Airborne meteorological radar space-time combined ground clutter suppression method and device and medium

Abstract

The invention discloses an airborne meteorological radar space-time combined ground clutter suppression method and device and a medium, and the method comprises the steps: firstly, constructing an airborne meteorological radar pitching multichannel array signal receiving model, carrying out the distance pulse pressure processing of each channel data, and obtaining the distance-pulse-pitching array three-dimensional echo data; secondly, Doppler velocity offset of each distance-pulse unit is calculated, and phase difference caused by corresponding multi-spectrum offset is compensated in coherent processing time; secondly, designing an elliptical low-pass filter, filtering meteorological signal components in echo data to obtain a data sample for estimating an interference and noise covariance matrix, and calculating a weight vector of each distance-pulse unit; and finally, returning to the airspace to carry out adaptive beam forming, and outputting a ground clutter suppression result. According to the method, the number of required samples is smaller, the robustness is higher, on the premise that the ground clutter is effectively suppressed, the meteorological signal component is hardly lost, and the ground clutter suppression effect is better.

Description

Technical Field

[0001] The present invention belongs to the field of airborne meteorological radars, and in particular relates to a space-time combined ground clutter suppression method, device and medium for airborne meteorological radars. Background Art

[0002] Airborne weather radar is mainly used to detect dangerous weather such as thunderstorms and severe convection on the flight route in real time, provide weather warnings to pilots in advance, and plan safe flight paths in time. Usually in the downward-looking mode, the beam will inevitably illuminate the ground, causing the weather echo to be submerged in the wide-range and strong ground clutter, which has a significant impact on the detection and identification of meteorological targets, and may lead to a significant increase in false alarm rate and false detection rate, and even endanger flight safety. Therefore, the research on ground clutter suppression technology is of great significance. In meteorological radar, ground clutter can be suppressed from multiple dimensions, and meteorological echoes and ground clutter can be separated in the fast time domain, such as beam multi-scanning method, ground clutter suppression method based on terrain database, by changing the downward angle of the radar, receiving and separating weather and ground at different range gates, but when the weather and ground are in the same range gate, the suppression effect is not ideal. In the Doppler domain, due to the difference in radar echoes between meteorological targets and ground targets, the ground clutter spectrum is near zero frequency and has a narrow spectrum width, while the meteorological target spectrum is far from zero frequency and has a wide spectrum width. For example, the classic Gaussian model adaptive method, but due to the motion of the airborne platform, the echo signal spectrum is broadened, which makes this method difficult to apply to airborne radars. In the airspace, the difference in the pitch angle between the ground and the weather relative to the airborne radar is used, and an adaptive beamforming algorithm is used on the equidistant gate to form a depression at the angle of the ground clutter. At present, there are several robust beamforming algorithms to improve the algorithm robustness of the adaptive beamforming algorithm when the expected signal exists in the training sample, such as the diagonal loading algorithm of the sample covariance matrix (LSMI), the covariance matrix gradient method (CMT), the worst case optimal algorithm (WPO), etc. However, a serious disadvantage of the diagonal loading technique and the covariance matrix gradient method is that there is no closed solution, and there is no reliable method to select the diagonal loading coefficient. If this parameter is not selected properly, the robustness of the diagonal loading method may be insufficient. However, the two load factors of the optimal algorithm in the worst case cannot obtain closed solutions and can only be set manually, which is a disadvantage. In addition, the convergence speed of the algorithm still cannot meet the needs of real-time processing.

[0003] In the spatial adaptive beamforming method, the desired signal often exists in the training samples, that is, the meteorological target is often submerged in the ground clutter background, which will affect the accuracy of the covariance matrix estimation, and then affect the ground clutter suppression performance. As the meteorological signal becomes stronger, the ground clutter suppression performance will deteriorate significantly, and the traditional method is less robust to this situation. In addition, the number of samples will also affect the performance of the ground clutter suppression algorithm. Considering the real-time requirements of airborne meteorological radars, it is necessary to maintain good performance with fewer samples, so both problems need to be solved. Summary of the invention

[0004] Purpose of the invention: The present invention provides a space-time combined ground clutter suppression method, device and medium for airborne meteorological radar, which can improve the ground clutter suppression effect.

[0005] Technical solution: The method for suppressing ground clutter by combining space and time in an airborne meteorological radar according to the present invention comprises the following steps:

[0006] (1) Construct an airborne meteorological radar elevation multi-channel array signal receiving model. Each channel data is first processed by range pulse compression to obtain the three-dimensional echo data of the range-pulse-elevation array.

[0007] (2) Calculate the Doppler velocity offset of each range-pulse unit Compensate for the phase difference caused by the corresponding multi-spectrum offset during the coherent processing time;

[0008] (3) Design a low-pass filter, adjust the corresponding parameters, perform Doppler domain filtering on the range-pulse data of each channel, and perform group filtering on the pulse dimension to filter out the meteorological signal and eliminate the influence of the meteorological signal on the interference plus noise covariance matrix estimation;

[0009] (4) Using the data after filtering out the meteorological signal as samples, the interference plus noise covariance matrix is ​​estimated to calculate the weight vector required by the adaptive beamforming algorithm, and the weight vector is returned to the spatial domain for adaptive filtering. The estimated weight vector and the original data of step (1) are used to perform adaptive beamforming to suppress ground clutter.

[0010] Furthermore, the implementation process of step (1) is as follows:

[0011] A uniform linear array with one transmit and multiple receive is used to scan the area in front of the airborne radar. A linear frequency modulated pulse signal is emitted at a pulse repetition frequency. The reflected echo is received by the array antenna, and the echo is sampled to obtain multi-channel sampling data to complete the scan of the entire area. After distance pulse compression, the three-dimensional echo data of the range-pulse-elevation array is obtained.

[0012] Furthermore, the three-dimensional echo data of the range-pulse-elevation array is subjected to dimensionality reduction processing, and sum and difference beam data are synthesized to obtain sum or difference-range-pulse data.

[0013] Furthermore, the implementation process of step (2) is as follows:

[0014] The corresponding pitch angle is calculated according to the aircraft altitude and the distance of each clutter block. The Doppler shift corresponding to each range-pulse unit of each channel is calculated using the aircraft movement speed, radar wavelength, azimuth and pitch angle of each clutter block. The Doppler shift of the (i, j)th range-pulse unit is:

[0015]

[0016] Among them, v a is the aircraft flying speed, θ i , are the elevation and azimuth angles of the (i,j)th range-pulse unit, θ i The calculation method is:

[0017] θ i =arcsin(H / R i,j )

[0018] Among them, H is the flight altitude of the aircraft, R i,j is the distance of the (i, j)th range-pulse ground clutter unit relative to the radar; the elevation angles of the N range gates θ = [θ1, θ2, …, θ N ] T , M pulse azimuths Then the Doppler shift of the entire scanning area is:

[0019]

[0020] in, represents the Kronecker product; then the Doppler compensation result of the (i, j)th range-pulse ground clutter unit data of a certain channel is for:

[0021]

[0022] Among them, x i,j is the data of the (i, j)th range-pulse unit of a channel before Doppler compensation, f r is the pulse repetition frequency, and t is the coherent processing time.

[0023] Furthermore, the low-pass filter in step (3) is an elliptical low-pass filter.

[0024] Furthermore, the group filtering process for the pulse dimension in step (3) is to select 32 or 64 pulses as a group; for the interval scanning working mode, the number of pulses resident at each wave position is used as a group.

[0025] Furthermore, the implementation process of step (4) is as follows:

[0026] Use the data after filtering out the meteorological data as samples to estimate the interference plus noise covariance matrix but:

[0027]

[0028] in, To filter out the multi-channel data after the meteorological signal, K represents the number of samples taken. As the number of samples K increases, the sample covariance matrix Converges to the theoretical covariance matrix R c+n , and the corresponding SCNR will tend to the optimal value; the adaptive filtering weight vector is:

[0029]

[0030] The final output of the elevation array is:

[0031]

[0032] Where x is the original multi-channel echo data.

[0033] Furthermore, the number of samples K is greater than twice the number of channels.

[0034] A device described in the present invention includes a memory and a processor, wherein: the memory is used to store a computer program that can be run on the processor; the processor is used to execute the steps of the above-mentioned airborne meteorological radar space-time combined ground clutter suppression method when running the computer program.

[0035] A storage medium described in the present invention stores a computer program on the storage medium, and when the computer program is executed by at least one processor, the steps of the above-mentioned method for suppressing ground clutter by combining space and time for an airborne meteorological radar are implemented.

[0036] Beneficial effects: Compared with the prior art, the present invention has the following beneficial effects: According to the difference between the Doppler domain spectra of the meteorological echo and the ground clutter, the present invention preliminarily filters out the meteorological signal in the training sample and eliminates the influence of the meteorological signal on the covariance matrix, thereby calculating a more accurate weight vector; and then combining the spatial domain adaptive beamforming algorithm to suppress the ground clutter, greatly improving the antenna directivity pattern, the output signal-to-noise ratio improvement factor, and the convergence speed; compared with the traditional spatial domain adaptive beamforming algorithm, the present invention requires fewer samples and has higher robustness, and under the premise of effectively suppressing the ground clutter, the meteorological signal component is almost not lost, and the ground clutter suppression effect is better. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a flow chart of the space-time combined ground clutter suppression method for airborne weather radar proposed by the present invention;

[0038] Figure 2 The geometric model of the airborne weather radar and the target space schematic diagram proposed by the present invention;

[0039] Figure 3 The antenna pattern proposed by the present invention;

[0040] Figure 4 It is the SCNR improvement factor curve;

[0041] Figure 5 This is a local enlarged view of the SCNR improvement factor curve;

[0042] Figure 6 This is the curve of output SCNR changing with the number of samples;

[0043] Figure 7 It is the curve diagram of output SCNR changing with input SNR;

[0044] Figure 8 It is the true value diagram of simulated meteorological data;

[0045] Fig. 9 To simulate the original data of radar echo;

[0046] Fig.10 This is the result diagram of clutter suppression of traditional spatial adaptive filtering algorithm simulation data;

[0047] Fig.11 This is a diagram showing the clutter suppression results of the simulation data of the present invention;

[0048] Fig.12 This is a cross-sectional comparison diagram of the processing results of the present invention and the traditional method;

[0049] Fig.13 The actual measurement and channel data for a helicopter;

[0050] Fig.14 This is the measured difference channel data of a helicopter;

[0051] Fig.15 This is the clutter suppression result diagram of the measured data of the traditional spatial adaptive filtering algorithm;

[0052] Fig.16 This is a graph of clutter suppression results of measured data of the present invention. DETAILED DESCRIPTION

[0053] The present invention is further described in detail below with reference to the accompanying drawings.

[0054] When estimating the covariance matrix, the traditional spatial adaptive beamforming algorithm does not consider the expected meteorological signal in the sample, which will affect the accuracy of the covariance matrix estimation. This leads to a decrease in the ground clutter suppression performance of the traditional spatial algorithm in the unit where the meteorological target is located, and some meteorological information will be lost. To solve the above problems, Figure 1 As shown, the present invention proposes a space-time combined ground clutter suppression method for airborne meteorological radar, and the specific implementation process is as follows:

[0055] Step 1: Construct an airborne weather radar elevation multi-channel array signal receiving model, and obtain the range-pulse-elevation array three-dimensional echo data after range pulse compression processing.

[0056] Figure 2 The geometric model of the airborne pitch array weather radar is shown, as well as the spatial position of the weather targets and the ground on the equidistant ring relative to the radar. The aircraft is moving at a speed of v a Flying at a uniform speed, the array radar is placed parallel to the Z axis. The array is a uniform linear array with one transmit and multiple receive. The array elements meet the half-wavelength spacing (i.e. d = λ / 2). The meteorological radar usually works in the downward scanning mode to scan the area in front of the airborne radar. A linear frequency modulated pulse signal is emitted at every pulse repetition frequency. The reflected echo is received by the array antenna, and the echo is sampled to obtain multi-channel sampling data to complete the scan of the entire area. After the distance pulse is compressed, the three-dimensional echo data of the range-pulse-pitch array is obtained. In actual processing, in order to meet the real-time requirements, the data of each channel is usually reduced in dimension to synthesize the sum and difference beam data, that is, the sum / difference-range-pulse data is finally obtained.

[0057] Step 2: Calculate the Doppler velocity compensation value of each range-pulse pixel point based on the aircraft flight speed, beam scanning azimuth, altitude, and distance parameters to compensate for the Doppler shift caused by aircraft motion.

[0058] Due to the movement of the aircraft, each clutter unit produces a Doppler shift. The pitch angle and azimuth angle of clutter units in different places relative to the radar are different, and the Doppler shifts produced are also different. Therefore, it is necessary to calculate the corresponding Doppler shift for each unit. The Doppler shift of the (i, j)th range-pulse clutter unit is:

[0059]

[0060] Among them, θ i , are the elevation and azimuth angles of the (i, j)th range-pulse pixel, θ i The calculation method is:

[0061] θ i =arcsin(H / R i,j )

[0062] Among them, H is the flight altitude of the aircraft, R i,j is the distance of the (i,j)th range-pulse ground clutter unit relative to the radar. The elevation angles of the N range gates θ=[θ1,θ2,…,θ N ] T , M pulse azimuths Then the Doppler shift of the entire scanning area is:

[0063]

[0064] in, represents the Kronecker product. Then the Doppler compensation result of the (i, j)th range-pulse ground clutter unit data of a certain channel is for:

[0065]

[0066] Among them, x i,j is the data of the (i, j)th range-pulse unit of a channel before Doppler compensation, f r is the pulse repetition frequency, and t is the coherent processing time.

[0067] Step 3: Design a low-pass filter to perform Doppler domain filtering on the range-pulse data of each channel and group filtering on the pulse dimension to filter out meteorological signals and eliminate the influence of meteorological signals on the interference plus noise covariance matrix estimation.

[0068] After compensating for the Doppler shift caused by aircraft motion, ground clutter is usually near zero frequency and has a narrow spectrum width, while meteorological echo is far away from zero frequency and has a wide spectrum width. Therefore, a low-pass filter can be designed (the present invention uses an elliptical filter), and the pulse dimension data of each channel is grouped (usually 32 or 64 pulses can be selected as a group, and the specific number of pulses is selected according to actual conditions. For the interval scanning working mode, the number of pulses resident at each wave position is used as a group) for filtering, and the meteorological component is filtered out to obtain ground clutter plus noise data samples.

[0069] Step 4: Use the data after filtering out the meteorological signal in step 3 as samples, estimate the interference plus noise covariance matrix, which is used to calculate the weight vector required by the adaptive beamforming algorithm, return to the spatial domain for adaptive filtering processing, and use the estimated weight vector and the original data in step 1 to perform adaptive beamforming to suppress ground clutter.

[0070] Use the data after filtering out the meteorological data in step 3 as samples to estimate the interference plus noise covariance matrix but

[0071]

[0072] in, To filter out the multi-channel signal after the meteorological signal, K represents the number of samples taken, which is usually greater than twice the number of channels (i.e., K>2P). As the number of samples K increases, the sample covariance matrix Converges to the theoretical covariance matrix R c+n , and the corresponding SCNR will also tend to the optimal value. Then the adaptive filtering weight vector of this method is:

[0073]

[0074] The final output of the elevation array is:

[0075]

[0076] Where x is the original multi-channel echo data.

[0077] The present invention also provides a device, including a memory and a processor, wherein: the memory is used to store a computer program that can be run on the processor; the processor is used to execute the steps of the above-mentioned airborne weather radar space-time combined ground clutter suppression method when running the computer program.

[0078] The present invention also provides a storage medium having a computer program stored thereon, and when the computer program is executed by at least one processor, the steps of the above-mentioned method for suppressing ground clutter by combining space and time for an airborne meteorological radar are implemented.

[0079] To prove the feasibility of the present invention, a simulation experiment was first carried out, in which the pitch angle of the meteorological target was set to 7°~9°, the pitch angle of the ground clutter was set to 22°, the Doppler center frequency and spectral width of the meteorological target were set to 150Hz and 100Hz respectively, the Doppler center frequency and spectral width of the ground clutter were set to 0Hz and 10Hz respectively, and the number of array elements P=16. Figure 3 A comparison diagram of the antenna radiation patterns of the method proposed in the present invention and several traditional algorithms is given. It can be seen that the traditional method has higher sidelobe power and a shallow clutter notch, while the method of the present invention has lower sidelobe power and a deeper clutter notch, which is closer to the theoretical optimal algorithm. Figure 4 The signal-to-noise ratio improvement factor curve is given. Figure 5 A local enlarged image is given, and by comparison it can be seen that the main lobe notch of the method of the present invention is narrower, while the side lobe improvement effect is better. Figure 6 A curve showing the variation of the output signal-to-noise ratio with the number of samples is given. The present invention can still maintain good performance when the number of samples is small. Figure 7 The curve of output signal-to-noise ratio changing with signal-to-noise ratio is given. As the signal-to-noise ratio increases, the performance of other methods gradually decreases, while the performance of the present invention can still approach the theoretical optimal value.

[0080] In order to further verify the effectiveness of the present invention, the sum / difference beam simulation data is first generated, and the simulation parameters are as follows: radar height H = 10 km, carrier speed v a =150m / s, pulse repetition frequency PRF=1000Hz, beam scanning angle range is -20°~20°, meteorological target Doppler center frequency and spectrum width are 150Hz and 100Hz respectively, meteorological target Doppler center frequency and spectrum width are 0Hz and 20Hz. Figure 8 The true value diagram of the simulated meteorological target is given. Fig. 9 The simulated radar echo raw and channel data are given, and it can be seen that the meteorological target is submerged in the ground clutter background. Fig.10 The clutter suppression results of the traditional spatial adaptive filtering algorithm are given. Fig.11 The clutter suppression results of the method of the present invention are given. By comparison, it can be seen that the method of the present invention has less loss of meteorological targets and more obvious suppression effect of surrounding ground clutter. In order to reflect the performance improvement of the method of the present invention in ground clutter suppression, the signal with a range gate in the range of 650-680 (the area where the meteorological target is located) is selected, the echo intensity is calculated, and the echo intensity-pulse profile is obtained after the range gate is averaged. Fig.12 The corresponding comparison results are given. It can be seen from the profile that at the location of the meteorological target (the number of pulses is in the 200-550 region), compared with the ground clutter suppression results of the traditional spatial adaptive filtering algorithm, the method of the present invention has almost no meteorological signal loss and is closer to the real meteorological signal. This shows that the method of the present invention can retain the meteorological signal as much as possible while suppressing the ground clutter, and the performance is improved.

[0081] The method of the present invention is continued to be used to suppress ground clutter on the measured data of a helicopter weather radar. Fig.13 , Fig.14 The sum and difference data before ground clutter suppression are given respectively. The meteorological targets are completely submerged in the high-intensity ground clutter background, which poses a great challenge to the subsequent meteorological target detection. Fig.15 , Fig.16 The ground clutter suppression results of the measured data of the traditional spatial adaptive filtering method and the method of the present invention are given respectively. By comparison, it can be seen that the ground clutter suppression effect of the traditional method is not ideal, and the meteorological target is still submerged in the ground clutter background, while the clutter suppression effect of the method of the present invention is more obvious, the meteorological target is highlighted, and it is more convenient for subsequent meteorological target detection and processing.

[0082] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A space-time combined ground clutter suppression method for airborne meteorological radar, characterized in that: The following steps are involved: (1) Construct an airborne meteorological radar elevation multi-channel array signal receiving model. Each channel data is first processed by range pulse compression to obtain the three-dimensional echo data of the range-pulse-elevation array. (2) Calculate the Doppler velocity offset of each range-pulse unit Compensate for the phase difference caused by the corresponding multi-spectrum offset during the coherent processing time; (3) Design a low-pass filter, adjust the corresponding parameters, perform Doppler domain filtering on the range-pulse data of each channel, and perform group filtering on the pulse dimension to filter out the meteorological signal and eliminate the influence of the meteorological signal on the interference plus noise covariance matrix estimation; (4) Using the data after filtering out the meteorological signal as samples, the interference plus noise covariance matrix is ​​estimated to calculate the weight vector required by the adaptive beamforming algorithm, and the weight vector is returned to the spatial domain for adaptive filtering. The estimated weight vector and the original data of step (1) are used to perform adaptive beamforming to suppress ground clutter.

2. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 1, characterized in that: The implementation process of step (1) is as follows: A uniform linear array with one transmit and multiple receive is used to scan the area in front of the airborne radar. A linear frequency modulated pulse signal is emitted at a pulse repetition frequency. The reflected echo is received by the array antenna, and the echo is sampled to obtain multi-channel sampling data to complete the scan of the entire area. After distance pulse compression, the three-dimensional echo data of the range-pulse-elevation array is obtained.

3. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 1, characterized in that: The three-dimensional echo data of the range-pulse-elevation array is subjected to dimensionality reduction processing, and sum and difference beam data are synthesized to obtain sum or difference-range-pulse data.

4. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 1, characterized in that: The implementation process of step (2) is as follows: The corresponding pitch angle is calculated according to the aircraft altitude and the distance of each clutter block. The Doppler shift corresponding to each range-pulse unit of each channel is calculated using the aircraft movement speed, radar wavelength, azimuth and pitch angle of each clutter block. The Doppler shift of the (i, j)th range-pulse unit is: Among them, v a is the aircraft flying speed, θ i , are the elevation angle and azimuth angle of the (i, j)th range-pulse unit, θ i The calculation method is: θ i =arcsin(H / R i，j ) Among them, H is the flight altitude of the aircraft, R i，j is the distance of the (i, j)th range-pulse ground clutter unit relative to the radar; the elevation angles of the N range gates θ = [θ1, θ2, …, θ N ] T , M pulse azimuths Then the Doppler shift of the entire scanning area is: in, represents the Kronecker product; then the Doppler compensation result of the (i, j)th range-pulse ground clutter unit data of a certain channel is for: Among them, x i，j is the data of the (i, j)th range-pulse unit of a channel before Doppler compensation, f r is the pulse repetition frequency, and t is the coherent processing time.

5. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 1, characterized in that: The low-pass filter in step (3) is an elliptical low-pass filter.

6. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 1, characterized in that: The grouping and filtering process for the pulse dimension in step (3) is to select 32 or 64 pulses as a group; for the interval scanning working mode, the number of pulses resident at each wave position is used as a group.

7. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 1, characterized in that: The implementation process of step (4) is as follows: Use the data after filtering out the meteorological data as samples to estimate the interference plus noise covariance matrix but: in, To filter out the multi-channel data after the meteorological signal, K represents the number of samples taken. As the number of samples K increases, the sample covariance matrix Converges to the theoretical covariance matrix R c+n , and the corresponding SCNR will tend to the optimal value; the adaptive filtering weight vector is: The final output of the elevation array is: Where x is the original multi-channel echo data.

8. The method for suppressing ground clutter by combining space and time for airborne weather radar according to claim 7, characterized in that: The sample number K is greater than twice the number of channels.

9. A device, characterized in that: comprising a memory and a processor, wherein: A memory for storing computer programs that can be run on the processor; A processor is used to execute the steps of the space-time combined ground clutter suppression method for airborne meteorological radar as described in any one of claims 1 to 8 when running the computer program.

10. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by at least one processor, implements the steps of the space-time combined ground clutter suppression method for airborne meteorological radar as described in any one of claims 1 to 8.

Images