Ground moving target refocusing method without parameter search

By constructing a distance frequency difference function in a synthetic aperture radar, eliminating the coupling of the distance frequency of the moving target and the slow time, the refocusing of the moving target energy at a low signal-to-noise ratio is achieved, and it is solved that it is difficult to correct the high at a low signal-to-noise ratio in the prior art. The problem of order distance migration improves calculation efficiency and focus effect.

CN119087377AActive Publication Date: 2024-12-06XIDIAN UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411153431.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-12-06
Estimated Expiration
2044-08-21

AI Technical Summary

Technical Problem

The prior art is difficult to achieve high-order distance migration correction and motion parameter estimation of dynamic motion targets under low signal-to-noise ratio conditions, and the traditional method has high computational complexity, so it is impossible to effectively deal with the distance caused by Doppler spectrum splitting and higher-order motion. Movement leads to a decrease in focus performance of moving targets.

Method used

Receive the moving target echo signal through the radar platform to compress the distance to pulse, convert it to the distance-frequency domain, construct the distance frequency difference function, and multiply it with the echo signal, eliminating the coupling between the moving target distance frequency and the slow time, realizing the The moving target energy refocusing without parameter estimates.

Benefits of technology

The stable distance migration correction of the moving target is achieved under the low signal-to-noise ratio, which reduces the computational complexity and improves the energy focusing effect of the moving target. It is suitable for Doppler central fuzzy and Doppler spectrum splitting.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119087377A_ABST
    Figure CN119087377A_ABST
Patent Text Reader

Abstract

The invention provides a ground moving target refocusing method without parameter search, which comprises the following steps of: constructing a distance frequency difference function without parameter estimation in a distance frequency domain according to moving target echo data received by a radar, and performing conjugate multiplication on the echo data and the distance frequency difference function to obtain a refocusing target. Coupling between the moving target distance frequency and the slow time can be eliminated, and the purpose of moving target energy refocusing is achieved. The distance frequency difference reference function completely comes from actual echoes, and extra parameters are not needed, so that the problems of complex distance envelope migration and Doppler defocus caused by high-order phase coupling can be well solved. According to the invention, the robust correction of the moving target range migration can be directly realized even under the conditions of platform motion error, moving target Doppler center ambiguity and Doppler spectrum splitting; according to the ground moving target refocusing method without parameter search, moving target refocusing can be realized without parameter estimation and search processing, and the calculation efficiency is high.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of moving target detection, and in particular relates to a ground moving target refocusing method without parameter search. Background Art

[0002] Synthetic aperture radar (SAR) has been widely used in various fields such as terrain mapping, maritime surveillance, environmental and disaster monitoring, and battlefield reconnaissance due to its all-weather observation capability. As one of the core tasks of radar detection, moving target detection plays a very important role in battlefield situation awareness and traffic control. Therefore, the ability to obtain real-time motion information of moving targets is very important in both military and civilian applications.

[0003] In the radar downward detection mission, the radar receives the echo signal of the weak moving target signal superimposed with the background clutter. Due to the movement of the platform, the Doppler spectrum of the background clutter becomes significantly wider. After the received data is transformed into the range-Doppler domain, the moving target echo signal will overlap with the clutter, making it impossible to detect the moving target directly.

[0004] Although the method of suppressing background clutter can retain the information of moving targets while suppressing clutter, improve the signal-to-noise ratio, and make moving targets easier to detect, SAR radar imaging requires a longer synthetic aperture time in order to obtain higher two-dimensional resolution capabilities. During this time, the complex relative motion between the moving target and the platform will cause range migration and Doppler spectrum diffusion. In addition, when imaging under fixed scene parameters, due to the mismatch of compensation functions, the moving target will inevitably be defocused, which will cause the peak energy of the moving target to drop seriously, which is not conducive to the detection of the moving target.

[0005] In order to solve the energy focusing problem of moving targets in moving target detection, range migration (RCM) correction must be performed. Most correction methods usually use the signal envelope autocorrelation characteristics and the minimum entropy criterion, which only perform well under high signal-to-noise ratio conditions and are difficult to apply in practice.

[0006] The Keystone transform (KT) method can be used to correct the first-order range motion under low signal-to-noise ratio conditions, but it cannot correct high-order motion. Although the second-order Keystone transform method can also be used to correct the second-order motion of moving targets, when the moving target has Doppler center blur or Doppler spectrum splitting, both methods require fuzzy search operations, resulting in a significant increase in computational complexity. In addition, these methods cannot correct the range motion caused by high-order motion, and will produce residual range motion that is not conducive to refocusing the moving target.

[0007] After the range migration is properly corrected, in order to accurately achieve the refocusing of the moving target, the motion parameters of the moving target need to be estimated. The moving target parameter estimation methods are mainly divided into two categories: linear time-frequency transformation method and nonlinear time-frequency transformation method.

[0008] The main linear time-frequency transform methods include maximum likelihood transform (ML), fractional Fourier transform (FRFT) and discrete Fourier transform (DCFT), but these methods require full-dimensional search and have high computational complexity.

[0009] The main nonlinear time-frequency transform methods include high-order ambiguity function (HAF), discrete polynomial phase transform (DPT), Wigner-Ville distribution (WVD) and scaled Fourier transform (SCFT) methods. These methods will encounter cross-terms of polynomial moving target signals, thus affecting the estimation of moving target parameters in multi-moving target scenarios. In the case of low signal-to-noise ratio, these methods are difficult to obtain moving target parameters.

[0010] From the above, we can see that in practical applications, it is difficult to obtain accurate parameters of moving targets, and these methods of obtaining moving target parameters are prone to large errors, which will lead to a decrease in the performance of moving target focusing methods that rely on parameter estimation. Therefore, there are also some methods that do not need to know the moving target parameters to achieve moving target focusing. These methods include two-dimensional frequency domain matched filter algorithm; De-skewed wedgestone transform (DerampKT, DKT) algorithm; Time Reversal Transform (Time Reversal Transform, TRT).

[0011] The two-dimensional frequency domain matched filter algorithm constructs a specific matched filter in the two-dimensional frequency domain to achieve moving target refocusing without estimating motion parameters. Since this fixed phase method has a large error when the time-bandwidth product (TBP) of the moving target is small, this method is only applicable to signals with a large TBP. If there is a significant error in the filter function, a large amount of residual RCM will be retained, which is not conducive to the refocusing of moving targets. The DerampKT algorithm uses the static scene information to construct an azimuth deramp function in the range-frequency domain to compensate for the range curvature and Doppler frequency modulation, and uses the KT algorithm to correct the RCM, and then focuses on the moving target in the azimuth frequency domain. Because the above two methods ignore the azimuth motion of the moving target, neither of them can compensate for the high-order phase effect caused by the motion of the complex moving target, and the moving target will inevitably be defocused. In response to the problems of the above two algorithms, the TRT method directly constructs a compensation function in the slow time domain of time reversal to compensate for the second-order Doppler phase of the moving target. Although the azimuth velocity of the moving target is taken into account in the TRT method, Doppler shift will still occur when the moving target has high-order motion such as acceleration. And because both the DerampKT algorithm and the TRT method use KT for preprocessing to correct the motion of the moving target, when there is motion error on the platform or high-order motion on the moving target, the diffusion of the range envelope will be more obvious in the high-resolution radar, making it impossible to effectively correct the range migration.

[0012] The existing technology for range motion correction only works well under high signal-to-noise ratio conditions, and fails under low signal-to-noise ratio conditions. Although some methods such as the Keystone transform (KT) method can complete the correction under low signal-to-noise ratio conditions, they can only handle first-order range motion correction and cannot correct the range motion caused by high-order motion. The algorithms for high-order range motion correction all require fuzzy search operations, which will significantly increase the computational complexity. In addition, these methods cannot correct the range motion caused by high-order motion and will also produce range motion residuals that are not conducive to focusing on the moving target.

[0013] After the range migration correction, in order to complete the moving target refocusing, the motion parameters of the moving target need to be estimated. Some existing algorithms for the motion parameter estimation of moving targets, such as the linear time-frequency transformation method, require full-dimensional search and have high computational complexity. The nonlinear time-frequency transformation method will encounter cross terms of polynomial moving target signals and cannot handle the moving target parameter estimation in multi-moving target scenarios. In addition, both methods are difficult to obtain the moving target parameters when the signal-to-noise ratio is low.

[0014] There are also moving target refocusing methods that do not require parameter estimation, such as the two-dimensional frequency domain matched filter algorithm, the DerampKT algorithm, and the TRT algorithm. The two-dimensional frequency domain matched filter algorithm is only applicable to signals with large TBP, and the filter function must be very accurate, otherwise a large amount of residual RCM will be generated. The DerampKT algorithm cannot compensate for the high-order phase effects caused by the complex motion of the moving target, and the moving target will inevitably be defocused. When the TRT method is used, Doppler frequency shift will occur when the moving target has high-order motion such as acceleration, and the moving target cannot be refocused. Summary of the invention

[0015] In order to solve the above problems existing in the prior art, the present invention provides a ground moving target refocusing method without parameter search. The technical problem to be solved by the present invention is achieved by the following technical solutions:

[0016] The present invention provides a ground moving target refocusing method without parameter search, comprising:

[0017] S100, receiving an echo signal of a moving target through a radar platform, and performing range-direction pulse compression on the echo signal to obtain a pulse-compressed signal;

[0018] S200, converting the pulse compressed signal into a distance-frequency domain to obtain a distance-compressed signal in the distance-frequency domain;

[0019] S300, constructing a distance-frequency difference function according to the distance compression signal;

[0020] S400, obtaining a moving target corrected signal according to the range-frequency difference function and the range compression signal in the range-frequency domain;

[0021] S500, performing inverse Fourier transform on the corrected signal of the moving target in the range direction, and then performing Fourier transform on the azimuth direction to obtain refocused image data of the moving target.

[0022] Beneficial effects:

[0023] 1. The ground moving target refocusing method proposed by the present invention does not require parameter search. According to the moving target echo data received by the radar, a distance frequency difference function without parameter estimation is constructed in the distance frequency domain. By multiplying the echo data by the conjugate of the distance frequency difference function, the coupling between the moving target distance frequency and the slow time can be eliminated, so as to achieve the purpose of moving target energy refocusing. Since the distance frequency difference reference function is completely derived from the actual echo and does not require additional parameters, it can well solve the complex distance envelope migration and Doppler defocusing problems caused by high-order phase coupling.

[0024] 2. Different from the traditional kt-based moving target focusing method, the ground moving target refocusing method proposed in the present invention without parameter search can directly achieve robust correction of moving target range migration even in the presence of platform motion error, moving target Doppler center ambiguity and Doppler spectrum splitting.

[0025] 3. Compared with the traditional method, the ground moving target refocusing method without parameter search of the present invention can achieve moving target refocusing without parameter estimation and search processing, and has high calculation efficiency.

[0026] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of a ground moving target refocusing method without parameter search provided by the present invention;

[0028] Figure 2 It is a flow chart of a radar optoelectronic integration method for detecting and tracking low, slow and small moving targets provided by the present invention;

[0029] Figure 3 This is a simulation result diagram of detecting a slow-moving target provided by the present invention;

[0030] Figure 4 This is a simulation result diagram of detecting fast-moving targets provided by the present invention;

[0031] Figure 5 These are three different moving target refocusing effect diagrams provided by the present invention;

[0032] Figure 6 It is a simulation result diagram of each algorithm of the slow-motion target 1 provided by the present invention;

[0033] Figure 7 It is a simulation result diagram of each algorithm of the moving target 2 with a relatively large speed provided by the present invention;

[0034] Figure 8 It is a simulation result diagram of each algorithm of fast moving target 3 provided by the present invention; DETAILED DESCRIPTION

[0035] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0036] For a single-channel synthetic aperture radar ground moving target detection (SAR-GMTI) system, the present invention performs range pulse compression on the moving target echo signal received by the radar, converts the signal after the range pulse compression to the range-frequency domain, constructs a corresponding specific range-frequency difference function for slow-moving targets and fast-moving targets according to the range-frequency domain data in the range-frequency domain, and then multiplies the range-frequency difference function with the moving target echo to achieve range migration correction. At the same time, by setting an appropriate frequency deviation, the azimuth signal can be effectively expressed as a small linear frequency modulation (LFM) signal function, which can simplify the analysis process. Finally, the processed data is subjected to inverse Fourier transform in the range direction and Fourier transform in the azimuth direction to be converted to the range-Doppler domain, so as to achieve focusing on the moving target.

[0037] like Figure 1 As shown, the present invention provides a ground moving target refocusing method without parameter search, comprising:

[0038] S100, receiving an echo signal of a moving target through a radar platform, and performing range-direction pulse compression on the echo signal to obtain a pulse-compressed signal;

[0039] This step uses the geometric relationship between the radar platform and the moving target to establish the echo signal expression of the moving target observed by the zero-slant side-looking radar platform, which is expressed as:

[0040]

[0041] In the formula, s(t,t m ) represents the echo signal of the moving target, σ s is the complex scattering coefficient of the moving target, w(t m ) is the azimuth window function, t is the fast time, R s (t m ) is the instantaneous slant distance between the platform and the moving target, T p is the pulse duration, K r is the pulse modulation frequency, c is the speed of light, and λ is the carrier wavelength;

[0042] Instantaneous slope distance R s (t m ) is expressed as:

[0043]

[0044] Where v represents the platform speed, t m It is slow time, v a represents the velocity in the azimuth direction, a a represents the acceleration in the azimuth direction, R B represents the initial slant distance between the platform and the moving target, vc represents the speed in the distance direction, a c Represents the acceleration in the distance direction.

[0045] The signal after pulse compression without noise is expressed as:

[0046]

[0047] Where G is the distance compression gain and B is the bandwidth of the transmission signal.

[0048] Considering Gaussian white noise, the signal after pulse compression in the presence of noise is expressed as:

[0049]

[0050] In the formula, n(t,t m ) represents Gaussian white noise in the range time domain and azimuth time domain.

[0051] S200, converting the pulse compressed signal into a distance-frequency domain to obtain a distance-compressed signal in the distance-frequency domain;

[0052] This step converts the pulse compressed signal into the distance-frequency domain to obtain distance compression signals in the distance-frequency domain without noise and with noise;

[0053] If the noise is not considered, the distance compression signal in the distance-frequency domain without noise is expressed as:

[0054]

[0055] In the formula, A t is the amplitude of the echo signal of the moving target in the frequency domain, P(f r )=rect(f r / B) is the Fourier transform of the envelope of the linear frequency modulation signal that approximately satisfies the rectangular envelope, f r is the range frequency, f c is the carrier frequency, rect(x) is the rectangular window function, expressed as x is the input data;

[0056] If the noise (Gaussian white noise) is considered, the distance compression signal converted to the distance-frequency domain in the presence of noise is expressed as:

[0057]

[0058] In the formula, n(f r ,t m ) represents Gaussian white noise in the range frequency domain and azimuth time domain.

[0059] S300, constructing a distance-frequency difference function according to the distance compression signal;

[0060] This step constructs the distance-frequency difference function in the range-frequency domain and the azimuth time domain when there is no noise and when there is noise according to the distance compression signal;

[0061] The range-frequency difference function in the range-frequency domain and the azimuth time domain in the absence of noise is expressed as:

[0062]

[0063] When there is noise, the range-frequency difference function in the range-frequency domain and the azimuth time domain is expressed as:

[0064]

[0065] Where Δf represents the distance frequency interval n(f r -Δf,t m ) represents Gaussian white noise in the range frequency domain and azimuth time domain with frequency difference.

[0066] Where Δf represents the distance frequency interval, which satisfies the following conditions:

[0067]

[0068] In order to avoid range spectrum aliasing Δf, the following formula needs to be satisfied:

[0069] |Δf|<F s -B;

[0070] Δf≠0;

[0071] In the formula, F s / N is the range frequency sampling interval, N is the number of range gate sampling points, T a is the synthetic aperture time, F s is the sampling frequency, and B is the signal bandwidth.

[0072] At the same time, in order to avoid the coherent accumulation of noise in the range-time domain, Δf cannot be zero. Because the radar signal is discretely sampled after reception, in order to reduce the frequency shift error and ensure that the reference signal and the original moving target signal do not completely overlap in the range frequency domain, Δf should be the range frequency sampling interval F. s / an integer multiple of N, that is, Δf = DF s / N; N is the number of range gate sampling points, D is a positive integer, T a is the synthetic aperture time, F s is the sampling frequency, and B is the signal bandwidth.

[0073] It is worth noting that the distance-frequency difference function does not require prior knowledge such as the moving target and platform motion parameters, so it is easy to construct in practice.

[0074] S400, obtaining moving target corrected data according to the range-frequency difference function and the range compression signal in the range-frequency domain;

[0075] In this step, the conjugate function is taken for the range-frequency difference function, and the conjugate function is multiplied by the range compression signal in the range-frequency domain in S200 to obtain the echo signal after the moving target correction, which is expressed as:

[0076]

[0077] In the formula, A t is the amplitude of the moving target signal in the frequency domain, w(t m ) is the azimuth windowing function, P * It represents the conjugate function.

[0078] S500, the corrected signal of the moving target is subjected to inverse Fourier transform in the range direction, and then subjected to Fourier transform in the azimuth direction to obtain a refocused image of the moving target, which is expressed as:

[0079]

[0080] In the formula, A t ′ is the amplitude of the echo signal of the moving target in the range-time domain, represents the convolution operation, f a is the azimuth frequency, in the range of -PRF / 2≤f a ≤PRF / 2, PRF is the pulse repetition frequency emitted by the radar. E(f a ) represents a small linear frequency modulation signal Compared with the original echo signal of the moving target, the azimuth frequency modulation of the moving target is significantly reduced by a factor Δf / f c , which means that the Doppler support area of ​​the moving target has been greatly reduced.

[0081] The effect of the present invention is verified by simulation experiments below.

[0082] The simulation parameters for slow moving targets are shown in Table 1. The range-frequency difference function s(f r -Δf,t m ), the corresponding frequency interval Δf is selected as 5MHz. After using the processing method proposed by the present invention, the signal azimuth angle signal is approximately equivalent to a linear frequency modulation signal with a smaller modulation frequency. For the slow moving target simulation, the specific calculation value of the minor modulation frequency is γ new =2Δf((vva ) 2 -a c R B ) / cR B =0.095Hz / s 2 , and the original modulation frequency: γ=2f0((vv a ) 2 -a c R B ) / cR B =95.23Hz / s 2 ,Since Δf is much smaller than f0, the frequency modulation rate after processing has been greatly reduced, which means that the Doppler spread of the moving target is greatly reduced after RFD processing.

[0083] Table 1 Simulation parameters of radar system and moving target

[0084]

[0085] Figure 3 Figure (a) shows the trajectory of the moving target after range compression, with a cross-track speed of 9 m / s and a along-track speed of 10 m / s. Figure 3 It can be seen that the moving target trajectory is not a straight line, and the moving target energy spans several distance units. After the processing method proposed by the present invention, the following is obtained: Figure 3 The moving target image shown in Figure (b) shows that the moving target is well focused.

[0086] For the detection of fast moving targets by single-channel synthetic aperture radar ground moving targets, the simulation parameters are the same as those for slow moving targets, and a specific range-frequency difference function is constructed to achieve range migration correction.

[0087] For fast moving targets, the Doppler frequency of the moving target will exceed the radar pulse repetition frequency PRF. Then the Doppler center of the moving target will be blurred. At this time, there are two different situations in the moving target spectrum after sampling along the azimuth direction: one is a complete spectrum, and the other is a spectrum divided into two different fuzzy numbers. A1~A2 is the complete spectrum processing process, and B1~B2 is the processing process of two different fuzzy number spectra.

[0088] A1, for the complete spectrum, assuming that the moving target spectrum is not divided into two parts, and the moving target is located in the mth PRF of the azimuth frequency axis, the range-frequency domain and azimuth-time domain functions of the moving target signal received by the radar can be rewritten as:

[0089]

[0090] In the formula, M is a fuzzy number, and the specific value can be M=…,-2,-1,1,2….

[0091] A2: Construct the range-frequency difference function s of the moving target signal received by the radar in S3 fast1 (f r -Δf,t m ), and multiply the moving target signal by the conjugate function of the range-frequency difference function After the correction, the distance-frequency domain function of the signal is expressed as follows:

[0092]

[0093] From the above formula, we can see that the coupling effect between range frequency and azimuth time (including linear components and high-order components) has been completely removed. This shows that even if there is Doppler center ambiguity in the moving target, the proposed method still only needs to construct the RFD function to realize the range envelope migration correction of the moving target without the need for search operation.

[0094] B1: The spectrum of moving targets with different fuzzy number spectra needs to be divided into two parts. The range-frequency domain and azimuth-time domain functions of the moving target signal received by the radar can be expressed as:

[0095]

[0096] Among them, w1(t m ) and w2(t m ) represent the azimuth envelope of the mth PRF and m+1th PRF parts of the moving target signal on the azimuth frequency axis respectively.

[0097] B2: According to the time-frequency relationship, the moving target signal is also separated into a supporting envelope w1(t m ) and w2(t m ), construct the range-frequency difference function in each azimuth pulse separately, and add the radar receiving signal in B1 to the complex conjugate of the constructed range-frequency difference function After multiplication, we can get the processed signal:

[0098]

[0099] Where w(t m ) is w1(t m ) and w2(t m ) combination.

[0100] The simulation results of detecting fast moving targets are shown in Figure 4 As shown, Figure 4 The simulation results for two different types of fast-moving targets are shown in Figure . Figure 4Figure (a) shows a moving target after range compression, with a cross-track speed of 30 m / s (Doppler center ambiguity number of 1) and a down-track speed of 10 m / s. It can be seen from the figure that within the synthetic aperture time, the signal energy spans a large number of range units. After the moving target energy refocusing algorithm of the present invention, the following is obtained: Figure 4 The moving target image shown in (b) shows that the fast-moving target is well focused. Figure 4 Figure (c) shows the trajectory of a moving target with a cross-track speed of 13 m / s after range pulse compression. Figure 4 As shown in Figure (d), the Doppler spectrum of the moving target is divided into two parts, and the two parts correspond to different fuzzy numbers. After the refocusing method of the present invention is used for focusing, the following is obtained: Figure 4 The focused moving target shown in Figure (e).

[0101] Measured data simulation example:

[0102] The airborne radar has three channels along the range direction. The first channel transmits signals and all three channels receive echoes. The radar is set to work in the x-band broadband mode. The relevant system parameters are shown in Table 2.

[0103] Table 2 Measured data simulation parameters

[0104] Parameter / Unit Value Carrier frequency / GHz 8.85 Bandwidth / MHz 40 Pulse repetition frequency / Hz 1000 Slope distance / m 9000 Platform speed / (m / s) 120 Sampling frequency / MHz 72 Number of channels 3 Synthetic aperture size / m 0.56 Number of pulses 3072 Distance Units 500

[0105] Figure 5 The middle (a) figure is the image of the original echo data converted to the range-Doppler domain. Figure 5 It can be seen that the moving target is blocked by strong background clutter. Figure 5 The middle (b) figure shows the clutter suppression result, which effectively suppresses the strong background clutter and retains the moving target. Figure 5 Three moving targets with different motion characteristics can be obtained, such as Figure 5 As shown in the yellow oval in Figure (b).

[0106] Assume that the target speed does not exceed 120km / h and the acceleration is less than 8m / s 2 , substitute the system and motion target parameters into the formula

[0107]

[0108] The range frequency interval for obtaining the focused image of the moving target should be less than 16.7MHz, and the range frequency sampling interval F is calculated. s / N is 0.2MHz, and the selectable range of Δf is between 0.2MHz and 16.7MHz. The larger the Δf is, the lower the accumulated energy of the moving target is. In order to minimize the loss of energy accumulation of the moving target, Δf can be selected to be 0.4MHz.

[0109] The simulation results of each algorithm for slow-motion target 1 are as follows Figure 6 As shown in (a)-(k); the motion trajectory of the slow-moving target 1 after pulse compression is as follows Figure 6 As shown in Figure (a). Figure 6 (b) Figure and Figure 6 The middle (c) figure shows the result of azimuth FFT after KT processing. It can be seen that the energy of the moving target is diffused in both distance and azimuth dimensions. Figure 6 Figure (d) and Figure 6 Figure (e) shows the processing result after the two-dimensional frequency matching algorithm. It can be seen that the refocusing effect of the moving target is improved to a certain extent, but due to the inability to compensate for high-order phase modulation and residual range envelope migration, the diffusion of the moving target energy still exists. Figure 6 Figure (f) and Figure 6 The middle (g) figure shows the result of the moving target after DKT processing. Due to the mismatch of high-order terms in azimuth, the moving target appears defocused; Figure 6 Middle (h) Figure and Figure 6 Figure (i) shows the processing result of the TRT method. Although this method reduces the mismatch caused by the azimuth velocity of the moving target, the high-order phase error caused by the non-ideal motion still leads to azimuth defocus. Figure 6 As shown in Figure (k), it can be seen from the figure that the algorithm proposed in the present invention can effectively compensate for high-order phase errors and correct non-ideal range migration, and the moving target 1 has a good refocusing effect in the range-Doppler domain.

[0110] The simulation results of each algorithm for moving target 2 with a larger speed are as follows Figure 7 (a)-7(k) shows the trajectory of the moving target 2 with a larger speed in the distance after distance compression. Figure 7 As shown in the middle figure (a), since the range velocity of this moving target is relatively large, a more obvious range migration is observed. Figure 7 (b) Figure and Figure 7 The middle (c) figure is the result of azimuth FFT processing. It can be seen from the figure that the energy of the moving target is obviously diffused in two dimensions, which is not conducive to the detection of the moving target. Figure 7 Figure (d) and Figure 7 The middle (e) figure is the result of the two-dimensional frequency matching filter algorithm. Figure 7 Figure (f) and Figure 7 Figure (g) is the processing result of the DKT method. It can be seen that due to the rapid motion of the moving target, the mismatch when constructing the matched filter is more serious than that of the slow moving target 1. The results of the TRT method are as follows Figure 7 Middle (h) Figure and Figure 7As shown in Figure (i), the TRT method cannot process third-order and higher phase errors, resulting in azimuth defocusing. The processing results of the algorithm proposed in the present invention are shown in Figure 7 Figure (j) and Figure 7 As shown in the middle figure (k), it can be seen that after processing, it has a better energy focusing effect on the moving target.

[0111] The simulation results of each algorithm for fast moving target 3 are as follows Figure 8 (a)-8(k) shows the motion trajectory of the fast moving target 3. Figure 8 As shown in the middle (a) figure, the Doppler blur of the moving target 3 is -1. It can be seen that the moving target has obvious range migration, resulting in the inability to refocus the energy of the moving target. Figure 8 Middle (b) to middle (c) are the direct azimuth FFT processing results of the moving target without focusing the energy; Figure 8 Middle (d) to middle (e) are the processing results of the two-dimensional frequency matched filtering algorithm. The results show that due to the fast speed of the moving target and insufficient range migration correction, the SAR image of the moving target appears blurred. Figure 8 Figures (f) to (i) describe the processing results of the DKT and TRT methods. From the figures, we can get a result that although the first-order motion can be corrected by combining Doppler fuzzy search and KT processing, there is still obvious inaccuracy when compensating for the high-order phase error of the azimuth angle. Compared with moving target 2, this difference leads to a larger Doppler energy dispersion of the moving target. In addition, in actual situations, since the speed of non-joint moving targets is unknown, the DKT method and the TRT method need to search for Doppler fuzzy numbers in order to effectively accumulate the energy of the moving target. Therefore, the KT processing and Doppler fuzzy search operation significantly increase the computational complexity of the two algorithms; Figure 8 Middle (j) to middle (k) show the processing results of the method of the present invention, which are consistent with the simulation results. The method can achieve accurate correction of the motion of the moving target and focusing of the moving target without the need for a search operation, even in the case of Doppler center blur.

[0112] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A method for refocusing a ground moving target without parameter search, characterized in that: include: S100, receiving an echo signal of a moving target through a radar platform, and performing range-direction pulse compression on the echo signal to obtain a pulse-compressed signal; S200, converting the pulse compressed signal into a distance-frequency domain to obtain a distance-compressed signal in the distance-frequency domain; S300, constructing a distance-frequency difference function according to the distance compression signal; S400, obtaining a moving target corrected signal according to the range-frequency difference function and the range compression signal in the range-frequency domain; S500, performing inverse Fourier transform on the corrected signal of the moving target in the range direction, and then performing Fourier transform on the azimuth direction to obtain refocused image data of the moving target.

2. The method for refocusing a ground moving target without parameter search according to claim 1, characterized in that: S100 includes: S110, using the geometric relationship between the radar platform and the moving target, establish an expression for the echo signal of the moving target observed by the zero-slant-angle side-looking radar platform, expressed as: In the formula, s(t,t m ) represents the echo signal of the moving target, σ s is the complex scattering coefficient of the moving target, w(t m ) is the azimuth window function, t is the fast time, R s (t m ) is the instantaneous slant distance between the platform and the moving target, T p is the pulse duration, K r is the pulse modulation frequency, c is the speed of light, and λ is the carrier wavelength; Instantaneous slope distance R s (t m ) is expressed as: Where v represents the platform speed, t m It is slow time, v a represents the velocity in the azimuth direction, a a represents the acceleration in the azimuth direction, R B represents the initial slant distance between the platform and the moving target, v c represents the speed in the distance direction, a c Represents the acceleration in the distance direction. S120, performing range-direction pulse compression on the echo signal to obtain pulse compressed signals with and without noise.

3. The method for refocusing a ground moving target without parameter search according to claim 2, characterized in that: The signal after pulse compression when there is no noise in S120 is expressed as: In the formula, G is the distance compression gain, and B is the bandwidth of the transmission signal; When there is noise, the signal after pulse compression is expressed as: In the formula, n(t,t m ) represents Gaussian white noise in the range time domain and azimuth time domain.

4. The method for refocusing a ground moving target without parameter search according to claim 3, characterized in that: S200 includes: Converting the pulse compressed signal to the range-frequency domain to obtain range compression signals in the range-frequency domain without noise and with noise; The range compression signal converted to the range-frequency domain in the absence of noise is expressed as: Where At is the amplitude of the echo signal of the moving target in the frequency domain, P(fr)=rect(fr / B) is the Fourier transform of the envelope of the linear frequency modulation signal that approximately satisfies the rectangular envelope, and f r is the range frequency, f c is the carrier frequency, rect(x) is the rectangular window function, expressed as x is the input data; The range compression signal converted to the range-frequency domain in the presence of noise is expressed as: In the formula, n(f r ,t m ) represents Gaussian white noise in the range frequency domain and azimuth time domain.

5. The method for refocusing a ground moving target without parameter search according to claim 4, characterized in that: S300 includes: Constructing a range-frequency difference function in the range-frequency domain and the azimuth time domain when there is no noise and when there is noise according to the range compression signal; The range-frequency difference function in the range-frequency domain and the azimuth time domain in the absence of noise is expressed as: When there is noise, the range-frequency difference function in the range-frequency domain and the azimuth time domain is expressed as: Where △f represents the distance frequency interval, n(f r -△f,t m ) represents Gaussian white noise in the range frequency domain and azimuth time domain with frequency difference.

6. The method for refocusing a ground moving target without parameter search according to claim 5, characterized in that: The distance frequency interval △f satisfies the following conditions: △f|<F s -B; △f≠0; △f=DF s / N; In the formula, F s / N is the range frequency sampling interval, D is a positive integer, N is the number of range gate sampling points, T a is the synthetic aperture time, F s is the sampling frequency, and B is the signal bandwidth.

7. The method for refocusing a ground moving target without parameter search according to claim 6, characterized in that: S400 includes: The conjugate function is taken for the range-frequency difference function in the conversion domain, and the conjugate function is multiplied by the range compression signal in the range-frequency domain in S200 to obtain the echo signal after the moving target correction, which is expressed as: In the formula, A t is the amplitude of the moving target signal in the frequency domain, w(t m ) is the azimuth windowing function, P * It represents the conjugate function.

8. The method for refocusing a ground moving target without parameter search according to claim 6, characterized in that: The refocused image data of the moving target in S500 is expressed as: In the formula, A t ′ is the amplitude of the echo signal of the moving target in the range-time domain, represents the convolution operation, E(f a ) represents a small linear frequency modulation signal The Fourier transform of .

Citation Information

Patent Citations

  • Imaging method of synthetic aperture radar in large squint angle mode

    CN103576147A

  • SAR two-dimension autofocus method based on prior phase structure knowledge

    CN103792534A

  • SAR imaging method targeted at optional ground moving target

    CN106772373A

  • Parameter estimation method and system for unmanned aerial vehicle cooperative TDOA / FDOA composite positioning

    CN112114296A

  • Ultrahigh-speed target echo signal processing method, system and device

    CN114527442A