Method for suppressing ground clutter of motion platform
By combining MTI cascade FFT and digital comprehensive algorithm in radar signal processing, the problem of poor ground clutter suppression effect and large computing volume in traditional methods is solved, and effective clutter on the motion platform is achieved Suppression and computational efficiency improvement.
Patent Information
- Application Number
- CN202310657714.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-02
- Publication Date
- 2024-12-27
AI Technical Summary
When traditional methods suppress ground clutter on the motion platform, there are problems such as widening spectrum, high computational volume and high engineering difficulty, which makes it difficult to effectively suppress ground clutter, affecting the detection performance of the radar.
The adaptive filter is designed using the MTI cascade FFT method and digital comprehensive algorithm, combined with the bandpass filter bank, adaptively adjust the filter notch to suppress ground clutter, and use the MTI cascade FFT in the intermediate frequency band to reduce the calculation amount.
It effectively suppresses ground clutter under the motion platform, improves radar detection performance, reduces computing volume and engineering complexity, and has versatility and high computing efficiency.
Smart Images
Figure CN119199751A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of radar signal processing, and in particular to a method for suppressing clutter on a moving platform. Background Art
[0002] In the ever-changing battlefield environment of modern times, mobility, concealment, and survivability are basic requirements for all weapons and equipment on the battlefield. In particular, for ground intelligence radars that obtain battlefield air situation information, they are always facing the threat of anti-radiation missiles, and their survivability in actual combat is particularly important. At present, most ground intelligence radars adopt a fixed working mode, which has problems such as poor mobility, poor concealment, and limited working environment. Compared with fixed-station radars, vehicle-mounted ground intelligence radars with on-the-go detection technology can move flexibly and effectively avoid the attack of enemy precision-guided weapons by relying on their mobile detection performance, greatly reducing the possibility of being detected and attacked.
[0003] When the radar is placed on a moving platform, the ground clutter itself becomes "moving" clutter, with a certain Doppler frequency, which changes with the platform speed and the azimuth and elevation of the clutter unit relative to the radar. In order to ensure the clutter suppression effect, the traditional method is to compensate the radar platform movement speed in real time so that the Doppler frequency of the ground clutter received by the main lobe of the beam is approximately zero, and then use the conventional Doppler filtering method to suppress the ground clutter.
[0004] Although traditional methods can achieve real-time compensation for vehicle speed, the compensation is equivalent to changing the Doppler frequency of ground clutter in the direction of the maximum value of the main lobe of the radar beam to zero. The beam of the radar antenna has a certain width, and the Doppler frequency of ground clutter in other directions of the main lobe is not zero (near zero), which is equivalent to widening the spectral width of the radar clutter. The wider the main lobe of the radar antenna beam, the wider the spectrum of the radar clutter, and the worse the clutter suppression effect. At the same time, the real-time compensation of vehicle speed requires a large amount of calculation, and the project implementation is also difficult.
[0005] Therefore, a method is needed to suppress the ground clutter with spectrum broadening when the ground intelligence radar is working on the move, and at the same time it needs to have the characteristics of low computational complexity. Summary of the invention
[0006] In view of the deficiencies of the prior art, the present invention provides a method for suppressing ground clutter on a moving platform. The method of the present invention includes comprehensively utilizing the MTI cascade FFT method and the digital synthesis algorithm in the entire Doppler channel, and using the digital synthesis algorithm to design an adaptive filter in the bandpass filter group near the ground clutter. The notch of the filter changes adaptively according to the speed of the moving platform. The filter also has the characteristics of deep null and low side lobe, and can suppress ground clutter at different speeds of the moving platform. The other part of the bandpass filter group is designed by the MTI cascade FFT method to reduce the amount of calculation. The present invention has the advantages of versatility, high computational efficiency, and easy implementation in engineering, and has been successfully verified for use in a certain model of radar.
[0007] The technical solution of the present invention is: a method for suppressing clutter on a moving platform, comprising the following steps:
[0008] Step 1: Get the radar echo signal and set the number of coherent accumulation pulses to M.
[0009] The beneficial effect of this step is to use coherent pulse train echoes to suppress ground clutter and improve the target detection signal-to-noise ratio by distinguishing ground clutter from moving targets in the spectrum.
[0010] Step 2: Determine that the number of adaptive filters designed using the digital synthesis algorithm is 2N, and the number of filters of the MTI cascade FFT designed using the moving target display cascade fast Fourier transform method is M-2N, where N is less than or equal to
[0011] The beneficial effect of this step is that the filter group near the ground clutter uses a digital synthesis algorithm to design an adaptive filter, so that the pulse frequency response in this area has a high main lobe, a low side lobe and a deep null. The filter group in the intermediate frequency band uses the MTI cascade FFT design method, so that the filter within this frequency range has a small amount of calculation and can detect the target better. This method makes the final filter group have a better filtering effect and can effectively reduce the amount of calculation.
[0012] Step 3: Adaptive filters are designed for the first N filters and the last N filters respectively using a digital synthesis algorithm, and the M coherent accumulation pulses are filtered by the first N filters and the last N filters to obtain adaptive coherent accumulation results.
[0013] The beneficial effect of this step is to design an adaptive filter using a digital synthesis algorithm, which has a low sidelobe characteristic while suppressing clutter, so that it is conducive to suppressing ground clutter with spectrum broadening. The digital synthesis algorithm is based on the adaptive array theory, and a large number of interference signals (each interference signal has a corresponding frequency) are placed in the sidelobe area of the filter frequency response, and the sidelobe level of the filter frequency response is adaptively controlled by changing the power intensity of the interference signal. At the same time, according to the speed of the moving platform, the filter notch is adaptively adjusted so that the notch of the filter is adaptively aligned with the Doppler frequency corresponding to the antenna main lobe, that is, aligned with the antenna main lobe clutter, and the notch is also correspondingly widened, adapted to the spectrum broadening of the main lobe clutter, while widening the filter notch, the notch depth is still kept deeper, so that the frequency response of the adaptive filter in the clutter area has a high main lobe and a low side lobe and a deep null, so that the suppression of ground clutter under the moving platform can be effectively achieved.
[0014] Step 4: For the M-2N filters in the middle, the design method of the moving target display MTI cascade FFT filter is adopted (generally, MTI adopts 3-pulse cancellation), and then the frequency response of the M-2N MTI cascade FFT filters is obtained. The M coherent accumulation pulses are filtered through the M-2N MTI cascade FFT filters to obtain the conventional coherent accumulation results.
[0015] The beneficial effect of this step is that the filter group of MTI cascade FFT is used in the non-clutter area. The filter group does not need to be calculated in real time, but is calculated in advance and stored in the memory, and is read from the memory in real time during calculation. Therefore, this method greatly reduces the amount of calculation and complexity, which is convenient for engineering applications.
[0016] Step 5: Finally, the moving target detection filter group within the entire pulse repetition period is obtained, and the MTD result of the M-order filter is obtained. The first N filters and the last N FIR filters are the output frequency responses corresponding to the 2N digital synthesis algorithm adaptive filters, and the middle M-2N filters are the frequency responses of the M-2N MTI cascade FFT filters.
[0017] The beneficial effect of this step is that the two filter design methods are integrated, which can not only adaptively suppress ground clutter at different movement speeds, but also reduce complexity and computational complexity, thus facilitating engineering implementation.
[0018] The beneficial effect of the present invention is that a comprehensive moving target detection filter design method is adopted to suppress ground clutter when the vehicle-mounted ground intelligence radar is working while moving, so as to ensure the clutter suppression effect and improve the detection performance of the radar, thereby solving the problems of poor clutter suppression effect, large amount of calculation, and difficult engineering implementation caused by the traditional speed compensation method.
[0019] This comprehensive moving target detection filter design method is a method that comprehensively utilizes the MTI cascade FFT method and the digital synthesis algorithm in the entire channel, wherein the bandpass filter group near the ground clutter uses the digital synthesis algorithm, while the bandpass filter group of another part is designed using the MTI cascade FFT method. The filter group near the ground clutter uses the digital synthesis algorithm, so that the pulse frequency response in this area has a high main lobe, a low side lobe and a deep null. The filter group in the intermediate frequency band uses the MTI cascade FFT design method, so that the filter within this frequency range has a small amount of calculation and can detect the target better. This method makes the final filter group have a better filtering effect and can effectively reduce the amount of calculation. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 The present invention is a flow chart of steps 1 to 5 in a specific embodiment of a method for suppressing ground clutter on a moving platform based on a vehicle-mounted ground intelligence radar.
[0021] Figure 2 It is a schematic diagram of the filter bank obtained by using the MTI cascade FFT design method.
[0022] Figure 3 Frequency response diagram of the 45th channel of the 63rd-order MTD filter when v=10km / h.
[0023] Figure 4 Frequency response diagram of the 45th channel of the 63rd-order MTD filter at v=30km / h.
[0024] Figure 5 Frequency response diagram of the 45th channel of the 63rd-order MTD filter at v=50km / h.
[0025] Figure 6 Frequency response diagram of the 90th channel of the 127th-order MTD filter when v=10km / h.
[0026] Figure 7 Frequency response diagram of the 90th channel of the 127th-order MTD filter at v=30km / h.
[0027] Figure 8 Frequency response diagram of the 90th channel of the 127th-order MTD filter at v=50km / h. DETAILED DESCRIPTION
[0028] The preferred implementation examples of the present invention are described in detail below in conjunction with the accompanying drawings so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more definite definition of the protection scope of the present invention.
[0029] The embodiment of the present invention provides a method for suppressing ground clutter on a moving platform, which can be applied to the field of radar signal processing.
[0030] See also Figure 1 A method for suppressing clutter on a moving platform comprises the following steps:
[0031] Step 1: Obtain radar echo signal and set the number of coherent accumulation pulses to M;
[0032] Step 2: According to the set number of moving target detection filters M, the number of adaptive filters designed by digital synthesis algorithm is determined to be 2N, and the number of filters of MTI cascade FFT designed by moving target display cascade fast Fourier transform method is determined to be M-2N, where N is less than or equal to Because the number of filters is divided into at least three equal parts, preferably equal to
[0033] Step 3: Adaptive filters are designed for the first N filters and the last N filters using a digital synthesis algorithm. The digital synthesis algorithm is based on the adaptive array theory, which places interference signals in the sidelobe area and adaptively controls the sidelobe level of the antenna pattern by changing the power intensity of the interference signal. At the same time, the filter notch is adaptively adjusted according to the speed of the moving platform, so that the frequency response in the clutter area has a high main lobe, a low side lobe and a deep notch. The specific implementation steps are as follows:
[0034] Step 3a): Assume the radar pulse repetition period is T r , the repetition frequency is F r =1 / T r , the passband center frequency of the i-th filter in the filter bank is f i According to the adaptive filter theory, the weight vector W of the i-th filter is i for:
[0035] W i =R -1 a(f i )
[0036] Where a(f i )=[1,exp(-j2πf i T r ),exp(-j2πf i 2T r )…exp(-j2πf i (N-1)T r )] T is the pilot vector, j represents the imaginary part, R = R c +σ 2 I is the clutter plus noise covariance matrix, R cis the ground clutter covariance matrix, I is the identity matrix, σ 2 is the noise power, R -1 The purpose is to make the filter adaptively form a notch at the frequency of the clutter, thereby suppressing the clutter. In practice, it is generally assumed that the ground clutter power spectrum is a Gaussian spectrum with a spectrum center of zero. According to the form of the ground clutter spectrum, the autocorrelation function of the ground clutter can be obtained as: δ c 2 is the power spectrum variance of the ground clutter. Assuming that the vehicle-mounted platform of the radar moves at a speed of v, in a direction of θ1, and the main lobe of the antenna beam points to θ2, the Doppler frequency of the main lobe clutter is f c =2v / λcos(θ1-θ2), the change of velocity v will cause the Doppler frequency f c The change of Doppler frequency f c The change of will cause the change of autocorrelation function rc, the ground clutter covariance matrix R c The filter notch is adaptively adjusted according to the change of the autocorrelation function; where λ is the operating wavelength of the radar, so the covariance matrix R of the ground clutter is c for:
[0037]
[0038] Step 3b): Generate L interference signals and input them into the first N filters and the last N filters, with frequencies f i1 ,f i2 ,…f iL , and these interference frequencies cover the entire period of the Doppler frequency response, that is, the pulse repetition frequency F r Divide it into L equal parts. Usually L is an integer multiple of N. In general, L = 2N.
[0039] Step 3c): Assume that the sidelobe level of the required filter frequency response is lower than the main lobe peak D(f)dB, generally D(f)=60dB, a sufficiently low sidelobe level can effectively suppress the sidelobe of clutter and improve the clutter improvement performance of the filter;
[0040] Step 3d): Let k = 0, interference power σi1 2 (k)=σi2 2 (k)=…=σi L 2 (k) = 0, k represents the number of iterations;
[0041] Step 3e): Calculate the covariance matrix R of interference, clutter and noise a (k):
[0042]
[0043] Step 3f): Find the weight coefficients of the filter
[0044] Step 3g): Find the frequency response of the filter
[0045] Step 3h): Obtain the current expected interference level from the current frequency response main lobe peak
[0046] Step 3i): Calculate the interference power at the k+1th step:
[0047]
[0048] In the above formula [f left (k),f right (k)] is the main lobe area of the frequency response of the k-th filter, {[0,f z1 (k)]∪[f z2 (k), Fr]} is the null zone of the kth step (near zero frequency for suppressing ground clutter),
[0049] ηi l =σi l 2 (k)+K[h i (k,fi l )-d(k,fi l )], K is a constant, called iterative gain. If the K value is too large, the algorithm will become unstable. If the K value is too small, the convergence speed of the algorithm will be very slow. In the example of the present invention, K=8;
[0050] Step 3j): Let k = k + 1, jump to step 3e), and repeat the above steps until h i (k,f)=h i (k+1,f), that is, until the ideal filter frequency response no longer changes;
[0051] The frequency response of the first N filters and the last N filters is h i (k, f), the M coherent integration pulses are filtered by the first N filters and the last N filters to obtain the adaptive coherent integration result.
[0052] Step 4: For the middle M-2N filters, adopt the design method of the moving target display MTI cascade FFT filter (generally, MTI uses 3 pulse cancellation), and then obtain the frequency response of the M-2N MTI cascade FFT filters. Filter the M coherent accumulation pulses through the M-2N MTI cascade FFT filters to obtain the conventional coherent accumulation results. Compared with the calculation of the M-order adaptive filter, the conventional method in step 4 reduces the amount of calculation by nearly 30%, and the filter can be stored in the memory and called from the memory during the project implementation. Step 4 can use the conventional method;
[0053] Step 5: Combine the adaptive coherent accumulation result with the conventional coherent accumulation result to obtain the moving target detection result.
[0054] Figure 2 The simulation parameters are: the number of MTD filters N = 20, the pulse repetition frequency is 1000 Hz, and the variance of the ground clutter spectrum is δ c 2 is 4Hz, ground clutter power For 10 6 , Figure 2 Schematic diagram of the filter bank obtained using the MTI cascade FFT design method.
[0055] pass Figure 2 It can be clearly seen that the filter designed by the MTI cascade FFT method has relatively poor performance near zero frequency, and has a loss in detection performance for low-speed targets. It can also be found that the number of filters with poor performance near ground clutter accounts for about 1 / 3 of the total number of filter groups. The effectiveness of the selected N / 3 number of filters designed by the FIR method is also verified through simulation.
[0056] Figure 3-Figure 8 The simulation parameters are: radar azimuth beam width BW = 3°, elevation angle φ = 10°, radar azimuth beam pointing θ = 30°, aircraft speed v = 10km / h, 30km / h, 50km / h, carrier wavelength λ = 0.05, pulse repetition frequency 2000Hz, and variance δ of ground clutter spectrum. c 2 is 4Hz, ground clutter power For 10 6 , the MTD filter orders are N=63 and N=127 respectively. From the above simulation results, it can be seen that the weights of the MTD filter can be updated in real time according to the platform movement speed, and a notch is generated at the relative Doppler frequency of the ground clutter. As the platform movement speed increases, the notch of the filter is also appropriately widened. At the same time, the filter also has the characteristics of low sidelobe and deep null, and has the ability to effectively suppress ground clutter within the range of platform movement speed.
[0057] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A method for suppressing clutter on a moving platform, characterized in that: The following steps are involved: Step 1: Obtain radar echo signal and set the number of coherent accumulation pulses to M; Step 2: Determine that the number of adaptive filters designed using the digital synthesis algorithm is 2N, and the number of filters of the MTI cascade FFT designed using the moving target display cascade fast Fourier transform method is M-2N, where N is less than or equal to Step 3: Design the adaptive filter using digital synthesis algorithm for the first N filters and the last N filters. The steps are as follows: Step 3a): The radar pulse repetition period is T r , the repetition frequency is F r , the passband center frequency of the i-th filter in the filter bank is f i , then the weight vector W of the i-th filter i for: W i =R -1 a(f i ) Where a(f i )=[1,exp(-j2πf i T r ),exp(-j2πf i 2T r )…exp(-j2πf i (N-1)T r )] T is the pilot vector, whether j represents the imaginary part; R=R c +σ 2 I is the clutter plus noise covariance matrix, where R c is the ground clutter covariance matrix, I is the identity matrix, σ 2 is the noise power, The autocorrelation function of ground clutter is: δ c 2 is the power spectrum variance of the ground clutter. The vehicle-mounted platform of the radar moves at a speed of v, in a direction of θ1, and the main lobe of the antenna beam points to θ2. The Doppler frequency of the main lobe clutter is f c =2v / λcos(θ1-θ2); The covariance matrix R of ground clutter c for: Step 3b): Generate L interference signals and input them into the first N filters and the last N filters, with frequencies f i1 ,f i2 ,…f iL ; Step 3c): Assume that the sidelobe level of the required filter frequency response is D(f) dB lower than the main lobe peak; Step 3d): Let k = 0, interference power σi1 2 (k)=σi2 2 (k)=…=σi L 2 (k) = 0, k represents the number of iterations; Step 3e): Calculate the covariance matrix R of interference, clutter and noise a (k): Step 3f): Find the weight coefficients of the filter Step 3g): Find the frequency response of the filter Step 3h): Obtain the current expected interference level from the current frequency response main lobe peak Step 3i): Calculate the interference power at the k+1th step: In the above formula [f left (k),f right (k)] is the main lobe area of the frequency response of the k-th filter, {[0,f z1 (k)]∪[f z2 (k),Fr]} is the zero-sink area of the kth step, ηi l =σi l 2 (k)+K[h i (k,fi l )-d(k,fi l )], K is a constant; Step 3j): Let k = k + 1, jump to step 3e), and repeat the above steps until h i (k,f)=h i (k+1,f); The frequency response of the first N filters and the last N filters is h i (k, f), filtering the M coherent accumulation pulses through the first N filters and the last N filters to obtain the adaptive coherent accumulation result; Step 4: Filter the M coherent accumulation pulses through the filters of M-2N MTI cascade FFTs to obtain a conventional coherent accumulation result; Step 5: Combine the adaptive coherent accumulation result with the conventional coherent accumulation result to obtain the moving target detection result.
2. The method for suppressing clutter on a moving platform according to claim 1, characterized in that: N is equal to 3. The method for suppressing clutter on a moving platform according to claim 1, characterized in that: The L interference signal frequencies cover the entire period of the Doppler frequency response, and the pulse repetition frequency F r Divide it into L equal parts, usually L is an integer multiple of N.
4. The method for suppressing clutter on a moving platform according to claim 1, characterized in that: L=2N.
5. The method for suppressing clutter on a moving platform according to claim 1, characterized in that: The sidelobe level of the filter frequency response is 60dB lower than the main lobe peak.