Radon-lvd transform-based robust detection method and system for high-speed and high-maneuvering targets

Abstract

This invention discloses a robust detection method and system for high-speed, highly maneuverable targets based on Radon-LVD transform, belonging to the field of radar signal processing. The method estimates the target's velocity and acceleration from the pulse-compressed echo signal using Radon-LVD transform; constructs a slow-time-range frequency domain quadratic phase compensation function using the velocity and acceleration estimates; performs migration compensation on the slow-time-range frequency domain of the echo signal; and then performs moving target detection in the range-Doppler domain. The two searches during this process improve the accuracy of target parameter estimation. This invention combines Radon transform and LVD transform, which can compensate for range migration and Doppler migration respectively, achieving coherent accumulation of energy in the echo signal of a uniformly accelerated target.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a robust detection method and system for high-speed, highly maneuverable targets based on Radon-LVD transform. Background Technology

[0002] With the continuous advancement of aerospace technology and the development and utilization of near-space resources, the focus of radar detection is gradually shifting to near-space hypersonic maneuvering targets and various high-speed, highly maneuverable targets. These targets generally possess typical characteristics such as hypersonic speed, strong maneuverability, and small radar cross-section (RCS). To achieve effective detection, it is necessary to accumulate echo data energy through long-term accumulation techniques that "trade time for energy," thereby improving the output signal-to-noise ratio (SNR) and the estimation accuracy of target parameters.

[0003] However, compared to conventional slow-moving targets, radar echoes from high-speed, highly maneuverable targets are highly susceptible to range migration and Doppler migration during the observation period. Range migration is triggered by the target's high speed, while Doppler migration exhibits complex characteristics under strong maneuverability. Both cause signal energy to spread along the range and velocity dimensions, respectively, severely reducing the performance of traditional moving target detection (MTD) methods and increasing the difficulty of radar detection of such targets. Therefore, how to address the migration problem to achieve effective detection of high-speed, highly maneuverable, and weak targets by existing radars, and further improve the accumulation gain, has become an urgent problem to be solved in the field of radar detection. Summary of the Invention

[0004] The purpose of this invention is to provide a robust detection method and system for high-speed, high-maneuverability targets based on Radon-LVD transform.

[0005] The technical solution to achieve the objective of this invention is as follows: Firstly, this invention provides a robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform, comprising the following steps:

[0006] (1) Perform pulse compression processing on radar echoes of high-speed maneuvering targets;

[0007] (2) Set the range and velocity coarse search intervals based on the target prior information, and set the search step size based on the radar parameters;

[0008] (3) Extract the target's slow-time-dimensional migration trajectory according to the search parameters;

[0009] (4) Perform LVD transformation on the slow-time dynamic trajectory to achieve fast coherent accumulation and estimate the target velocity and acceleration:

[0010] (5) Obtain the maximum value of all LVD transformation results in the distance-velocity search space, obtain the target acceleration estimate based on the frequency modulation slope calculated by the LVD transformation, and obtain the target velocity coarse estimate based on the velocity search number corresponding to the maximum value;

[0011] (6) Set a fine search interval based on the coarse velocity estimate, with a step size of the maximum unambiguous velocity; construct a phase compensation function using the velocity search value and the acceleration estimate during each search;

[0012] (7) The pulse compression echo signal is transformed to the slow time-range frequency domain. After migration compensation using the phase compensation function, the range-dimensional inverse Fourier transform and the slow time-dimensional Fourier transform are performed in sequence to obtain the range-Doppler domain target echo signal after coherent accumulation.

[0013] (8) Based on the maximum value among all velocity fine search accumulated spectral peaks, the velocity estimate and range estimate of the target are obtained;

[0014] (9) Detect the echo signal in the range-Doppler domain using the square law, and calculate the constant false alarm rate detection threshold according to the calculation formula of the detection threshold. Determine whether the signal is greater than or equal to the detection threshold. If yes, determine that there is a target in the echo signal; otherwise, determine that there is no target in the echo signal.

[0015] Furthermore, the expression for establishing the radar high-speed, high-maneuverability target echo model in step (1) is as follows:

[0016]

[0017] in, For fast time variables, i.e., distance-time, and These represent the transmitted signal pulse width and frequency modulation slope, respectively. , and These represent the echo signal amplitude, speed of light, and signal wavelength, respectively. The target's initial slant range, radial velocity, and acceleration relative to the radar are respectively... , and Then the instantaneous slant range between the target and the radar can be expressed as: ,in For slow time variables, and These represent the number of pulses and the pulse repetition interval, respectively.

[0018] Expression of the echo signal after pulse compression:

[0019]

[0020] in, The amplitude of the pulse pressure signal. This refers to the signal bandwidth.

[0021] Furthermore, in step (2), based on the prior information of the target, a coarse search interval for the target parameters is set, with the distance search range being... For pulse radar, This is a radar blind spot. That is, the maximum unambiguous velocity, and the velocity search range is set based on prior information. Considering only the joint search of the initial slant range and radial velocity, to minimize the number of searches and improve computational efficiency, the distance and velocity search step sizes can be set as follows:

[0022]

[0023]

[0024] in, Indicates the coherent accumulation time. At the speed of light, For bandwidth.

[0025] when and When the search value equals the true value, a slow-time dimension echo signal can be extracted along the target's motion trajectory.

[0026]

[0027] make , , ,but It is represented as a linear frequency modulated (LFM) signal:

[0028]

[0029] in Indicates the center frequency of the signal. This indicates the frequency modulation slope of the signal.

[0030] Subsequently, a search-free LVD transformation is used to perform rapid coherent accumulation.

[0031] Furthermore, step (4) specifically involves:

[0032] (4a) Construct the parameterized symmetric autocorrelation matrix of the extracted slow-time dimension signal;

[0033] (4b) Use the set scaling factor to perform scaling transformation on the matrix to eliminate the coupling between time and time delay variables;

[0034] (4c) Perform a two-dimensional fast Fourier transform on the scaled matrix to obtain the center frequency-frequency modulation slope (CFCR) domain of the signal, and extract the in-plane peak value and the corresponding two-dimensional coordinates.

[0035] Furthermore, the specific steps of the fast coherent accumulation of LVD transformation in step (4) are as follows:

[0036] Step 1: Construct a parameterized symmetric autocorrelation function

[0037]

[0038] in For time delay variables, It is a fixed delay.

[0039] Step 2: Scale Transformation

[0040]

[0041] in is the scale factor.

[0042] The formula Substitute The expression yields:

[0043]

[0044] Step 3: Two-dimensional Fast Fourier Transform (FFT)

[0045]

[0046] Therefore, it can be seen that LVD will The signal energy is focused on the CFCR (center frequency-frequency modulation slope) plane, thereby achieving long-term coherent accumulation of the target echo signal.

[0047] Furthermore, in step (5), the maximum value of all LVD transformation results is obtained within the distance-velocity search space, thus yielding the coarse velocity estimate. For the corresponding velocity search number, the acceleration estimate is... , This is the frequency modulation slope calculated from the LVD transformation result at this time.

[0048] Furthermore, step (6) uses the rough speed estimate obtained in the previous step. Set a fine-grained search range Let the speed refinement search step size be the maximum unambiguous speed. During each search, the search speed value is used. and acceleration estimates Construct the migration compensation phase as follows:

[0049]

[0050] Transforming the pulse compression echo signal to the slow time-distance frequency domain yields:

[0051]

[0052] in Represents the distance-frequency variable. Indicates the signal amplitude.

[0053] use right After phase compensation, a range-dimensional inverse Fourier transform and a slow-time Fourier transform are performed sequentially to obtain the target's MTD result. When the velocity search value is closest to the target's true velocity, the MTD accumulation spectrum peak will reach its maximum peak value, which can be obtained as follows:

[0054]

[0055] in, This represents the azimuth frequency variable.

[0056] In actual noise environments, The result of the range-Doppler domain transformation of the target echo.

[0057] Furthermore, in step (9), constant false alarm rate (CFAR) detection is performed on the target, using cell-averaged CFAR (CA-CFAR) processing. Let... This represents the probability of a false alarm. Indicates the reference window Each sample value, The number of reference units is the same as the reference window length. Therefore, the threshold product factor... Background power estimated by the reference cell Detection threshold According to the likelihood ratio test of the Neyman-Pearson criterion, the following judgment can be made:

[0058]

[0059] in, Let be the sampled value of the unit to be detected. When the sampled value of the unit to be detected is greater than the detection threshold, it is assumed that... If the condition is met, the unit is deemed to have a target; otherwise, it is deemed to have no target.

[0060] In a second aspect, the present invention provides a robust detection system for high-speed, high-maneuverability targets based on Radon-LVD transform, for implementing the method described in the first aspect, comprising:

[0061] The first module is used to perform pulse compression processing on radar echoes from high-speed, highly maneuverable targets.

[0062] The second module uses Radon-LVD joint transformation to estimate the target acceleration and constructs a migration compensation phase.

[0063] The third module involves a refined search based on the coarse velocity estimation results, and coherent accumulation of MTD on the migration-corrected signal.

[0064] The fourth module is used to perform constant false alarm rate (CFAR) processing on the MTD results to achieve target detection.

[0065] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method described in the first aspect.

[0066] Compared with the prior art, the significant advantages of the present invention are:

[0067] (1) For radar moving targets with high speed and high maneuverability, a method based on Radon transform and Lv distribution (LVD) is proposed, which can compensate for range migration and Doppler migration respectively, and realize coherent accumulation of the echo signal energy of uniformly accelerated moving targets;

[0068] (2) Compared to RFT, the method of this invention combines Radon transform with LVD transform to compensate for range migration and Doppler migration caused by acceleration. Compared to conventional fractional Fourier transform (FRFT), LVD transform can achieve higher accumulation under the same computational cost, thus improving the detection performance of weak targets. It can also achieve good detection results for some targets with small RCS and weak reflectivity. Attached Figure Description

[0069] Figure 1 This is a flowchart illustrating the complete technical solution of the present invention.

[0070] Figure 2 This is a schematic diagram of the target's motion trajectory after pulse compression.

[0071] Figure 3 This diagram illustrates the signals extracted along the slow time dimension after Radon transform with different search numbers.

[0072] Figure 4 This is a schematic diagram of the center frequency-frequency modulation slope plane after LVD transformation.

[0073] Figure 5 This is a schematic diagram of the peak values ​​of the LVD transformation results within the distance-velocity search space.

[0074] Figure 6 This is a schematic diagram of the target's trajectory after migration phase compensation.

[0075] Figure 7 A schematic diagram showing the peak values ​​of MTD results corresponding to different search speeds when performing a speed-refined search.

[0076] Figure 8 This is a schematic diagram of the distance-Doppler plane corresponding to the peak values ​​of all MTD results. Detailed Implementation

[0077] This invention proposes a robust detection method for high-speed, highly maneuverable targets based on Radon-LVD transform, belonging to the field of radar signal processing. The method includes: pulse compression processing of the received radar echo signal; determining the search interval and step size for range and velocity, and determining the target migration trajectory based on the search parameters; further achieving long-term coherent accumulation of the echo signal extracted along the slow time dimension after Radon transform using LVD transform; storing the CFCR (center frequency-modulation slope) plane peak values ​​from each search, and obtaining target range and velocity estimates based on the maximum peak value in the range-velocity search space after the search is completed, and obtaining target acceleration estimates based on the corresponding LVD transform results; performing a small-range search around the coarse velocity estimate obtained after Radon-LVD processing using a smaller search step size, performing migration phase compensation and MTD processing during each search; and using the MTD result corresponding to the maximum accumulated peak value as the range-Doppler domain transform output of the target echo signal, performing constant false alarm rate (CFAR) processing on it, thus achieving robust detection of high-speed, highly maneuverable targets.

[0078] This method estimates the target's velocity and acceleration from the pulse-compressed echo signal using Radon-LVD transform; it then constructs a slow-time-range frequency domain quadratic phase compensation function using the velocity and acceleration estimates; migration compensation is applied to the slow-time-range frequency domain of the echo signal, before transferring it to the range-Doppler domain for moving target detection. This process involves two searches to improve the accuracy of target parameter estimation. The following section combines... Figure 1 The steps of the method of the present invention will be described in detail.

[0079] like Figure 1 As shown, a robust detection method for high-speed, highly maneuverable targets based on Radon-LVD transform includes the following steps:

[0080] (1) Assume that the initial slant range, radial velocity, and acceleration of the target relative to the radar are respectively , and Then the instantaneous slant range between the target and the radar can be expressed as: ,in For slow time variables, and These represent the pulse number and the pulse repetition interval, respectively. The echo model expression is then as follows:

[0081]

[0082] in, For fast time variables, i.e., distance-time, and These represent the transmitted signal pulse width and frequency modulation slope, respectively. , and These represent the echo signal amplitude, the speed of light, and the signal wavelength, respectively.

[0083] By performing pulse compression on the received signal, the expression for the pulse-compressed echo signal can be obtained:

[0084]

[0085] in, The amplitude of the pulse pressure signal. This refers to the signal bandwidth.

[0086] (2) Based on the prior information of the target, set the coarse search interval for the target parameters, with the distance search range being... For pulse radar, This is a radar blind spot. That is, the maximum unambiguous velocity, and the velocity search range is set based on prior information. Then, an appropriate search step size is set. Considering only the joint search of the initial slant range and radial velocity, to minimize the number of searches and improve computational efficiency, the distance and velocity search step sizes can be set as follows:

[0087]

[0088]

[0089] in, Indicates the coherent accumulation time. At the speed of light, For bandwidth.

[0090] when and When the search value equals the true value, a slow-time dimension echo signal can be extracted along the target's motion trajectory.

[0091]

[0092] make , , ,but It is represented as a linear frequency modulated (LFM) signal:

[0093]

[0094] in Indicates the center frequency of the signal. This indicates the frequency modulation slope of the signal.

[0095] Subsequently, a search-free LVD transformation is used to perform rapid coherent accumulation.

[0096] The specific steps of LVD transformation are as follows:

[0097] Step 1: Construct a parameterized symmetric autocorrelation function

[0098]

[0099] in For time delay variables, It is a fixed delay.

[0100] Step 2: Scale Transformation

[0101]

[0102] in is the scale factor.

[0103] The formula Substitute The expression yields:

[0104]

[0105] Step 3: Two-dimensional Fast Fourier Transform (FFT)

[0106]

[0107] Therefore, it can be seen that LVD will The signal energy is focused on the CFCR (center frequency-frequency modulation slope) plane, thereby achieving long-term coherent accumulation of the target echo signal.

[0108] (3) Obtain the maximum value of all LVD transformation results within the distance-velocity search space, then the coarse velocity estimate is obtained. For the corresponding velocity search number, the acceleration estimate is... , This is the frequency modulation slope calculated from the LVD transformation result at this time. Based on the rough speed estimate... Set a fine-grained search range Let the speed refinement search step size be the maximum unambiguous speed. During each search, the search speed value is used. and acceleration estimates Construct the migration compensation phase as follows:

[0109]

[0110] Transforming the pulse compression echo signal to the slow time-distance frequency domain yields:

[0111]

[0112] in Represents the distance-frequency variable. Indicates the signal amplitude.

[0113] use right After phase compensation, a range-dimensional inverse Fourier transform and a slow-time Fourier transform are performed sequentially to obtain the target's MTD result. When the velocity search value is closest to the target's true velocity, the MTD accumulation spectrum peak will reach its maximum peak value, which can be obtained as follows:

[0114]

[0115] in, This represents the azimuth frequency variable.

[0116] In actual noise environments, The result of the range-Doppler domain transformation of the target echo.

[0117] (4) Perform constant false alarm rate (CFAR) detection on the target, using cell-averaged CFAR (CA-CFAR) processing. Let... This represents the probability of a false alarm. Indicates the reference window Each sample value, The number of reference units is the same as the reference window length. Therefore, the threshold product factor... Background power estimated by the reference cell Detection threshold According to the likelihood ratio test of the Neyman-Pearson criterion, the following judgment can be made:

[0118]

[0119] in, Let be the sampled value of the unit to be detected. When the sampled value of the unit to be detected is greater than the detection threshold, it is assumed that... If the condition is met, the unit is deemed to have a target; otherwise, it is deemed to have no target.

[0120] The invention will be further illustrated below with simulation examples.

[0121] The radar transmission signal is set to a carrier frequency of 35GHz, a signal bandwidth of 10MHz, and a pulse width of 21. The pulse repetition frequency is 3333Hz, and the accumulation time is 9.6ms. The initial slant range, radial velocity, and acceleration of the target relative to the radar are 30km, 10km / s, and 200m / s², respectively. 2 .

[0122] Figure 2 The target trajectory after pulse compression is given, showing that the target's high-speed and high-maneuverability motion leads to severe migration. Figure 3 The target trajectories extracted along the slow time dimension of the pulse compression matrix for different search numbers after Radon transform are shown. It can be seen that when the search number is close to the true value, the trajectory takes the form of a linear frequency modulated signal. Figure 4 The CFCR plane after LVD transformation shows that the target signal is coherently accumulated, and the frequency modulation slope unit corresponding to the peak value can be used to calculate the target's acceleration estimate. Figure 5 This represents the peak LVD transformation result for all distance-velocity search parameters, with the maximum value taken as the target's LVD transformation result. Figure 6 This is the result after phase compensation, and it can be seen that the distance migration has been corrected. Figure 7 It is the peak value of the MTD results corresponding to all search counts when speed is refined for search. Figure 8 yes Figure 7 The MTD distance at the mid-peak value - Doppler plane.

[0123] The above embodiments are described in a relatively specific and detailed manner, but they should not be construed as limiting the scope of the invention patent.

Claims

1. A robust detection method for high-speed, highly maneuverable targets based on Radon-LVD transform, characterized in that, include: (1) Perform pulse compression processing on radar echoes of high-speed maneuvering targets; (2) Set the range and velocity coarse search intervals based on the target prior information, and set the search step size based on the radar parameters; (3) Extract the target's slow-time-dimensional migration trajectory according to the search parameters; (4) Perform LVD transformation on the slow-time dynamic trajectory to achieve fast coherent accumulation and estimate the target velocity and acceleration: (5) Obtain the maximum value of all LVD transformation results in the distance-velocity search space, obtain the target acceleration estimate based on the frequency modulation slope calculated by the LVD transformation, and obtain the target velocity coarse estimate based on the velocity search number corresponding to the maximum value; (6) Set a fine search interval based on the coarse velocity estimate, with a step size of the maximum unambiguous velocity; construct a phase compensation function using the velocity search value and the acceleration estimate during each search; (7) The pulse compression echo signal is transformed to the slow time-range frequency domain. After migration compensation using the phase compensation function, the range-dimensional inverse Fourier transform and the slow time-dimensional Fourier transform are performed in sequence to obtain the range-Doppler domain target echo signal after coherent accumulation. (8) Based on the maximum value among all velocity fine search accumulated spectral peaks, the velocity estimate and range estimate of the target are obtained; (9) Detect the echo signal in the range-Doppler domain using the square law, and calculate the constant false alarm rate detection threshold according to the calculation formula of the detection threshold. Determine whether the signal is greater than or equal to the detection threshold. If yes, determine that there is a target in the echo signal; otherwise, determine that there is no target in the echo signal.

2. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 1, characterized in that, In step (1), a radar high-speed, high-maneuverability target echo model is established. The expression is as follows: ； in, For fast time variables, i.e., distance-time, and These represent the transmitted signal pulse width and frequency modulation slope, respectively. , and Representing the echo signal amplitude, speed of light, and signal wavelength respectively; the target's initial slant range, radial velocity, and acceleration relative to the radar are respectively... , and The instantaneous slant range between the target and the radar is expressed as ,in For slow time variables, and These are the number of pulses and the pulse repetition interval, respectively. Echo signal after pulse compression expression: ； in, The amplitude of the pulse pressure signal. This refers to the signal bandwidth.

3. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 2, characterized in that, In step (2), based on the prior information of the target, a coarse search interval for the target parameters is set, with the distance search range being... For pulse radar, This is a radar blind spot. That is, the maximum unambiguous velocity, and the velocity search range is set according to prior information; considering only the joint search of initial slant range and radial velocity, the distance and velocity search step sizes are set as follows: ， ， in, Indicates the coherent accumulation time. At the speed of light, For bandwidth; when and When the search value equals the true value, the slow-time dimension echo signal is extracted along the target's motion trajectory: ； make , , ,but It is represented as a linear frequency modulated signal: ； in Indicates the center frequency of the signal. This indicates the frequency modulation slope of the signal.

4. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 1, characterized in that, Step (4) is as follows: (4a) Construct the parameterized symmetric autocorrelation matrix of the extracted slow-time dimension signal; (4b) Use the set scaling factor to perform scaling transformation on the matrix to eliminate the coupling between time and time delay variables; (4c) Perform a two-dimensional fast Fourier transform on the scaled matrix to obtain the center frequency-frequency modulation slope domain of the signal, and extract the peak value and corresponding two-dimensional coordinates in the plane.

5. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 1, characterized in that, Step (4) involves performing LVD transformation on the slow-time dynamic trajectory to achieve fast coherent accumulation. The specific steps are as follows: 1) Construct a parameterized symmetric autocorrelation function ； in For time delay variables, For fixed delay; 2) Scale transformation ； in Scale factor; The formula Substitute The expression yields: ； 3) Two-dimensional Fast Fourier Transform ； LVD will The signal energy is focused on the CFCR plane, thereby achieving long-term coherent accumulation of the target echo signal.

6. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 1, characterized in that, Step (5) Obtain the maximum value of all LVD transformation results within the distance-velocity search space, then the coarse velocity estimate is obtained. For the corresponding velocity search number, the acceleration estimate is... , This is the frequency modulation slope calculated from the LVD transformation result at this time.

7. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 6, characterized in that, Step (6): Based on the rough speed estimate obtained in the previous step... Set a fine-grained search range Let the speed refinement search step size be the maximum unambiguous speed. During each search, the speed search value is used. and acceleration estimates Construct the migration compensation phase as follows: ； Transforming the pulse compression echo signal to the slow time-distance frequency domain yields: ； in Represents the distance-frequency variable. Indicates signal amplitude; use right After phase compensation, the range-dimensional inverse Fourier transform and the slow-time Fourier transform are performed sequentially to obtain the target's MTD result. When the velocity search value is closest to the target's true velocity, the MTD accumulated spectral peak will reach its maximum peak value, which means: ； in, Represents the azimuth frequency variable; In actual noise environments, The result of the range-Doppler domain transformation of the target echo.

8. The robust detection method for high-speed, high-maneuverability targets based on Radon-LVD transform according to claim 1, characterized in that, In step (9), constant false alarm rate (CFAR) detection is performed on the target, using cell-averaged CFAR processing; assuming This represents the probability of a false alarm. Indicates the reference window Each sample value, , The threshold product factor is the number of reference units, i.e., the reference window length. Background power estimated by the reference cell Detection threshold According to the likelihood ratio test of the Neyman-Pearson criterion, the following judgment can be made: ； in, Let be the sampled value of the unit to be detected; when the sampled value of the unit to be detected is greater than the detection threshold, assume... If the condition is met, the unit is deemed to have a target; otherwise, it is deemed to have no target.

9. A robust detection system for high-speed, high-maneuverability targets based on Radon-LVD transform, used to implement the method described in any one of claims 1-8, the system comprising: The first module is used to perform pulse compression processing on radar echoes from high-speed, highly maneuverable targets. The second module uses Radon-LVD joint transformation to estimate the target acceleration and constructs a migration compensation phase. The third module involves a refined search based on the coarse velocity estimation results, and coherent accumulation of MTD on the migration-corrected signal. The fourth module is used to perform constant false alarm rate (CFAR) processing on the MTD results to achieve target detection.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1-7.

Images