, a radar calibration device is provided, comprising:

[0017] Memory, used to store programs;

[0018] A processor for executing a program stored in the memory;

[0019] When the program stored in the memory is executed, the processor performs the method as described in any of the first aspects.

[0020] Fourthly, a radar calibration system is provided, the system comprising:

[0021] Turntable system, the turntable system being used to fix the radar system;

[0022] A target simulation device, wherein the target simulation device is used to simulate a target point;

[0023] A controller connected to the turntable system, the controller being configured to drive the turntable system to rotate to each first angle to be calibrated, and to perform the method as described in any of the first aspects.

[0024] Fifthly, a computer-readable storage medium is provided that stores instructions for execution by a computing device, wherein the computing device, when executing the instructions, implements the method as described in any of the first aspects.

[0025] One of the above technical solutions has the following advantages or beneficial effects:

[0026] Compared with existing technologies, the radar calibration method of this application includes: acquiring first echo data received by each virtual channel of the radar system at each first angle to be calibrated; determining the current angle deviation of the turntable system; correcting each first angle based on the current angle deviation to obtain multiple second angles; and determining a first correction matrix based on the first steering vector and first echo data of each virtual channel at each second angle, so as to correct the first echo data through the first correction matrix. The method provided by this application can determine the current angle deviation of the turntable system and correct each angle to be calibrated, and then generate a correction matrix based on the corrected angles, thereby greatly mitigating the impact of the deterioration of the correction effect caused by the angle deviation of the turntable system, and still ensuring the calibration effect of the radar system when the angle of the turntable system has accuracy problems.

[0027] The radar calibration device disclosed in this application can greatly mitigate the impact of the deterioration of the calibration effect caused by the angular deviation of the turntable system, and can still ensure the calibration effect of the radar system when the angular accuracy of the turntable system is affected.

[0028] The radar calibration system disclosed in this application can still accurately calibrate the radar system even when the angle of the turntable system has accuracy problems, thereby reducing or even avoiding the impact of the angle deviation of the turntable system on the calibration effect of the radar system.

[0029] The present application provides a computer-readable storage medium that can greatly mitigate the impact of the deterioration of the calibration effect caused by the angular deviation of the turntable system, and can still ensure the calibration effect of the radar system when the angular accuracy of the turntable system is affected. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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 schematic diagram of the radar calibration system provided in the embodiments of this application;

[0032] Figure 2 This is a partial structural schematic diagram of the radar system provided in the embodiments of this application;

[0033] Figure 3 This is a schematic diagram of the overall process of the radar calibration method according to an embodiment of this application;

[0034] Figure 4 This is a schematic diagram of the antenna position according to an embodiment of this application;

[0035] Figure 5 yes Figure 3 A schematic diagram illustrating the specific execution process of step 302;

[0036] Figure 6 This is a schematic diagram illustrating an example of sampling points where the correlation between the second echo data and the second steering vector is higher than 0.99 in Case 1 of this application embodiment;

[0037] Figure 7 This is a schematic diagram illustrating an example of sampling points where the correlation between the second echo data and the second steering vector is higher than 0.99 in Case 2 of this application.

[0038] Figure 8 This is a schematic diagram illustrating an example of sampling points where the correlation between the second echo data and the second steering vector is higher than 0.99 in Case 3 of this application.

[0039] Figure 9 This is a schematic diagram of the software structure of the radar calibration device according to an embodiment of this application;

[0040] Figure 10 This is a schematic diagram of the hardware structure of the radar calibration device according to an embodiment of this application;

[0041] Figure label:

[0042] 10-Turntable system; 20-Target simulation device; 30-Controller; 40-Radar system; 41-Antenna array; 411-Transmitting antenna array; 412-Receiving antenna array; 901-Data acquisition unit; 902-Bias estimation unit; 903-Angle correction unit; 904-Matrix generation unit; 1001-Memory; 1002-Processor. Detailed Implementation

[0043] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0044] In the description of this application, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of this application, "multiple" means two or more, and "at least one" can mean one, two, or more, unless otherwise explicitly specified.

[0045] Please see Figure 1 , Figure 1 This is a schematic diagram of the radar calibration system provided in this application embodiment. The radar calibration system includes a turntable system 10, a target simulation device 20, and a controller 30, with the controller 30 connected to the turntable system 10. The turntable system 10 is used to fix the radar system 40. The turntable system 10 includes three mutually orthogonal axes, namely the α-axis, β-axis, and γ-axis, each moving along its corresponding rotation center. When the radar system 40 is mounted at the rotation center of the turntable system 10, the controller 30 can control the turntable system 10 to rotate the radar system 40, so that the target simulation device 20 is located at various first angles to be calibrated by the radar system 20. When the turntable system 10 rotates to any first angle, the radar system 40 collects echo data at the current first angle and sends it to the controller 30 as the basic information for calibration. The target simulation device 20 is used to simulate target points and may specifically include a corner reflector with a fixed RCS (Radar Cross Section) or a radar target simulator with adjustable velocity and distance. For example, the corner reflector may be a corner reflector with an RCS of 22 dBsm. The target simulation device 20 can be at the same height as the radar system 40 located on the turntable system 10. For example, with the initial angular deviation of all three axes of the turntable system 10 at 0°, the target simulation device 20 can be positioned 7.2m away from the turntable system 10 in the first positive X direction. The controller 30 is also used to calculate the correction matrix of the radar system 40 based on the acquired echo data to complete the calibration of the radar system 40. Furthermore, this radar calibration system is located in a microwave anechoic chamber, which is primarily a sealed space filled with pyramidal-shaped absorbing material, providing a clean environment for the radar system 40. For example, the microwave anechoic chamber can be an anechoic chamber that meets the requirements for E-band far-field range.

[0046] Please see Figure 2 , Figure 2 This is a partial structural schematic diagram of the radar system provided in this application embodiment. The radar system 40 mainly detects and tracks targets within its detection range using an antenna array 41. The antenna array 41 includes a transmitting antenna array 411 and a receiving antenna array 412. The transmitting antenna array 411 includes Nt transmitting antennas TX (Transmit), and the receiving antenna array 412 includes Nr receiving antennas RX (Receive) with a spacing of d. Nt is a positive integer greater than or equal to 1, and Nr is a positive integer greater than or equal to 2. The Nt transmitting antennas TX and the Nr receiving antennas RX can form Nt×Nr virtual channels. The transmitting antenna array 411 is used to transmit electromagnetic wave signals S, and the receiving antenna array 412 is used to receive echo signals.

[0047] Through the above scheme, the radar calibration system can create an environment with a realistic and accurate target position and angle, detect and compensate for the initial phase and amplitude differences of the radar transceiver channel, so that the radar angle measurement algorithm can be performed accurately.

[0048] However, in actual operation, radar calibration results occasionally become unsatisfactory. Aside from issues inherent to the radar system itself, a reality is that the accuracy of the radar calibration system gradually decreases with increasing usage time and frequency; for example, the turntable system 10 may gradually develop angular deviations. When encountering such problems, the conventional solution is often to use a more accurate system to calibrate the original radar calibration system. Specifically, this could involve seeking a professional organization to recalibrate the radar system. This method is not only time-consuming and labor-intensive, significantly disrupting production schedules during large-scale production tasks, but also, before recalibration, the radar calibration system can no longer accurately complete the calibration work, thus affecting the calibration results.

[0049] In view of this, the present application provides a radar calibration method that introduces a misalignment compensation function for the decrease in accuracy of the turntable system during the analysis of radar calibration data. This ensures the calibration effect of the radar system even when the calibration system has accuracy problems, thereby solving at least part of the above-mentioned technical problems.

[0050] Please see Figure 3 , Figure 3 This is a schematic flowchart of the radar calibration method according to an embodiment of this application. The radar calibration method can be used in the controller 30 of a radar calibration system, and specifically includes the following steps:

[0051] Step 301: Acquire the first echo data received by each virtual channel of the radar system at each first angle to be calibrated.

[0052] Specifically, the various first angles to be calibrated can be represented by a list of angles to be calibrated. Each first angle can be set according to the azimuth range, elevation range, and sampling interval of the radar system.

[0053] For example, the FOV (Field of View) of the radar system's azimuth plane is [-75°, +75°], and the FOV of the elevation plane is [-12°, +12°]. The sampling interval is set to 1°, the azimuth range is [-90°, +90°], and the corresponding elevation angles for each azimuth are -6°, 0°, and +6°, respectively. Therefore, the total number of first angles in the list of angles to be calibrated is M = 181 × 3 = 543, that is, the total number of sampling points is M = 543. It should be noted that, for ease of description, the first angle described in subsequent embodiments can represent the sampling point corresponding to the first angle.

[0054] The first echo data is used to represent the echo signal received by each virtual channel at each first angle. For example, if the number of transmitting channels of the radar system is Nt, the number of receiving channels is Nr, and the total number of virtual channels is N = Nt × Nr, then the first echo data is an N × M matrix, which can be specifically represented by the following formula (1):

[0055]

[0056] In formula (1), X represents the first echo data. For a specific angle and a specific virtual channel, the beam vector is... For the first virtual channel under the first angle θ1, Let θM be the beam vector of the first virtual channel at the first first angle. For the first virtual channel with angle θ1, the beam vector is... Let θM be the beam vector under the first angle θM of the Mth virtual channel.

[0057] According to the classical signal model, the beam vector can be specifically represented by the following formula (2):

[0058]

[0059] In formula (2), A(θ) is the beam vector under a specific angle and a specific virtual channel, and A(θ) is the guide vector. For signal source strength, It is noise.

[0060] The steering vector, which is also an N×M matrix, can be determined based on the first angle to be calibrated and the position of the antenna in the radar system. Specifically, it can be expressed by the following formula (3):

[0061]

[0062] In formula (3), A is the guide vector, and N is the total number of virtual channels. For a specific angle and a specific virtual channel, the guide vector. The first virtual channel is the guide vector at the first angle θ1. The guide vector under the first virtual channel at the first first angle θM of the Mth time. The guide vector under the first angle θ1 and the Nth virtual channel. Let θM be the guide vector under the first angle θM of the Mth virtual channel.

[0063] Please see Figure 4 , Figure 4 This is a schematic diagram of the antenna positions according to an embodiment of this application. A planar coordinate system is established on the plane where the antenna array 41 is located, where the direction parallel to the ground is defined as the Az direction, and the direction perpendicular to the ground is defined as the E1 direction. The coordinates of each antenna in the transmitting antenna TX and the receiving antenna RX in the Az direction in this planar coordinate system are denoted as d. az The coordinates in the El direction are denoted as d. el .

[0064] By combining the antenna positions and the initial angles to be calibrated, the corresponding steering vectors can be determined. Specifically, the steering vector under a specific virtual channel at a specific angle. It can be determined by the following formula (4):

[0065]

[0066] In formula (4), az is the azimuth angle corresponding to the first angle, el is the pitch angle corresponding to the first angle, and d az Let d be the coordinate of the antenna in the Az direction. el Let be the coordinates of the antenna in the E1 direction.

[0067] Step 302: Determine the current angle deviation of the turntable system.

[0068] Specifically, the current angle deviation is used to characterize the deviation between the current angle of the turntable system and the reference angle. For example, if the reference angle of the three axes of the turntable system is 0°, and the reference angle of any axis is not 0°, then the whole system will have an angle deviation.

[0069] In some embodiments, the current angular deviation of the turntable system can be determined in the following ways:

[0070] Step 1: Search the three-axis angle deviations of the turntable system sequentially according to the preset search order, and determine the three-axis angle deviation corresponding to the maximum correlation between the second guide vector and the second echo data.

[0071] The second guide vector is determined based on each third angle, which is obtained by correcting the first angle based on the three-axis angle deviation. The second echo data is obtained by correcting the first echo data using the second correction matrix, which is determined based on the second guide vector and the first echo data.

[0072] In some examples, the steps of searching for the three-axis angular deviations of the turntable system in a preset search order can be specifically performed as follows:

[0073] The first step is to determine the initial angular deviation of each axis of the turntable system.

[0074] The initial angle deviation of each axis is the three-axis angle deviation at the end of the previous search.

[0075] Specifically, the initial angle deviations of the three axes of the turntable system can be represented as α0, β0, and γ0. The initial angle deviations during each search are the same as the angle deviations of the three axes at the end of the previous search, and so on, continuously iterating and updating the initial angle deviations of each axis.

[0076] Understandably, in the initial state of the turntable system, there is no initial angular deviation among the three axes, so α0 = 0°, β0 = 0°, and γ0 = 0°.

[0077] The second step is to determine the updated angle deviation for each axis based on the initial angle deviation of each axis, the search step size of each axis, and the preset search order.

[0078] Specifically, the search step size for each axis represents the increment of the angle deviation for each axis during each search, and the preset search order represents the search sequence of the three axes during the search process. The preset search range for each axis can be lim α lim β lim γ Then the update angle deviations of the three axes are represented as α1, β1, and γ1, respectively. 1＝ α0+Δα,β 1＝ β0+Δβ,γ 1＝ γ0+Δγ, where the angular deviation increment Δα∈lim α ,Δβ∈lim β ,Δγ∈lim γ The preset search scope, preset search count, and search order are not specifically limited in the embodiments of this application.

[0079] For example, the preset search order is α-axis, β-axis, γ-axis, α-axis, β-axis, γ-axis... in a loop, with a preset search range of 5° and angle deviation increments Δα, Δβ, and Δγ all of 0.1°. In the first search, the initial angle deviation for each axis is 0°, so the updated angle deviation α1 for the α-axis is 0.1°, the updated angle deviation β1 for the β-axis is 0°, and the updated angle deviation γ1 for the γ-axis is 0°. Correspondingly, in the second search, the initial angle deviations for the three axes are 0.1°, 0°, and 0°, respectively, and the updated angle deviation α1 for the α-axis is 0.1°, the updated angle deviation β1 for the β-axis is 0.1°, and the updated angle deviation γ1 for the γ-axis is 0°, and so on. Further details are omitted.

[0080] It is understandable that the preset search range can represent absolute values, and the update angle deviation of the three axes can be in different directions. That is to say, in the actual search process, you can search 5° in the forward direction and then search 5° in the reverse direction, or you can start from -5° and search sequentially to +5°. The specific search is performed according to the set search order, and no specific limitation is made here.

[0081] The third step is to determine the updated angle deviations of the three axes as the three-axis angle deviations for the current search.

[0082] Understandably, due to different search orders and search ranges, the update angle deviation of the three axes can have multiple combinations, and each combination will produce a corresponding three-axis angle deviation. The specific search process is quite flexible.

[0083] In some examples, the steps to determine the triaxial angular deviation corresponding to the maximum correlation between the second steering vector and the second echo data can be performed in the following ways:

[0084] The first step is to correct each first angle based on the three-axis angle deviation for any three-axis angle deviation within the preset search range, thereby obtaining multiple third angles.

[0085] For example, a subset of first angles can be selected from multiple first angles for subsequent correlation calculations. For instance, the first angles within the radar system's FOV range can be selected, i.e., from the aforementioned 543 sampling points, the angle range within the FOV [-75°, +75°] of the azimuth plane can be selected, with each azimuth angle corresponding to an elevation angle of -6°, 0°, and +6°, respectively. Then, the total number of selected first angles M0 = 151 × 3 = 453.

[0086] Specifically, for each selected first angle θ, the corresponding rotation angle list of the turntable system can be represented as [α M0 β M0 γ M0], where M0 is the total number of selected first angles. The rotation angle list of the turntable system after three-axis angle deviation correction can be updated to [α1 M0 β1 M0 γ1 M0 Since the coordinate system of the turntable system differs from that of the radar system, a coordinate system transformation is required during angle correction. The radar system's coordinate system is typically a spatial XYZ coordinate system. Specifically, the rotation transformation from the initial azimuth and elevation angles corresponding to the first angle to the final azimuth and elevation angles corresponding to the third angle can be calculated using the following formula (5):

[0087]

[0088] In formula (5), R αβγ To transform the three axes of the turntable system into rotation matrices in the XYZ coordinate system, Az0 M0 and El0 M0 These represent the initial azimuth and elevation angles corresponding to each first angle, respectively. Az1 M0 and El1 M0 These represent the final azimuth and elevation angles corresponding to each third angle.

[0089] The second step is to determine the second correction matrix based on the second guide vector and the first echo data of each virtual channel under each third angle, and then correct the first echo data using the second correction matrix to determine the second echo data.

[0090] Specifically, based on each third angle and the position of the antenna in the radar system, the second steering vector A2 can be obtained by combining formulas (3) and (4).

[0091] The second correction matrix C2 is determined based on the second steering vector A2 and the first echo data X1. First, the target expression can be constructed based on the second steering vector A2, the first echo data X1, the second correction matrix C2, and the amplitude and phase change data Z3 of each virtual channel at each third angle. Specifically, it is represented by the following formula (6):

[0092] C2X1=A2Z3(6);

[0093] Then, the objective expression is solved to obtain the second correction matrix C2. Specifically, under certain norms, the minimum value of C2X1-A2Z3 among all possible C2 and Z3 is found. That is, C2 and Z3 are iteratively solved to ensure that ‖C2X1-A2Z3‖≤ε, where ε is the threshold for stopping the loop. The specific solution method can be a Kalman loop or the method described in the embodiments of this application, where the solution method of the embodiments of this application will be described in subsequent embodiments.

[0094] After obtaining the second correction matrix C2, the first echo data X1 is corrected using the second correction matrix C2, and the second echo data X2 = C2X1 is determined.

[0095] The third step is to determine the correlation between the second echo data and the second steering vector.

[0096] In some examples, the correlation between the second echo data and the second steering vector can be specifically determined through the following steps:

[0097] First, based on the second echo data and the second steering vector, the correlation corresponding to each third angle is determined.

[0098] Then, the mean of the correlations corresponding to all third angles is determined as the correlation between the second echo data and the second steering vector.

[0099] Specifically, the correlation corresponding to each third angle can be determined using the following formula (7):

[0100]

[0101] In formula (7), Cor i Let represent the correlation corresponding to the i-th third angle, where i is an integer greater than or equal to 1 and less than or equal to M, M is the total number of third angles, and N is the total number of virtual channels. Let A be the conjugate transpose of the vector corresponding to the i-th third angle in the second echo data, and let A be the vector corresponding to the i-th third angle in the second steering vector.

[0102] After determining the correlation corresponding to each third angle, the sum of the correlations corresponding to all third angles and the quotient of the number of third angles are determined as the mean of the correlations corresponding to all third angles, which is the correlation between the second echo data and the second steering vector.

[0103] Understandably, correlation can be used to evaluate the correction effect, with an upper limit of 1. The higher the correlation between the second echo data and the second steering vector, the better the correction effect.

[0104] The above scheme uses the mean of the correlations corresponding to all third angles to characterize the correlation between the second echo data and the second steering vector, which has a more accurate effect and facilitates subsequent calculations.

[0105] The fourth step is to determine the most relevant three-axis angle deviation from all three-axis angle deviations within the preset search range.

[0106] In some examples, the most relevant triaxial angle deviations can be determined using the following steps:

[0107] When searching in a preset order, if the difference between the correlation of any three-axis angle deviation within the preset search range and the correlation of the previous three-axis angle deviation is less than a preset threshold, then the three-axis angle deviation is determined as the three-axis angle deviation with the highest correlation.

[0108] For example, the preset threshold can be set to 0.1.

[0109] In other words, the condition for determining whether the iterative calculation has converged is achieved by judging whether the change in the correlation corresponding to the three-axis angle deviation between two consecutive searches is less than a preset threshold. If the change in the correlation corresponding to the three-axis angle deviation between two consecutive searches is less than the preset threshold, the iterative optimization process is considered to have converged, and the optimization can be stopped and the current three-axis angle deviation can be output. This reduces the amount of computation data, improves processing speed, and thus outputs the required three-axis angle deviation more quickly.

[0110] Step 2: Determine the current angle deviation as the three-axis angle deviation corresponding to the maximum correlation between the second guide vector and the second echo data.

[0111] To more clearly illustrate the execution process of step 302, please refer to [link / reference]. Figure 5 , Figure 5 yes Figure 3 The following is a schematic diagram illustrating the specific execution process of step 302. In determining the current angle deviation, a preset search range can be set first. The list of angles to be calibrated is updated based on the three-axis angle deviation of the current search. A second steering vector A is generated from the updated angle list. A second correction matrix C is generated based on the second steering vector A and the first echo data X. The first echo data X is then corrected using the second correction matrix C, and the correlation Cor between the second echo data X2 and the second steering vector A is calculated. Finally, the mean correlation Cor of all points within the radar system's FOV range is calculated. mean It compares the current angle deviation with the mean correlation calculated in the previous search to determine whether it has converged. If it has converged, it outputs the current angle deviation. If it has not converged, it updates the three-axis angle deviation and performs the next search.

[0112] The above method can accurately and quickly estimate the current angle deviation, thereby accurately correcting the first angle and improving the generation accuracy of the correction matrix.

[0113] Step 303: Correct each first angle based on the current angle deviation to obtain multiple second angles.

[0114] Specifically, for each first angle θ, the corresponding rotation angle list of the turntable system can be represented as [α M β M γ M], where M is the total number of the first angles. The rotation angle list of the turntable system after correction based on the current angle deviation can be updated to [α1 M β1 M γ1 M Based on the description of the relevant content in the foregoing embodiments, the rotation transformation from the initial azimuth and pitch angles corresponding to the first angle to the final azimuth and pitch angles corresponding to the second angle can be calculated using the following formula (8):

[0115]

[0116] In formula (8), R αβγ To transform the three axes of the turntable system into rotation matrices in the XYZ coordinate system, Az0 M and El0 M These represent the initial azimuth and elevation angles corresponding to each first angle, respectively. Az1 M and El1 M These represent the final azimuth and elevation angles corresponding to each of the second angles.

[0117] Step 304: Based on the first steering vector and first echo data of each virtual channel at each second angle, determine the first correction matrix to correct the first echo data.

[0118] In some embodiments, the first correction matrix can be determined in the following manner:

[0119] Step 1: Based on the first steering vector, the first echo data, the first correction matrix, and the amplitude and phase change data of each virtual channel at each second angle, construct the target expression.

[0120] Specifically, the calibration problem is actually a mathematical problem of transforming the beam vector into a steering vector. That is, using a 16*16 complex matrix C, the beam vector can be transformed at each test angle. Convert into guide vector It can be represented by a series of equations as shown in the following formula (9):

[0121]

[0122] The set of equations shown in formula (9) can be represented as a matrix, as shown in formula (10) below:

[0123] CX = AZ(10);

[0124] In formula (10), C represents the first correction matrix, X represents the first echo data, A represents the first steering vector, and Z represents the amplitude and phase change data. Z is a diagonal matrix of size M, representing the amplitude and phase changes of all channels of the radar system at each second angle. The expression for Z is given in formula (11) below:

[0125]

[0126] Solving for the first correction matrix C in formula (10) can be transformed into finding the minimum value of ‖CX-AZ‖ under the Frobenius norm. The solution method used in this application embodiment is to first estimate Z by singular value decomposition, and then estimate the first correction matrix C by least squares method.

[0127] Step 2: Based on the objective expression, the amplitude and phase change data are obtained by solving the singular value decomposition method.

[0128] Specifically, according to formula (10), C = AZX + X + Let X be the pseudo-inverse matrix of the first echo data. Then the problem of finding the minimum value can be expressed as the following formula (12):

[0129]

[0130] In formula (12), It is the orthogonal projection matrix of the first echo data X in the subspace.

[0131] Assumption Convert the matrix into a column vector, that is:

[0132]

[0133] Let z = Vec(Z), in This is the Kronecker product. The problem of Bz = 0 can be solved using Singular Value Decomposition (SVD). From SVD, we know that:

[0134] B=UΣV H (14);

[0135] In formula (14), Σ and V are both M*M matrices. Σ is a diagonal matrix composed of all singular values, specifically represented by the following formula (15):

[0136]

[0137] Let m0 be the row corresponding to the smallest singular value in Σ, and let Z be the column vector of the m0th position in V, as shown in the following formula (16):

[0138] Z = diag(V) m0 (16);

[0139] The above method is relatively simple to solve for the amplitude and phase change data Z, and can improve the overall computational efficiency and accuracy.

[0140] Step 3: Based on the objective expression and the amplitude and phase change data obtained from the solution, the first correction matrix is ​​obtained by solving the least squares method.

[0141] Specifically, after solving for the amplitude and phase change data Z, the first correction matrix C = AZX can be obtained by using the least squares method in conjunction with formula (10). + The solution to the correction matrix is ​​now complete.

[0142] Solving the first correction matrix using the above method can yield results faster and more accurately, thus enabling precise correction of the first echo data. This can better improve the deterioration of radar calibration results caused by the inherent errors of the radar calibration system, making it highly practical.

[0143] The application effect of the radar calibration method of this application embodiment will be described below with reference to the accompanying drawings.

[0144] The current angular deviations of the three axes of the turntable system are set to 5°, 0°, and 0° respectively. Using the angle range of [-75°, +75°] in the azimuth plane of the radar system mentioned in the previous embodiment, and the elevation angles corresponding to each azimuth angle are -6°, 0°, and +6° respectively, and the azimuth angles are sampled at 1° intervals, then the total number of selected first angles M0 = 151 × 3 = 453. The radar system is calibrated for these angle positions within the implied range, based on three cases: the actual angular deviation of the turntable system is unknown, and the angle deviation estimation function of this application embodiment is not enabled (case 1); the actual angular deviation of the turntable system is unknown, and the angle deviation estimation function of this application embodiment is enabled (case 2); the actual angular deviation of the turntable system is known, and there is no ideal result of angular deviation (case 3). The proportion of the second echo data after correction and the second steering vector with a correlation higher than 0.99 in the above three cases is statistically calculated by the following formula (17), and summarized in Table 1:

[0145]

[0146] In formula (17), Cor r99 M1 represents the proportion of first angles with a correlation higher than 0.99, M2 represents the number of first angles with a correlation higher than 0.99, and M0 represents the total number of first angles selected.

[0147] Table 1: The proportion of cases where the correlation between the second echo data and the second steering vector is higher than 0.99 in three different scenarios.

[0148] Scenario 1 Scenario 2 Scenario 3 0.44％ 32.22％ 33.92％

[0149] Please refer to the following: Figures 6 to 8 , Figure 6 This is a schematic diagram illustrating an example of sampling points where the correlation between the second echo data and the second steering vector is higher than 0.99 in Case 1 of this application embodiment. Figure 7 This is a schematic diagram illustrating an example of sampling points where the correlation between the second echo data and the second steering vector is higher than 0.99 in Case 2 of this application. Figure 8 This is a schematic diagram illustrating an example of sampling points where the correlation between the second echo data and the second steering vector is higher than 0.99 in Case 3 of this application. Here, AzimuthAngle is the azimuth angle, Elevation Angle is the elevation angle, BV Correlation (>0.99) after Calibration is the echo data correlation greater than 0.99 after correction, and coverage is the average value.

[0150] Combined with Table 1 and Figures 6 to 8 The radar calibration method of this application can greatly alleviate the impact of the deterioration of the correction effect caused by the turntable angle deviation. At the same time, the radar correction effect after compensation using the radar calibration method of this application is very close to the best correction effect under no deviation. Therefore, the embodiment of this application is a very innovative and practical method.

[0151] It is understood that the method in this application embodiment can determine the current angle deviation of the turntable system and correct each angle to be calibrated, and then generate a calibration matrix based on the corrected angle, thereby greatly mitigating the impact of the deterioration of the calibration effect caused by the angle deviation of the turntable system, and still ensuring the calibration effect of the radar system when the angle of the turntable system has accuracy problems.

[0152] Accordingly, please refer to Figure 9 , Figure 9 This is a schematic diagram of the software structure of a radar calibration device according to an embodiment of this application. The radar calibration device provided in this embodiment includes a data acquisition unit 901, a deviation estimation unit 902, an angle correction unit 903, and a matrix generation unit 904.

[0153] The data acquisition unit 901 is used to acquire the first echo data received by each virtual channel of the radar system at each first angle to be calibrated.

[0154] Deviation estimation unit 902 is used to determine the current angular deviation of the turntable system;

[0155] Angle correction unit 903 is used to correct each first angle based on the current angle deviation to obtain multiple second angles;

[0156] The matrix generation unit 904 is used to determine a first correction matrix based on the first steering vector and the first echo data of each virtual channel at each second angle, so as to correct the first echo data through the first correction matrix.

[0157] In some embodiments, the deviation estimation unit 902 is specifically used for:

[0158] The three-axis angle deviations of the turntable system are searched sequentially according to a preset search order. The three-axis angle deviation corresponding to the maximum correlation between the second guide vector and the second echo data is determined. The second guide vector is determined based on each third angle. The third angle is obtained by correcting the first angle based on the three-axis angle deviation. The second echo data is obtained by correcting the first echo data through the second correction matrix. The second correction matrix is ​​determined based on the second guide vector and the first echo data.

[0159] The triaxial angle deviation corresponding to the maximum correlation between the second guide vector and the second echo data is determined as the current angle deviation.

[0160] In some embodiments, the deviation estimation unit 902 is specifically used for:

[0161] For any three-axis angle deviation within a preset search range, each first angle is corrected based on the three-axis angle deviation to obtain multiple third angles;

[0162] Based on the second steering vector and first echo data of each virtual channel under each third angle, a second correction matrix is ​​determined, and the first echo data is corrected by the second correction matrix to determine the second echo data;

[0163] Determine the correlation between the second echo data and the second steering vector;

[0164] The most relevant three-axis angle deviation is determined from all three-axis angle deviations within the preset search range.

[0165] In some embodiments, the deviation estimation unit 902 is specifically used for:

[0166] When searching in a preset order, if the difference between the correlation of any three-axis angle deviation within the preset search range and the correlation of the previous three-axis angle deviation is less than a preset threshold, then the three-axis angle deviation is determined as the three-axis angle deviation with the highest correlation.

[0167] In some embodiments, the deviation estimation unit 902 is specifically used for:

[0168] Based on the second echo data and the second steering vector, the correlation corresponding to each third angle is determined;

[0169] The mean of the correlations corresponding to all third angles is determined as the correlation between the second echo data and the second steering vector.

[0170] In some embodiments, the deviation estimation unit 902 is specifically used for:

[0171] The correlation for each third angle is determined using the following formula:

[0172]

[0173] Among them, Cor i Let represent the correlation corresponding to the i-th third angle, where i is an integer greater than or equal to 1 and less than or equal to M, M is the total number of third angles, and N is the total number of virtual channels. Let A be the conjugate transpose of the vector corresponding to the i-th third angle in the second echo data, and let A be the vector corresponding to the i-th third angle in the second steering vector.

[0174] In some embodiments, the deviation estimation unit 902 is specifically used for:

[0175] Determine the initial angular deviation of each axis of the turntable system. The initial angular deviation of each axis is the three-axis angular deviation at the end of the previous search.

[0176] The updated angle deviation for each axis is determined based on the initial angle deviation of each axis, the search step size of each axis, and the preset search order.

[0177] The update angle deviations of the three axes are determined as the three-axis angle deviations for the current search.

[0178] In some embodiments, the matrix generation unit 904 is specifically used for:

[0179] Based on the first steering vector, the first echo data, the first correction matrix, and the amplitude and phase change data of each virtual channel at each second angle, a target expression is constructed.

[0180] Based on the objective expression, the amplitude and phase change data are obtained by solving the singular value decomposition method.

[0181] Based on the objective expression and the amplitude and phase change data obtained from the solution, the first correction matrix is ​​obtained by solving using the least squares method.

[0182] It is understood that the radar calibration device of this application embodiment can greatly alleviate the impact of the deterioration of the correction effect caused by the angle deviation of the turntable system, and can still ensure the calibration effect of the radar system when the angle of the turntable system has accuracy problems.

[0183] Accordingly, please refer to Figure 10 , Figure 10 This is a schematic diagram of the hardware structure of a radar calibration device according to an embodiment of this application. This application also provides a radar calibration device, including a memory 1001 and a processor 1002. The memory 1001 is used to store programs. The processor 1002 is used to execute the programs stored in the memory 1001. When the programs stored in the memory 1001 are executed, the processor 1002 executes the radar calibration method of the foregoing embodiments of this application.

[0184] Accordingly, the radar calibration system provided in this application embodiment can still accurately calibrate the radar system even when the angle of the turntable system has accuracy problems, thereby reducing or even avoiding the impact of the angle deviation of the turntable system on the calibration effect of the radar system.

[0185] Accordingly, embodiments of this application also provide a computer-readable storage medium that stores instructions for execution by a computing device, which, when executing the instructions, implements the radar calibration method as described in the foregoing embodiments of this application.

[0186] The radar calibration method, apparatus, system, and storage medium provided in the embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the technical solutions and core ideas of this application. Those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A radar calibration method, characterized by, The method comprises: acquiring first echo data received by each virtual channel of a radar system at each first angle to be calibrated; determining a current angle deviation of a turntable system; correcting each first angle based on the current angle deviation to obtain a plurality of second angles; determining a first correction matrix based on a first steering vector of each virtual channel and the first echo data at each second angle, so as to correct the first echo data by using the first correction matrix.

2. The radar calibration method of claim 1, wherein, The determination of the current angle deviation of the turntable system comprises: sequentially searching the three-axis angle deviations of the turntable system according to a preset search order, and determining a three-axis angle deviation corresponding to a maximum correlation between a second steering vector and second echo data, the second steering vector being determined based on each third angle, the third angle being obtained by correcting the first angle based on the three-axis angle deviation, and the second echo data being obtained by correcting the first echo data by using a second correction matrix, the second correction matrix being determined based on the second steering vector and the first echo data; determining the three-axis angle deviation corresponding to the maximum correlation between the second steering vector and the second echo data as the current angle deviation.

3. The radar calibration method of claim 2, wherein, The determination of the three-axis angle deviation corresponding to the maximum correlation between the second steering vector and the second echo data comprises: for any one of the three-axis angle deviations in a preset search range, correcting each first angle based on the three-axis angle deviation to obtain a plurality of third angles; determining a second correction matrix based on the second steering vector of each virtual channel and the first echo data at each third angle, and correcting the first echo data by using the second correction matrix to determine the second echo data; determining the correlation between the second echo data and the second steering vector; determining a three-axis angle deviation corresponding to the maximum correlation from all the three-axis angle deviations in the preset search range.

4. The radar calibration method of claim 3, wherein, The determination of the three-axis angle deviation corresponding to the maximum correlation from all the three-axis angle deviations in the preset search range comprises: when searching according to the preset order, if a difference between the correlation corresponding to any one of the three-axis angle deviations in the preset search range and the correlation corresponding to a previous one of the three-axis angle deviations is less than a preset threshold, determining the three-axis angle deviation as the three-axis angle deviation corresponding to the maximum correlation.

5. The radar calibration method of claim 3, wherein, The determination of the correlation between the second echo data and the second steering vector comprises: determining the correlation corresponding to each third angle based on the second echo data and the second steering vector; determining a mean value of the correlations corresponding to all the third angles as the correlation between the second echo data and the second steering vector.

6. The radar calibration method of claim 5, wherein, The determination of the correlation corresponding to each third angle based on the second echo data and the second steering vector comprises: the correlation corresponding to each third angle is determined by the following formula: wherein Cor i is a correlation corresponding to the i-th third angle, i is an integer greater than or equal to 1 and less than or equal to M, M is a total number of the third angles, N is a total number of the virtual channels, X2 H is a conjugate transpose of a vector corresponding to the i-th third angle in the second echo data, and A is a vector corresponding to the i-th third angle in the second steering vector.

7. The radar calibration method of claim 2, wherein, The sequential search of the three-axis angle deviations of the turntable system according to the preset search order comprises: determining an initial angle deviation of each axis of the turntable system, the initial angle deviation of each axis being a three-axis angle deviation at the end of a previous search; determining an updated angle deviation of each axis based on the initial angle deviation of each axis, a search step of each axis, and the preset search sequence; determining the updated angle deviation of the three axes as a three-axis angle deviation of a current search.

8. The radar calibration method of claim 1, wherein, determining a first correction matrix based on a first steering vector of each of the virtual channels at each of the second angles and the first echo data, including: constructing a target expression based on the first steering vector, the first echo data, the first correction matrix, and amplitude and phase variation data of each of the virtual channels at each of the second angles; solving the amplitude and phase variation data by singular value decomposition based on the target expression; solving the first correction matrix by least square method based on the target expression and the solved amplitude and phase variation data.

9. A radar calibration device, characterized by The device includes: a data acquisition unit configured to acquire first echo data received by each virtual channel of a radar system at each first angle to be calibrated; a deviation estimation unit configured to determine a current angle deviation of a turntable system; an angle correction unit configured to correct each of the first angles based on the current angle deviation to obtain a plurality of second angles; a matrix generation unit configured to determine a first correction matrix based on a first steering vector of each of the virtual channels at each of the second angles and the first echo data, so as to correct the first echo data by the first correction matrix.

10. A radar calibration device, characterized by including: a memory configured to store a program; a processor configured to execute the program stored in the memory; when the program stored in the memory is executed, the processor executes the method of any one of claims 1-8.

11. A radar calibration system, characterized by The system includes: a turntable system configured to fix a radar system; a target simulation device configured to simulate a target point; a controller connected with the turntable system, the controller being configured to drive the turntable system to rotate to each first angle to be calibrated, and execute the method of any one of claims 1-8.

12. A computer-readable storage medium, characterized in that, The computer readable medium stores instructions for a computing device to execute, and the computing device executes the instructions to implement the method of any one of claims 1-8.

Images