A polarization calibration method based on reciprocity and symmetry of insect targets

Abstract

The present invention discloses a polarization calibration method based on the reciprocity and symmetry of insect targets. The present invention can be used to quickly and labor-savingly perform polarization calibration on a full polarization radar. The present invention first deforms the system model and decomposes the system parameters into two parts that can be solved by reciprocity constraints and symmetry constraints; then, the reciprocity and symmetry of the insect target are used as constraints to directly solve the system parameters subject to reciprocity and design cost functions to solve the system parameters of other parts through genetic algorithms; finally, the system error is reversed into the measured target scattering matrix to obtain the real target scattering matrix. Compared with the existing polarization calibration algorithms, the present method is more efficient and does not require the design of a calibration body. Polarization calibration can be performed at any time, avoiding the impact of harsh field environments.

Description

Technical Field

[0001] The invention belongs to the technical field of insect radars, and in particular relates to a polarization calibration method based on insect target reciprocity and symmetry. Background Art

[0002] Insect radar is an important tool for monitoring insect migration. Fully polarized insect radar is currently a radar system with high measurement efficiency, high accuracy and the greatest potential. Fully polarized radar can obtain target polarization information in a very short time, and then invert the direction, body length, weight and other information of the insect through the target's polarization information. Based on these parameters, we can determine the type of insect, analyze the direction of insect flight, and then give early warning. This is of great significance for preventing the outbreak of pests and diseases and studying the migration theory of insects.

[0003] Polarization is one of the important radar target characteristics. In order to accurately obtain the target polarization information, the full polarization radar needs to be polarization calibrated first.

[0004] Traditional full polarimetric radar polarization calibration is to estimate the systematic error by measuring the target with known polarization scattering matrix, and then compensate the measured target scattering matrix. This process is usually very complicated and requires precision-machined calibration bodies. Summary of the invention

[0005] In view of this, the present invention provides a polarization calibration method based on the reciprocity and symmetry of insect targets, which can avoid the complex operation of traditional radar polarization calibration and improve the efficiency of polarization calibration. In the field environment, the system is often affected by temperature and changes. At this time, the system needs to be re-polarized, otherwise the accuracy of the measured data will decrease. The newly proposed method can perform polarization calibration without relying on manpower, and can automatically use the measured target for polarization calibration, and can compensate for the time-varying characteristics of the system, thereby indirectly improving the accuracy of polarization measurement. This helps to study the migration behavior of insects, predict the migration direction of insects, and warn of the outbreak of pests and diseases.

[0006] A polarization calibration method based on insect target reciprocity and symmetry is proposed. The radar system error is parameterized and decomposed into a part solved by reciprocity and a part solved by symmetry. Then, the error caused by the system transmission link is first solved by reciprocity. c , and then use the symmetry to solve the error a caused by the system transmission link 1 , and the crosstalk C between channels, and then use the obtained a c 、a 1 and C compensate the system.

[0007] Preferably, the scattering matrix obtained after the parameter decomposition is: Among them, M hh Indicates the echo obtained by H polarization transmission and H polarization reception, M vh It represents the echo obtained by transmitting with H polarization and receiving with V polarization, M hv Indicates the echo obtained by transmitting with V polarization and receiving with H polarization, M vv Indicates the echo obtained by V polarization transmission and V polarization reception, g = R h T h , g will act on all elements of the scattering matrix; S hh represents the echo obtained by H polarization transmission and H polarization reception, where S vh It represents the echo obtained by transmitting with H polarization and receiving with V polarization, where S hv represents the echo obtained by transmitting with V polarization and receiving with H polarization, where S vv Indicates the echo obtained by V polarization transmission and V polarization reception; a c It represents the ratio of the difference between the H and V transmitting channels to the difference between the H and V receiving channels; a 1 represents the difference between the V polarized receiving channel and the H polarized receiving channel, and C represents the crosstalk between the channels.

[0008] Preferably, the error a caused by the system transmission link is solved by using reciprocity. c The method is: take the ratio of the echo obtained by V polarization transmission and H polarization reception to the echo obtained by H polarization transmission and V polarization reception as the error a c .

[0009] Preferably, the error a caused by the system transmission link is solved by using symmetry. 1 , and the specific method of crosstalk C between channels is: first, according to the error a obtained c The system is compensated, and then an optimization function is established at the cost of reciprocity to obtain the optimal solution.

[0010] Preferably, the optimal solution is solved using a genetic algorithm.

[0011] Preferably, when the solutions obtained by the optimization function are opposite to each other, the phases corresponding to the two solutions are selected to be respectively equal to the error a 1 The solution with the smallest difference in the initial phase is taken as the optimal solution.

[0012] Beneficial Effects

[0013] 1. The present invention can be used to quickly and labor-savingly perform polarization calibration on a full-polarization radar. The present invention first deforms the system model and decomposes the system parameters into two parts that can be solved by reciprocity constraints and symmetry constraints; then, the reciprocity and symmetry of the insect target are used as constraints to directly solve the system parameters subject to reciprocity and design cost functions to solve the system parameters of other parts through genetic algorithms; finally, the system error is reversed into the measured target scattering matrix to obtain the real target scattering matrix. Compared with the existing polarization calibration algorithms, this method is more efficient and does not require the design of a calibration body. Polarization calibration can be performed at any time, avoiding the impact of harsh field environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1(a) shows the reciprocity of insects at different frequencies;

[0015] Figure 1(b) shows the symmetry of insects at different frequencies.

[0016] Figure 2(a) shows the amplitude inconsistency distribution after calibration;

[0017] Figure 2(b) shows the phase inconsistency distribution after calibration;

[0018] Figure 2(c) shows the isolation distribution after calibration. DETAILED DESCRIPTION

[0019] The implementation of the method of the present invention is described below with reference to the accompanying drawings and examples.

[0020] The present invention provides a polarization calibration method based on the reciprocity and symmetry of insect targets. The basic idea is to directly solve the system parameters of the reciprocity system with the reciprocity and symmetry of insect targets as constraints, design the cost function with symmetry as constraints and solve the system parameters of other parts through genetic algorithms, and finally use the obtained system polarization error to compensate the measured data to obtain the real scattering matrix of the target.

[0021] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0022] Radar inverts insect parameters by measuring the polarization information of insects, and one of the main manifestations of polarization information is the scattering matrix. Suppose the insect scattering matrix is:

[0023]

[0024] Where S hh , S vv , S hv and S vh is the measurement result corresponding to each channel of the scattering matrix. hhrepresents the echo obtained by H polarization transmission and H polarization reception, where S vh It represents the echo obtained by transmitting with H polarization and receiving with V polarization, where S hv represents the echo obtained by transmitting with V polarization and receiving with H polarization, where S vv It represents the echo obtained by V polarization transmission and V polarization reception. The scattering matrix of insect targets measured by radar is:

[0025]

[0026] M is the scattering matrix containing the systematic errors. hh 、M vv 、M hv and M vh is the measurement result corresponding to each channel of the scattering matrix. hh It represents the echo obtained by H polarization transmission and H polarization reception, where M vh It represents the echo obtained by transmitting with H polarization and receiving with V polarization, where M hv It represents the echo obtained by transmitting with V polarization and receiving with H polarization, where M vv Indicates the echo obtained by V-polarization transmission and V-polarization reception.

[0027] A classic fully polarized system model is:

[0028]

[0029] R h , R v , T h , T v They represent the transmission characteristics of H receiving channel, V receiving channel, H transmitting channel and V transmitting channel respectively. C represents the crosstalk between channels.

[0030] Traditional polarization calibration algorithms require the use of drones to lift precision-machined metal balls into the air, or to place precision-machined corner reflectors at a distance by placing the radar high up. The radar is then used to measure these targets and estimate the system parameters in equation (3).

[0031] Since many measured targets have symmetry and reciprocity, but due to the existence of systematic errors, the directly measured targets lose reciprocity and symmetry. Therefore, it is possible to estimate the systematic error based on the reciprocity and symmetry of the target, and then use the systematic error to compensate the measured target scattering matrix with errors to obtain the real target scattering matrix.

[0032] In order to estimate the system parameters by using the reciprocity and symmetry of the target, this method first decomposes the system parameters into a part that can be solved by reciprocity and a part that can be solved by symmetry. Let g = Rh T h , a 1 =R v / R h 、a c =(T v / T h ) / (R v / R h ), and substituting it into (3) we get:

[0033]

[0034] Among them, g will act on all elements of the scattering matrix. c represents the ratio of the difference between the H and V transmitting channels to the difference between the H and V receiving channels, which can be solved with the reciprocity of the target as a constraint. 1 Represents the difference between the V polarization receiving channel and the H polarization receiving channel, a 1 and C can be solved with the symmetry of the objective as constraints.

[0035] First, solve for a c ; According to reciprocity, it can be concluded that the cross-channels of the target scattering matrix are equal, and a c Affects the relative size of the measured scattering matrix, so a is obtained by comparing the cross-channel of the measured scattering matrix c The value is:

[0036] From the target reciprocity, we know that:

[0037] S hv =S vh (5)

[0038] From formula (4), we can know that:

[0039]

[0040] From (6), we can see that the measured target cross-polarization channel is c There is a close relationship between them, and a can be solved by the following formula c :

[0041]

[0042] Indicates a c The estimated value of .

[0043] Afterwards, based on the obtained Try using Compensation is eliminated by matrix operation. However, in actual tests, it is found that when any Any After compensating the system measurement results, all symmetrical targets will restore symmetry. According to Cameron's method of measuring symmetry, the symmetry of the scattering matrix will not be affected if the non-diagonal elements of the scattering matrix are simultaneously changed to their opposite numbers. This shows that is a sufficient condition to restore the target reciprocity and bilateral symmetry. 1 When and C, the symmetry of the target can be used as a constraint to solve a 1 and C, the specific solution method is:

[0044] Definition 1 and C is estimated to be and Try to use the estimated and any and Compensation system:

[0045]

[0046] Therefore, at the cost of reciprocity, the optimization function is established:

[0047]

[0048] in and represents the simplified system error estimation result, ζ(·) represents the reciprocity degree of the scattering matrix, and i represents the i-th target. Equation (9) is a typical optimization problem that can be solved by genetic algorithm. Genetic algorithm is a mature global optimization algorithm suitable for solving equation (9). In this way, the estimated system error can be obtained: and

[0049] because May be equal to ±a 1 , need to Usually, after the radar is installed, a 1 The phase fluctuation range will not change greatly. So we can measure a once after the radar is installed. 1 The phase of In the subsequent solution Afterwards, if The phase and The absolute value of the difference is less than 90°, then Can be used directly; if The phase and If the absolute value of the difference is greater than 90°, it is considered Blurred, need to be corrected Multiply by -1 to deblur.

[0050] After solving the system error, the real scattering matrix of the target can be solved by matrix operation. The system can be compensated by the following formula to obtain the real S matrix of the target:

[0051]

[0052] Figure 1(a) shows the degree to which insects satisfy reciprocity measured in a darkroom, and Figure 1(b) shows the degree to which insects satisfy symmetry measured in a darkroom.

[0053] To verify the polarization calibration method described above, a Ku-band fully polarized radar was used to measure metal balls and about 40,000 targets in the air that passed through the radar beam over a period of 6 hours. An estimated group of 100 insects was obtained. and Each set of system parameters is used to compensate the metal ball. The amplitude inconsistency, phase inconsistency and isolation results of the ball after calibration are shown in Figure 2. Figure 2 (a), (b) and (c) respectively show the S after the ball is calibrated with insects. vv / S hh The amplitude and phase of the signal, as well as the isolation. vv / S hh The amplitude of S can be used to measure the amplitude inconsistency after calibration. vv / S hh The phase can be used to measure the phase inconsistency after calibration, and the isolation refers to the isolation after calibration. The closer the amplitude inconsistency is to 1, the closer the phase inconsistency is to 0, and the smaller the isolation, the better the calibration effect. As can be seen from Figure 2(a), the amplitude inconsistency is mainly distributed around 1, the phase inconsistency is mainly concentrated at 0 degrees, and the isolation is mostly less than -25dB.

[0054] This method is applicable to all full-polarization radars that can measure insects and can automatically solve system error parameters.

[0055] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A polarization calibration method based on insect target reciprocity and symmetry, characterized by: The radar system error is parameterized and decomposed into a part solved by reciprocity and a part solved by symmetry. Then, the error a caused by the system transmission link is first solved by reciprocity. c , and then use the symmetry to solve the error a1 caused by the system transmission link and the crosstalk C between channels, and then use the obtained a c , a1 and C compensate the system; a1=R v / R h 、a c =(T v / T h ) / (R v / R h ), a1 represents the difference between the V polarization receiving channel and the H polarization receiving channel, a c It represents the ratio of the difference between the H and V transmitting channels to the difference between the H and V receiving channels, R h , R v , T h , T v They represent the transmission characteristics of H receiving channel, V receiving channel, H transmitting channel and V transmitting channel respectively; the specific method of using symmetry to solve the difference a1 caused by the system transmission link and the crosstalk C between channels is: Define the estimates of a1 and C as and Try to use the estimated and any and Compensation system: At the expense of reciprocity, establish the optimization function: Get the optimal solution, where and represents the simplified system error estimation result, ζ(·) represents the reciprocity degree of the scattering matrix, and i represents the i-th target.

2. The polarization calibration method according to claim 1, wherein: The scattering matrix obtained after the parameter decomposition is: Among them, M hh Indicates the echo obtained by H polarization transmission and H polarization reception, M vh Indicates the echo obtained by transmitting with H polarization and receiving with V polarization, M hv Indicates the echo obtained by transmitting with V polarization and receiving with H polarization, M vv It represents the echo obtained by V polarization transmission and V polarization reception, g = R h T h , g will act on all elements of the scattering matrix; S hh represents the echo obtained by H polarization transmission and H polarization reception, where S vh It represents the echo obtained by transmitting with H polarization and receiving with V polarization, where S hv represents the echo obtained by transmitting with V polarization and receiving with H polarization, where S vv Indicates the echo obtained by V polarization transmission and V polarization reception; a c It represents the ratio of the difference between H and V transmitting channels to the difference between H and V receiving channels; a1 represents the difference between V polarized receiving channel and H polarized receiving channel, and C represents the crosstalk between channels.

3. The polarization calibration method according to claim 1 or 2, characterized in that: The error a caused by the transmission link of the system is solved by using reciprocity c The method is: take the ratio of the echo obtained by V polarization transmission and H polarization reception to the echo obtained by H polarization transmission and V polarization reception as the error a c .

4. The polarization calibration method according to claim 1 or 2, characterized in that: The specific method of using symmetry to solve the error a1 caused by the system transmission link and the crosstalk C between channels is: first, according to the obtained error a c The system is compensated, and then an optimization function is established at the cost of reciprocity to obtain the optimal solution.

5. The polarization calibration method according to claim 4, characterized in that: The optimal solution is solved by using a genetic algorithm.

6. The polarization calibration method according to claim 4, characterized in that: When the solutions obtained by the optimization function are opposite to each other, the solution with the smallest difference between the phases corresponding to the two solutions and the initial phase of the error a1 is selected as the optimal solution.

Images