Radar channel calibration method and device and storage medium

Abstract

The invention provides a radar channel calibration method and device and a storage medium, and aims to form a second array steering vector matrix by establishing a signal model which has no channel amplitude-phase error and simultaneously has an azimuth error, a pitch error and a roll error. And constructing a nonlinear least square optimization problem according to the relationship among the first array steering vector matrix, the amplitude-phase calibration coefficient of the multiple channels of the radar and the second array steering vector matrix, and minimizing and solving related parameters based on constraint conditions. Therefore, influences caused by radar azimuth errors, pitching errors and rolling errors in the radar darkroom calibration process can be completely eliminated. Decoupling is realized by the estimated multi-channel amplitude-phase calibration coefficient, the azimuth error, the pitching error and the roll error, namely, the calibration coefficient can accurately reflect the channel amplitude-phase error.

Description

Technical Field

[0001] The present invention relates to the field of radar technology, and in particular to a radar channel calibration method, device and storage medium. Background Art

[0002] Due to the influence of factors such as antenna processing errors, feeder line length errors, and antenna mutual coupling, the radar's multi-channels have inconsistent amplitudes and phases, resulting in errors in the radar's direction-finding results, or even direction-finding errors. Therefore, in order to realize the radar's direction-finding function, it is necessary to calibrate the radar's multi-channels for amplitude and phase. Amplitude and phase calibration needs to be performed in a darkroom. By placing angular invertors at multiple known angles and collecting radar data at the same time, the amplitude and phase calibration coefficients of each channel can be estimated.

[0003] Before darkroom calibration, it is necessary to ensure that when the servo carrying the radar is at azimuth 0 and pitch 0 degrees, the target angle is at least on the normal line of the radar, and the radar itself does not roll. In common technology, when the servo is at azimuth 0 and pitch 0 degrees, the azimuth and pitch offsets of the target angle relative to the radar normal are called azimuth error and pitch error respectively, and the roll offset is called roll error. If the radar has azimuth error, pitch error or roll error during the darkroom calibration process, the radar will have a direction finding error in actual applications. In order to reduce the azimuth error and pitch error, the method of laser irradiation plus mirror reflection is generally used; in order to reduce the roll error, tools such as a bubble level are required.

[0004] On the one hand, such methods require manual adjustment and cannot guarantee that the error is completely eliminated; on the other hand, such adjustments require delicate operations, which will take a lot of time and cost, and it is difficult to meet the requirements of large-scale mass production. If the azimuth pitch error and roll error cannot be completely eliminated, the obtained channel amplitude and phase calibration coefficients will couple these errors and cannot truly reflect the amplitude and phase errors of the channel, resulting in the calibrated radar having direction-finding errors caused by azimuth pitch error and roll error. Summary of the invention

[0005] The present invention provides a radar channel calibration method, device and storage medium, aiming to effectively solve the technical problems in the prior art that radar channel calibration is difficult and the calibrated radar has direction finding errors caused by azimuth pitch errors and roll errors.

[0006] According to one aspect of the present invention, the present invention provides a radar channel calibration method, wherein the radar comprises at least one transmitting antenna and at least one receiving antenna, wherein the at least one transmitting antenna and the at least one receiving antenna constitute a plurality of channels, and the method comprises:

[0007] Step S1: Drive the servo carrying the radar in the darkroom to rotate, so that the target corner reflector is respectively located at the positions of multiple azimuth and elevation angles of the radar. The radar respectively transmits radar test signals at each of the azimuth and elevation angles, and for each of the azimuth and elevation angles, obtain the echo signal fed back by the target corner reflector for the radar test signal and extract the first array steering vector matrix corresponding to the target corner reflector.

[0008] Step S2: Establish a signal model without channel amplitude-phase errors and with azimuth error, elevation error, and roll error existing simultaneously to form a second array steering vector matrix.

[0009] Step S3: Construct a non-linear least squares optimization problem according to the relationship between the first array steering vector matrix, the amplitude-phase calibration coefficients of multiple channels of the radar, and the second array steering vector matrix.

[0010] Step S4: Use an iterative sequential quadratic optimization algorithm to solve the original problem to obtain the estimated values of the amplitude-phase calibration coefficients, azimuth error, elevation error, and roll error of multiple channels of the radar.

[0011] Further, the multiple azimuth and elevation angles in Step S1 are multiple nominal azimuth and elevation angles, respectively expressed as: (α1,β1),(α2,β2),…,(α L ,β L );

[0012] The corresponding first array steering vector extracted is expressed as: To construct the first array steering vector matrix, the first array steering vector matrix is expressed as:

[0013]

[0014] Among them, A m is the first array steering vector matrix, and L is a positive integer.

[0015] Further, the l true azimuth and elevation angles of the target corner reflector relative to the radar can be written as:

[0016] (α1+Δα,β1+Δβ),(α2+Δα,β2+Δβ),…,(α L +Δβ,β L +Δβ);

[0017] Among them, αl is the l-th true azimuth angle, βl is the l-th true elevation angle, Δα is the azimuth error, Δβ is the elevation error, 0<l≤L, and l is a positive integer.

[0018] Furthermore, the signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error in step S2 is specifically:

[0019]

[0020] Among them, A t is the second array steering vector matrix, λ is the wavelength of the radar signal, Δγ is the rolling error of the radar multi-channel, P is the matrix composed of the array element coordinates, the first row of P corresponds to the first element of the two-dimensional coordinates of all array elements, the second row of P corresponds to the second element of the two-dimensional coordinates of all array elements, K is the inverse true direction matrix of the target angle, and the lth column of K corresponds to the azimuth pitch angle (α l +Δα,β l +Δβ), (·) T represents the transpose operation, and exp() is the exponential function.

[0021] Furthermore, the sub-steps of constructing the nonlinear least squares optimization problem in step S3 include:

[0022] S31. Design the following optimal estimator:

[0023]

[0024] stω T [η T ,Δα,Δβ,Δγ] T =1,

[0025] Among them, A m is the first array steering vector matrix, diag(·) represents a diagonal matrix formed by vectors, η is the amplitude and phase calibration coefficient of the radar multi-channel, ||·|| F represents the F norm, st is the constraint condition, and ω represents a column vector whose first element is 1 and the rest are 0.

[0026] Furthermore, the sub-steps of the process of solving the original problem using the iterative sequential quadratic optimization algorithm in step S4 include:

[0027] (1) Let q = 0, where q is the iteration count,

[0028] Let the result of the 0th iteration be Δα (0) =0,Δβ (0) =0,Δγ (0) =0,

[0029] By solving Get η (0) ;

[0030] Among them, K b is the nominal direction matrix of the target angle, K b The lth column corresponds to the azimuth pitch angle (α l ,β l );

[0031] (2) Perform iterative operation steps, setting q = q + 1;

[0032] In the q+1th iteration step, let the qth iteration result be η (q) , Δα (q) ,Δβ (q) ,Δγ (q) , solve the following system of equations:

[0033]

[0034] in, is the direction matrix of the target angle after the qth iteration update, The lth column corresponds to the azimuth pitch angle (α l +Δα (q) ,β l +Δβ (q) ), vec(·) means stacking the matrix into column vectors, (·) H represents the conjugate transpose, B (q) for Relative to [(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T The Jacobian matrix of ; x is the variable to be determined, δ is the auxiliary variable to be determined, and q is a positive integer;

[0035] (3) In the q+1th iteration step, the update formula is:

[0036] [(η (q+1) ) T ,Δα (q+1) ,Δβ (q+1) ,Δγ (q+1) ] T =[(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T +x,

[0037] Substitute the x obtained in step (2) into step (3);

[0038] (4) If the number of iterations reaches the set maximum number of iterations, or the 2-norm of x is less than the set threshold, the iteration is stopped and the estimated values ​​of the amplitude and phase calibration coefficients of the radar multi-channel are output. and the estimated azimuth error Pitch error estimate and the roll error estimate Otherwise, repeat steps (2) to (4).

[0039] Further, in step S1, before driving the servo carrying the radar in the darkroom to rotate, the method includes:

[0040] The radar is placed on a servo in the darkroom and the target angle is placed normal to the far field position of the radar.

[0041] According to another aspect of the present invention, a radar channel calibration device is provided, wherein the radar comprises at least one transmitting antenna and at least one receiving antenna, wherein the at least one transmitting antenna and the at least one receiving antenna constitute a plurality of channels, and the device comprises:

[0042] A driving and echo signal acquisition unit, used for driving the servo carrying the radar in the darkroom to rotate, so that the target angle reflector is located at multiple azimuth and elevation angles of the radar, the radar transmits a radar test signal at each azimuth and elevation angle, and for each azimuth and elevation angle, obtains an echo signal fed back by the target angle reflector to the radar test signal and extracts a first array steering vector matrix corresponding to the target angle reflector;

[0043] A signal model building unit, used for building a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error, so as to form a second array steering vector matrix;

[0044] An optimization problem constructing unit, used to construct a nonlinear least squares optimization problem according to the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel, and the azimuth error, the pitch error, and the roll error;

[0045] The optimization problem solving unit solves the original problem by using an iterative sequential quadratic optimization algorithm to obtain an estimated value of amplitude and phase calibration coefficients, an estimated value of azimuth error, an estimated value of pitch error, and an estimated value of roll error of the radar multi-channel.

[0046] According to another aspect of the present invention, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the radar channel calibration method described in any of the above embodiments is implemented.

[0047] Through one or more of the above embodiments of the present invention, at least the following technical effects can be achieved:

[0048] The technical solution disclosed in the present invention aims to establish a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error to form a second array steering vector matrix, and construct a nonlinear least squares optimization problem based on the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel and the second array steering vector matrix, and solve the relevant parameters based on the constraint conditions to completely eliminate the influence of the radar azimuth error, pitch error and roll error during the radar darkroom calibration process. The estimated multi-channel amplitude and phase calibration coefficients are decoupled from the azimuth error, pitch error and roll error, that is, the calibration coefficients can accurately reflect the channel amplitude and phase errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The technical solutions and other beneficial effects of the present invention will be made apparent by describing in detail the specific embodiments of the present invention in conjunction with the accompanying drawings.

[0050] Figure 1 Schematic diagram of radar array antenna echo signal reception.

[0051] Figure 2 A flowchart of a radar channel calibration method provided by an embodiment of the present invention.

[0052] Figure 3 A schematic diagram of the structure of a radar channel calibration device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0054] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " herein, unless otherwise specified, generally indicates that the associated objects before and after are in an "or" relationship.

[0055] The technical terms involved in the present invention are described below:

[0056] MIMO (Multiple-Input Multiple-Output) technology refers to the use of multiple transmitting antennas and receiving antennas at the transmitting end and receiving end respectively, so that the signal is transmitted on multiple channels, thereby improving the communication quality. It can make full use of spatial resources and achieve multiple transmission and multiple reception through multiple antennas. Without increasing spectrum resources and antenna transmission power, it can multiply the system channel capacity.

[0057] Figure 1 Schematic diagram of radar array antenna echo signal reception.

[0058] like Figure 1 As shown in the figure, under far-field conditions, the phase difference between radar array antennas is caused by the path difference of the target echo signal reaching each array antenna. This phase difference is recorded as Usually and array spacing d m And the target direction of arrival (DOA). The phase difference of the data collected by the radar receiver for each antenna channel is not only caused by the wavelength difference, but also related to the feeder length of each channel. The channel delay phase difference caused by the unequal length of the feeder of each channel, denoted as ε, is not related to the target direction of arrival. Therefore, when ε is superimposed on When the radar is on the target, it will affect the DOA estimation of the radar. As mentioned in the background, the calibrated radar still has direction finding errors caused by azimuth pitch error and roll error.

[0059] Figure 2 A flowchart of a radar channel calibration method provided by an embodiment of the present invention.

[0060] like Figure 2 As shown, the radar channel calibration method provided by the embodiment of the present invention includes:

[0061] Step S1, driving the servo carrying the radar in the darkroom to rotate so that the target angle reflector is located at multiple azimuth and elevation angles of the radar respectively, the radar transmits a radar test signal at each azimuth and elevation angle respectively, and for each azimuth and elevation angle, obtaining an echo signal fed back by the target angle reflector to the radar test signal and extracting a first array steering vector matrix corresponding to the target angle reflector;

[0062] Step S2, establishing a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error, so as to form a second array steering vector matrix;

[0063] Step S3, constructing a nonlinear least squares optimization problem according to the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel, and the second array steering vector matrix;

[0064] Step S4, using an iteration-based sequential quadratic optimization algorithm to solve the original problem, and obtaining the estimated values ​​of the amplitude and phase calibration coefficients, the estimated values ​​of the azimuth error, the estimated values ​​of the pitch error, and the estimated values ​​of the roll error of the radar multi-channel.

[0065] The technical solution disclosed in the present invention can be used for single-transmit multiple-receive radar, multiple-transmit single-receive radar, and MIMO (Multiple-Input Multiple-Output) radar; wherein, MIMO radar refers to a radar with a multiple-transmit multiple-receive antenna system, which provides an economical and effective way to improve the resolution of the radar, and the antenna volume is relatively small. Based on a multi-element antenna, the MIMO radar uses M channels to transmit mutually orthogonal signals, and the multi-waveform signals remain independent in space. After being scattered by the target, they are received by N receiving array elements, and each array element uses M matched filters to match the echo, so that the echo data of M*N channels can be obtained.

[0066] The above steps S1 to S4 are described in detail below.

[0067] In step S1, the servo carrying the radar in the darkroom is driven to rotate so that the target angle reflector is located at multiple azimuth and elevation angles of the radar respectively, and the radar transmits a radar test signal at each azimuth and elevation angle respectively, and for each azimuth and elevation angle, an echo signal fed back by the target angle reflector to the radar test signal is obtained and a first array steering vector matrix corresponding to the target angle reflector is extracted.

[0068] Exemplarily, in a MIMO radar system, a pulse sequence is transmitted in a time-division transmission mode, and the time required for all array antennas to transmit a pulse signal in turn is a Burst cycle (i.e., a pulse sequence cycle), and the time of a Burst cycle is divided into multiple Chirp signal cycles, wherein the multiple Chirp signal cycles within a Burst cycle can be the same or different, and only one antenna transmits a pulse signal in each Chirp signal cycle.

[0069] In the embodiment of the present invention, the multiple azimuth and elevation angles in step S1 are multiple nominal azimuth and elevation angles, which are respectively expressed as: (α1, β1), (α2, β2), …, (α L ,β L ), where L is a positive integer; that is, the servo is rotated to L nominal azimuth and pitch angles, and the L nominal azimuth and pitch angles are respectively expressed as (α1, β1), (α2, β2),…, (α L ,β L)。The radar transmits signals at each azimuth and elevation angle respectively, collects baseband data, and extracts the first array steering vector corresponding to the target angle reflection; the extracted first array steering vector is expressed as to construct the first array steering vector matrix A m , which is used for subsequent calculation of roll error and radar channel amplitude-phase error coefficient. Specifically, let Call A m the measurement array manifold. Due to the existence of azimuth error and elevation error, the L true azimuth and elevation angles of the target angle reflection relative to the radar can be written as (α1+Δα,β1+Δβ),(α2+Δα,β2+Δβ),…,(α L +Δβ,β L +Δβ); where, αl is the l-th true azimuth angle, βl is the l-th true elevation angle, Δα is the azimuth error, Δβ is the elevation error, 0<l≤L, and l is a positive integer.

[0070] In step S2, establish a signal model without channel amplitude-phase error and with azimuth error, elevation error, and roll error simultaneously to form a second array steering vector matrix.

[0071] The establishment of the signal model without channel amplitude-phase error and with azimuth error, elevation error, and roll error simultaneously in step S2 is specifically as follows:

[0072]

[0073] where, A t is the second array steering vector matrix, λ is the wavelength of the radar transmitted signal, Δγ is the roll error of the radar multi-channel, P is the matrix composed of array element coordinates, the first row of P corresponds to the first element of the two-dimensional coordinates of all array elements, the second row of P corresponds to the second element of the two-dimensional coordinates of all array elements, P is a known parameter, K is the true direction matrix of the target angle reflection, and the l-th column of K corresponds to the azimuth and elevation angle (α l +Δα,β l +Δβ), (·) T represents the transpose operation, and exp() is the exponential function.

[0074] Specifically, let the array manifold without channel amplitude-phase error but with azimuth error, elevation error, and roll error simultaneously be A t =[a1,a2,...,a L , where a l is the l-th array steering vector without channel amplitude-phase error, with azimuth and elevation error, and with roll error. A t can be modeled as to form the second array steering vector.

[0075] In step S3, a nonlinear least squares optimization problem is constructed according to the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel, and the second array steering vector matrix.

[0076] The sub-steps of constructing the nonlinear least squares optimization problem in step S3 include:

[0077] S31. Design the following optimal estimator:

[0078]

[0079] stω T [η T ,Δα,Δβ,Δγ] T =1,

[0080] Among them, A m is the first array steering vector matrix, diag(·) represents a diagonal matrix formed by vectors, η is the amplitude and phase calibration coefficient of the radar multi-channel, ||·|| F represents the F norm, st is the constraint condition, and ω represents a column vector whose first element is 1 and the rest are 0.

[0081] In step S4, an iteration-based sequential quadratic optimization algorithm is used to solve the original problem, and the estimated values ​​of amplitude and phase calibration coefficients, azimuth error, pitch error, and roll error of the radar multi-channel are obtained.

[0082] The sub-steps of the process of solving the original problem by using the iterative sequential quadratic optimization algorithm in step S4 include:

[0083] (1) Let q = 0, where q is the iteration count,

[0084] Let the result of the 0th iteration be Δα (0) =0,Δβ (0) =0,Δγ (0) =0,

[0085] By solving Get η (0) ;

[0086] Among them, K b is the nominal direction matrix of the target angle, K b The lth column corresponds to the azimuth pitch angle (α l ,β l );

[0087] (2) Perform iterative operation steps, setting q = q + 1;

[0088] In the q+1th iteration step, let the qth iteration result be η (q) , Δα (q) ,Δβ (q) ,Δγ (q) , solve the following system of equations:

[0089]

[0090] in, is the direction matrix of the target angle after the qth iteration update, The lth column corresponds to the azimuth pitch angle (α l +Δα (q) ,β l +Δβ (q) ), vec(·) means stacking the matrix into column vectors, (·) H represents the conjugate transpose, B (q) for Relative to [(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T The Jacobian matrix of ; x is the variable to be determined, δ is the auxiliary variable to be determined, and q is a positive integer;

[0091] (3) In the q+1th iteration step, the update formula is:

[0092] [(η (q+1) ) T ,Δα (q+1) ,Δβ (q+1) ,Δγ (q+1) ] T =[(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T +x,

[0093] Substitute the x obtained in step (2) into step (3);

[0094] (4) If the number of iterations reaches the set maximum number of iterations, or the 2-norm of x is less than the set threshold, the iteration is stopped and the estimated values ​​of the amplitude and phase calibration coefficients of the radar multi-channel are output. and the estimated azimuth error Pitch error estimate and the roll error estimate Otherwise, repeat steps (2) to (4).

[0095] By solving the four unknown parameters η, Δγ, Δα and Δβ in the optimization estimator, the result of the optimization estimator is minimized to obtain the estimated value of the amplitude and phase calibration coefficient of the radar multi-channel Roll error estimate Azimuth error estimate and the pitch error estimate Use the amplitude and phase calibration coefficients to estimate the value Approximate the amplitude and phase calibration coefficients of the real radar multi-channel, using the roll error estimate Approximate the true roll error value and use the azimuth error estimate Approximate the azimuth error of the real radar multi-channel and use the pitch error estimate Approximate the pitch error of real radar multi-channel.

[0096] The technical solution provided in this embodiment aims to form a second array steering vector matrix by establishing a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error. A nonlinear least squares optimization problem is constructed based on the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel and the second array steering vector matrix, and the relevant parameters are solved based on the constraint conditions to minimize, so as to completely eliminate the influence of the radar azimuth error, pitch error and roll error during the radar darkroom calibration process. The estimated multi-channel amplitude and phase calibration coefficient is decoupled from the azimuth error, pitch error and roll error, that is, the calibration coefficient can accurately reflect the channel amplitude and phase error.

[0097] Secondly, before calibration, there is no need to use laser irradiation combined with mirror reflection and bubble level to fine-tune the radar attitude. The existing attitude error is automatically and accurately compensated by the calibration algorithm.

[0098] Finally, the radar channel calibration method provided by the present invention is not based on Taylor approximation, and therefore is still effective when the azimuth, pitch or roll errors are large.

[0099] Further, in step S1, before driving the servo carrying the radar in the darkroom to rotate, the method includes: placing the radar on the servo in the darkroom, and placing the target angle reflector on the normal line of the far-field position of the radar.

[0100] Assume a 4x4 radar square array with 16 array elements and a half-wavelength spacing. Set the radar azimuth error Δα, pitch error Δβ, roll error Δγ and amplitude-phase error coefficient η. Measure a total of L = 161 nominal azimuth and pitch angle steering vectors. For example, first set the pitch angle to 0 degrees, and scan the azimuth angle from -40 degrees to 40 degrees at 1 degree intervals; then set the azimuth angle to 0 degrees, and scan the pitch angle from -40 degrees to 40 degrees at 1 degree intervals, and extract the array steering vector corresponding to the target angle to form the first array steering vector matrix A. m . Let Δα be 0.5 degrees, Δβ be 0.3 degrees, Δγ be 0.6 degrees, and the setting of the amplitude and phase errors is shown in the first column of Table 1. The amplitude and phase calibration coefficients and the azimuth pitch and roll error values ​​estimated by the radar channel calibration method provided by the embodiment of the present invention are shown in Tables 1 and 2.

[0101] Table 1

[0102]

[0103] Table 2

[0104] Settings First iteration value Second iteration value Iteration 3 Azimuth error (degrees) 0.5000 0.5027 0.5000 0.5000 Pitch error (degrees) 0.3000 0.2943 0.2999 0.3000 Roll error (degrees) 0.6000 0.6000 0.6000 0.6000

[0105] It can be found from the results in Table 2 that the method provided by the embodiment of the present invention only requires three iterative calculations to obtain the accurate values ​​of the channel amplitude and phase errors, azimuth pitch and roll errors.

[0106] Based on the same inventive concept as the radar channel calibration method in the embodiment of the present invention, the embodiment of the present invention further provides a radar channel calibration device 200, wherein the radar includes at least one transmitting antenna and at least one receiving antenna, and the at least one transmitting antenna and the at least one receiving antenna constitute a plurality of channels, please refer to Figure 3 , the device 200 comprises:

[0107] A driving and echo signal acquisition unit 201 is used to drive the servo carrying the radar in the darkroom to rotate so that the target angle reflector is located at multiple azimuth and elevation angles of the radar, the radar transmits a radar test signal at each azimuth and elevation angle, and for each azimuth and elevation angle, obtains an echo signal fed back by the target angle reflector to the radar test signal and extracts a first array steering vector matrix corresponding to the target angle reflector;

[0108] A signal model building unit 202 is used to build a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error to form a second array steering vector matrix;

[0109] An optimization problem construction unit 203 is configured to construct a non - linear least - squares optimization problem according to the first array steering vector matrix, the amplitude - phase calibration coefficients of multiple radar channels, and the relationships of azimuth error, pitch error, and roll error;

[0110] An optimization problem solving unit 204 is configured to solve the original problem by using an iterative sequential quadratic programming algorithm to obtain the estimated values of the amplitude - phase calibration coefficients, azimuth error, pitch error, and roll error of the multiple radar channels.

[0111] Exemplarily, the driving and echo signal acquisition unit 201 extracts the first array steering vector matrix corresponding to the target angle reflection in the following manner:

[0112] In the embodiment of the present invention, multiple nominal azimuth - pitch angles are selected, which are respectively represented as: (α1,β1),(α2,β2),…,(α L ,β L ), where L is a positive integer; that is, the servo is rotated to L nominal azimuth - pitch angles, and the L nominal azimuth - pitch angles are respectively represented as (α1,β1),(α2,β2),…,(α L ,β L ). The radar sends signals at each azimuth - pitch angle respectively, acquires baseband data, and extracts the first array steering vector corresponding to the target angle reflection; the extracted first array steering vector is represented as to construct the first array steering vector matrix A m , which is used for subsequent calculation of roll error and radar channel amplitude - phase error coefficients. Specifically, let It is called that A m is the measurement array manifold. Due to the existence of azimuth error and pitch error, the L true azimuth - pitch angles of the target angle reflection relative to the radar can be written as (α1 + Δα,β1 + Δβ),(α2 + Δα,β2 + Δβ),…,(α L +Δβ,β L +Δβ); where αl is the l - th true azimuth angle, βl is the l - th true pitch angle, Δα is the azimuth error, Δβ is the pitch error, 0 < l ≤ L, and l is a positive integer.

[0113] Exemplarily, the signal model establishment unit 202 establishes a signal model without channel amplitude - phase error and with azimuth error, pitch error, and roll error simultaneously in the following manner, specifically:

[0114]

[0115] where A t is the second array steering vector matrix, λ is the wavelength of the radar transmitted signal, Δγ is the roll error of the radar multi-channel, P is the matrix composed of the array element coordinates, the first row of P corresponds to the first element of the two-dimensional coordinates of all array elements, the second row of P corresponds to the second element of the two-dimensional coordinates of all array elements, P is an existing parameter, K is the inverse true direction matrix of the target angle, and the lth column of K corresponds to the azimuth pitch angle (α l +Δα,β l +Δβ), (·) T represents the transpose operation, and exp() is the exponential function.

[0116] Specifically, let the array flow pattern without channel amplitude and phase errors, but with azimuth error, pitch error and roll error be A t =[a1,a2,...,a L ], where a l A is the lth array steering vector without channel amplitude and phase errors but with azimuth pitch error and roll error. t Can be modeled as To form a second array steering vector.

[0117] Exemplarily, the optimization problem constructing unit 203 constructs a nonlinear least squares optimization problem in the following manner, specifically:

[0118] Design the following optimal estimator:

[0119]

[0120] stω T [η T ,Δα,Δβ,Δγ] T =1,

[0121] Among them, A m is the first array steering vector matrix, diag(·) represents a diagonal matrix formed by vectors, η is the amplitude and phase calibration coefficient of the radar multi-channel, ||·|| F represents the F norm, st is the constraint condition, and ω represents a column vector whose first element is 1 and the rest are 0.

[0122] Exemplarily, the sub-steps of the optimization problem solving unit 204 using the iterative sequential quadratic optimization algorithm to solve the original problem include:

[0123] (1) Let q = 0, where q is the iteration count,

[0124] Let the result of the 0th iteration be Δα (0) =0,Δβ (0) =0,Δγ (0) =0,

[0125] By solving Get η(0) ;

[0126] Among them, K b is the nominal direction matrix of the target angle, K b The lth column corresponds to the azimuth pitch angle (α l ,β l );

[0127] (2) Perform iterative operation steps, setting q = q + 1;

[0128] In the q+1th iteration step, let the qth iteration result be η (q) , Δα (q) ,Δβ (q) ,Δγ (q) , solve the following system of equations:

[0129]

[0130] in, is the direction matrix of the target angle after the qth iteration update, The lth column corresponds to the azimuth pitch angle (α l +Δα (q) ,β l +Δβ (q) ), vec(·) means stacking the matrix into column vectors, (·) H represents the conjugate transpose, B (q) for Relative to [(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T The Jacobian matrix of ; x is the variable to be determined, δ is the auxiliary variable to be determined, and q is a positive integer;

[0131] (3) In the q+1th iteration step, the update formula is:

[0132] [(η (q+1) ) T ,Δα (q+1) ,Δβ (q+1) ,Δγ (q+1) ] T =[(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T +x,

[0133] Substitute the x obtained in step (2) into step (3);

[0134] (4) If the number of iterations reaches the set maximum number of iterations, or the 2-norm of x is less than the set threshold, the iteration is stopped and the estimated values ​​of the amplitude and phase calibration coefficients of the radar multi-channel are output. and the estimated azimuth error Pitch error estimate and the roll error estimate Otherwise, repeat steps (2) to (4).

[0135] According to another aspect of the present invention, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the radar channel calibration method described in any of the aforementioned embodiments is implemented.

[0136] The specific definition and implementation of the above steps can refer to the steps and methods of the embodiment of the radar channel calibration method, which will not be repeated here. Those of ordinary skill in the art can understand that all or part of the processes in the above embodiment methods can be completed by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0137] In summary, although the present invention has been disclosed as above in terms of preferred embodiments, the above preferred embodiments are not intended to limit the present invention. A person skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined in the claims.

Claims

1. A radar channel calibration method, characterized in that: The radar comprises at least one transmitting antenna and at least one receiving antenna, wherein the at least one transmitting antenna and the at least one receiving antenna form a plurality of channels, and the method comprises: Step S1, driving the servo carrying the radar in the darkroom to rotate so that the target angle reflector is located at multiple azimuth and elevation angles of the radar respectively, the radar transmits a radar test signal at each azimuth and elevation angle respectively, and for each azimuth and elevation angle, obtaining an echo signal fed back by the target angle reflector to the radar test signal and extracting a first array steering vector matrix corresponding to the target angle reflector; Step S2, establishing a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error, so as to form a second array steering vector matrix; Step S3, constructing a nonlinear least squares optimization problem according to the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel, and the second array steering vector matrix; Step S4, using an iteration-based sequential quadratic optimization algorithm to solve the original problem, and obtaining the estimated values ​​of the amplitude and phase calibration coefficients, the estimated values ​​of the azimuth error, the estimated values ​​of the pitch error, and the estimated values ​​of the roll error of the radar multi-channel.

2. The radar channel calibration method according to claim 1, characterized in that: The multiple azimuth and elevation angles in step S1 are multiple nominal azimuth and elevation angles, which are respectively expressed as: (α1, β1), (α2, β2), …, (α L ,β L ); The corresponding extracted first array steering vector is expressed as: To construct a first array steering vector matrix, the first array steering vector matrix is ​​expressed as: Among them, A m is the first array steering vector matrix, and L is a positive integer.

3. The radar channel calibration method according to claim 2, characterized in that: The l true azimuth and elevation angles of the target angle relative to the radar can be written as: (α1+Δα,β1+Δβ),(α2+Δα,β2+Δβ),…,(α L +Δβ,β L +Δβ); where α l is the l-th true azimuth angle, β l is the l-th true elevation angle, Δα is the azimuth error, Δβ is the elevation error, 0 < l ≤ L, and l is a positive integer.

4. The radar channel calibration method according to claim 3, characterized in that: The signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error in step S2 is specifically: Among them, A t is the second array steering vector matrix, λ is the wavelength of the radar signal, Δγ is the rolling error of the radar multi-channel, P is the matrix composed of the array element coordinates, the first row of P corresponds to the first element of the two-dimensional coordinates of all array elements, the second row of P corresponds to the second element of the two-dimensional coordinates of all array elements, K is the inverse true direction matrix of the target angle, and the lth column of K corresponds to the azimuth pitch angle (α l +Δα,β l +Δβ), (·) T represents the transpose operation, and exp() is the exponential function.

5. The radar channel calibration method according to claim 4, characterized in that: The sub-steps of constructing the nonlinear least squares optimization problem in step S3 include: S31. Design the following optimal estimator: stω T [or T ,Dα,Δβ,Δγ] T =1, Among them, A m is the first array steering vector matrix, diag(·) represents a diagonal matrix formed by vectors, η is the amplitude and phase calibration coefficient of the radar multi-channel, ||·|| F represents the F norm, st is the constraint condition, and ω represents a column vector whose first element is 1 and the rest are 0.

6. The radar channel calibration method according to claim 5, characterized in that: The sub-steps of the process of solving the original problem using the iterative sequential quadratic optimization algorithm in step S4 include: (1) Let q = 0, where q is the iteration count, Let the result of the 0th iteration be Δα (0) =0,Δβ (0) =0,Δγ (0) =0, By solving Get η (0) ; Among them, K b is the nominal direction matrix of the target angle, K b The lth column corresponds to the azimuth pitch angle (α l ,β l ); (2) Perform iterative operation steps, setting q = q + 1; In the q+1th iteration step, let the qth iteration result be η (q) , Δα (q) ,Δβ (q) ,Δγ (q) , solve the following system of equations: in, is the direction matrix of the target angle after the qth iteration update, The lth column corresponds to the azimuth pitch angle (α l +Δα (q) ,β l +Δβ (q) ), vec(·) means stacking the matrix into column vectors, (·) H represents the conjugate transpose, B (q) for Relative to [(η (q) ) T ,Δα (q) ,Δβ (q) ,Δγ (q) ] T The Jacobian matrix of ; x is the variable to be determined, δ is the auxiliary variable to be determined, and q is a positive integer; (3) In the q+1th iteration step, the update formula is: [(or (q+1) ) T ,Da (q+1) ,Db (q+1) ,Dg (q+1) ] T =[(η (q) ) T ,Da (q) ,Db (q) ,Dg (q) ] T +x, Substitute the x obtained in step (2) into step (3); (4) If the number of iterations reaches the set maximum number of iterations, or the 2-norm of x is less than the set threshold, the iteration is stopped and the estimated values ​​of the amplitude and phase calibration coefficients of the radar multi-channel are output. and the estimated azimuth error Pitch error estimate and the roll error estimate Otherwise, repeat steps (2) to (4).

7. The radar channel calibration method according to claim 1, characterized in that: In step S1, before driving the servo carrying the radar in the darkroom to rotate, the method includes: The radar is placed on a servo in the darkroom and the target angle is placed normal to the far field position of the radar.

8. A radar channel calibration device, characterized in that: The radar comprises at least one transmitting antenna and at least one receiving antenna, wherein the at least one transmitting antenna and the at least one receiving antenna form a plurality of channels, and the device comprises: A driving and echo signal acquisition unit, used for driving the servo carrying the radar in the darkroom to rotate, so that the target angle reflector is located at multiple azimuth and elevation angles of the radar, the radar transmits a radar test signal at each azimuth and elevation angle, and for each azimuth and elevation angle, obtains an echo signal fed back by the target angle reflector to the radar test signal and extracts a first array steering vector matrix corresponding to the target angle reflector; A signal model building unit, used for building a signal model without channel amplitude and phase errors and with azimuth error, pitch error and roll error, so as to form a second array steering vector matrix; An optimization problem constructing unit, used to construct a nonlinear least squares optimization problem according to the relationship between the first array steering vector matrix, the amplitude and phase calibration coefficients of the radar multi-channel, and the azimuth error, the pitch error, and the roll error; The optimization problem solving unit solves the original problem by using an iterative sequential quadratic optimization algorithm to obtain an estimated value of amplitude and phase calibration coefficients, an estimated value of azimuth error, an estimated value of pitch error, and an estimated value of roll error of the radar multi-channel.

9. A storage medium, characterized in that: The storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the radar channel calibration method according to any one of claims 1 to 7.

Images