Radar detection and multi-dimensional parameter estimation method and system for high-maneuverability group target

Abstract

The invention discloses a radar detection and multi-dimensional parameter estimation method and system for a high-maneuverability group target. The method comprises the following steps: constructing a radar echo signal model of the high-maneuverability group target; utilizing a pulse compression technology to obtain a pulse pressure echo signal of the high-maneuverability group target; signal energy coherent accumulation is carried out through second-order phase compensation, slow time dimension non-uniform fast Fourier transform and distance dimension inverse fast Fourier transform, and group target detection and distance-speed-acceleration combined coarse estimation are achieved; constructing an over-complete dictionary according to a multi-dimensional parameter coarse estimation result, and performing sparse representation on the pulse pressure signal of the high-maneuverability group target; and performing sparse reconstruction on each target signal in the cluster in a distance-speed-acceleration estimation space by adopting a sparse Bayesian learning method to obtain a multi-dimensional parameter refined estimation result of each target. According to the method, the high-maneuverability group target can be rapidly detected without multi-dimensional search, coarse-to-fine estimation of multi-dimensional parameters of each target is realized, and both estimation precision and operation efficiency are considered.

Description

TECHNICAL FIELD

[0001] The present application relates to radar signal processing technology, in particular to a radar detection and multi-dimensional parameter estimation method and system for high-maneuvering group targets. BACKGROUND

[0002] With the rapid development of aerospace technology, high-maneuvering aircrafts such as unmanned aerial vehicles are increasingly present in the field of radar detection. Such targets have small radar cross section (RCS), are flexible and maneuverable, and can efficiently and accurately complete various specific tasks through high-density and large-scale deployment. Long-term observation is beneficial to the detection of weak targets by radar, but the high-maneuvering motion of the targets can cause range migration and Doppler migration of the echoes, causing the signal energy to spread along the range dimension and the velocity dimension, respectively, and reducing the performance of traditional moving target detection (MTD) methods. In addition, group targets that form dense formations often have very similar motion models, and the energy spectrum peaks of different targets can easily fall within the same range-velocity-acceleration resolution cell, making it difficult to detect and estimate the parameters of each target separately.

[0003] For the detection and parameter estimation of high-maneuvering targets, although some scholars have proposed algorithms such as generalized Radon-Fourier transform (GRFT) and polynomial Radon-polynomial Fourier transform, which can correct range-Doppler migration and coherently accumulate echo energy through joint search of target range, velocity, and acceleration, the computational complexity of multi-dimensional parameter traversal search is huge, and these methods are subject to resolution and do not have the ability to finely detect and estimate the parameters of group targets. SUMMARY

[0004] The present application relates to radar signal processing technology, in particular to a radar detection and multi-dimensional parameter estimation method and system for high-maneuvering group targets.

[0005] The technical solution for achieving the present application is as follows: In a first aspect, the present application provides a radar detection and multi-dimensional parameter estimation method for high-maneuvering group targets, comprising the following steps:

[0006] Constructing a radar echo signal model for high-maneuvering group targets;

[0007] Obtaining a pulse compression echo signal for high-maneuvering group targets using pulse compression technology;

[0008] Performing signal energy coherent accumulation through second-order phase compensation, slow-time NUFFT, and range IFFT to achieve group target detection and range-velocity-acceleration joint coarse estimation;

[0009] According to the multi-dimensional parameter rough estimation result, an over-complete dictionary is constructed, and the pulse compression signal of the high-maneuvering group target is sparsely represented.

[0010] A sparse Bayesian learning method is used to sparsely reconstruct the distance-speed-acceleration estimation space of each target signal in the cluster, so as to obtain the fine estimation result of the multi-dimensional parameter of each target.

[0011] In the second aspect, the present application provides a radar detection and multi-dimensional parameter estimation system for a high-maneuvering group target, which is used to realize the method in the first aspect, and the system comprises:

[0012] A first module is used to construct a radar echo signal model of the high-maneuvering group target.

[0013] A second module is used to obtain the pulse compression echo signal of the high-maneuvering group target by using the pulse compression technology.

[0014] A third module is used to realize the group target detection and the joint rough estimation of the distance-speed-acceleration by performing signal energy coherent accumulation through the second-order phase compensation, the slow-time NUFFT and the distance IFFT.

[0015] A fourth module is used to construct an over-complete dictionary according to the multi-dimensional parameter rough estimation result, and to sparsely represent the pulse compression signal of the high-maneuvering group target.

[0016] A fifth module is used to sparsely reconstruct the distance-speed-acceleration estimation space of each target signal in the cluster by using the sparse Bayesian learning method, so as to obtain the fine estimation result of the multi-dimensional parameter of each target.

[0017] In the third aspect, the present application provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the processor realizes the steps of the method in the first aspect when executing the program.

[0018] In the fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the program realizes the steps of the method in the first aspect when executed by a processor.

[0019] In the fifth aspect, the present application provides a computer program product, which comprises a computer program, and the computer program realizes the steps of the method in the first aspect when executed by a processor.

[0020] Compared with the prior art, the significant advantages of the present application are: for high maneuvering group targets under the joint influence of distance and Doppler migration, the present application combines second-order phase compensation, slow-time dimension NUFFT and distance dimension IFFT, and only needs to perform acceleration search to realize echo migration correction and coherent accumulation, so as to obtain target detection and parameter coarse estimation results, and avoids multi-dimensional parameter search. On this basis, a super-complete dictionary is constructed around the coarse estimation results and used to sparsely represent the group target signal, and then multi-dimensional sparse reconstruction is realized through sparse Bayesian learning to obtain refined estimation results of each target parameter, and the problem of insufficient resolution performance of the prior art on group targets is avoided. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 A flow chart of the radar detection and multi-dimensional parameter estimation method for high maneuvering group targets of the present application.

[0022] Figure 2 A schematic diagram of the motion trajectory of a high maneuvering group target after pulse compression.

[0023] Figure 3 A schematic diagram of the processing result of the MTD method.

[0024] Figure 4 A schematic diagram of the acceleration search result of the method of the present application.

[0025] Figure 5 A schematic diagram of the distance-velocity dimension coherent accumulation result of the method of the present application.

[0026] Figure 6 A schematic diagram of the three-dimensional sparse reconstruction result of the pulse compression signal of a high maneuvering group target.

[0027] Figure 7 A schematic diagram of the detection probability change curve of the method of the present application and the MTD method with input signal-to-noise ratio. DETAILED DESCRIPTION

[0028] The present application proposes a radar detection and multi-dimensional parameter estimation method for high maneuvering group targets, and the following will be combined with Figure 1 to specifically describe the steps of the method of the present application.

[0029] As Figure 1 shown, a radar detection and multi-dimensional parameter estimation method for high maneuvering group targets comprises the following steps:

[0030] Step 1, let denote the number of targets, then the radar echo signal model of a high maneuvering group target is constructed as .

[0031] ;

[0032] wherein, and These represent the fast-time variable and the slow-time variable, respectively. , , , , and These represent the pulse number, pulse repetition period, pulse width, speed of light, frequency modulation slope, and wavelength, respectively. For the goal The echo amplitude, Indicate target Relative radar instantaneous slant range, , and The target The initial slant range, radial velocity, and acceleration relative to the radar.

[0033] Step 2: High-mobility group target pulse compression echo signal obtained after pulse compression Represented as

[0034] ;

[0035] in, For the goal The amplitude of the pulse pressure signal, , and These are the distance-frequency variable, signal bandwidth, and carrier frequency, respectively.

[0036] according to It can be seen that, due to the distance frequency variable and slow time variables There is coupling between the target and the velocity components. The target's velocity and acceleration components will cause range migration in the echo, resulting in signal energy spreading along the range dimension. Additionally, the target's acceleration components will cause Doppler migration in the echo, causing signal energy to spread along the velocity dimension. Range and Doppler migration will make coherent energy accumulation of echoes from highly maneuverable targets difficult, severely reducing the target detection performance of traditional MTD methods.

[0037] Step 3: Obtain the coherent accumulation result of signal energy through second-order phase compensation, slow-time dimension NUFFT, and distance dimension IFFT. Represented as

[0038] ;

[0039] in, For azimuth frequency variables, To search for values ​​based on acceleration The constructed second-order phase compensation function, .

[0040] When the acceleration search value matches the target acceleration true value, The second-order phase term corresponding to the medium acceleration will be accurately compensated, and at this time The maximum peak value will be output, and its expression is

[0041] ;

[0042] Wherein, is the output signal amplitude of the target .

[0043] For the coherent accumulation result, target detection and joint rough estimation of distance, velocity and acceleration can be realized based on constant false alarm rate (CFAR) processing and main peak position.

[0044] Step 4, constructing an over-complete dictionary around the rough estimation value of the target multi-dimensional parameter , wherein and are the number of pulses and the number of distance dimension sampling points, respectively, , and are the number of dictionary grids in the distance, velocity and acceleration dimensions, respectively, and then the pulse compression signal of the high-maneuvering group target is vectorized and sparsely represented as

[0045] ;

[0046] Wherein and represent the observation signal vector and the sparse vector to be reconstructed, respectively, is a noise vector, which obeys a Gaussian distribution with zero mean and variance .

[0047] Step 5, using the sparse Bayesian learning method, the super-resolution sparse reconstruction of the target signals in the distance-velocity-acceleration estimation space in the cluster is carried out, and the parameter iterative process can be represented as

[0048] ;

[0049] Wherein is the mean of the posterior distribution, is the noise standard deviation, is the covariance matrix of the posterior distribution, , is the variance vector of , is the covariance matrix of , and are the first and second target parameters before and after updating, respectively. One hyperparameter, for The diagonal elements, for The Each element.

[0050] After the iteration is complete, Inverse vectorization is performed to obtain sparse reconstruction results of high-maneuverability group target signals in the three-dimensional estimation space of range-velocity-acceleration. Based on the coordinates of each reconstruction point, accurate estimates of the range, velocity, and acceleration of each target within the group can be obtained.

[0051] The present invention will be further described below with reference to specific embodiments.

[0052] The radar transmission signal carrier frequency was set to 0.3 GHz, signal bandwidth to 6 MHz, pulse width to 25 μs, pulse repetition frequency to 2000 Hz, coherent accumulation time to 1 s, and echo signal-to-noise ratio to -20 dB. The range, velocity, and acceleration parameters of each target within the cluster are shown in Table 1.

[0053] Table 1 Target parameters within the cluster

[0054]

[0055] Figure 2 The trajectory of a highly maneuverable group of targets after pulse compression is presented, showing that the highly maneuverable movement of the targets leads to a significant distance migration phenomenon. Figure 3 The results of the MTD method show that, due to the combined effects of range and Doppler migration, the echo energy of the group targets diffuses in the range-velocity dimension and cannot be effectively accumulated. Figure 4 and Figure 5 The acceleration search results and corresponding range-velocity coherent accumulation results of the method of the present invention are given respectively. It can be seen that the influence of range and Doppler migration on echo energy accumulation has been eliminated. However, due to the limitation of parameter estimation resolution, only the parameters of each dimension of target 2 in the cluster can be estimated.

[0056] Based on the construction of an overcomplete dictionary around the coarse estimation results, Figure 6 The results of three-dimensional sparse reconstruction of the pulse compression signal of a highly maneuverable group target are presented, where the dictionary grid spacing for range, velocity, and acceleration is 5 m, 0.1 m / s, and 0.1 m / s, respectively. 2 It can be seen that the three targets within the cluster were effectively distinguished, and the distance, velocity, and acceleration of each target were accurately estimated.

[0057] Set the false alarm probability to 10. -6 , Figure 7The change curves of the detection probability of the method and the MTD method with input signal-to-noise ratio are given, wherein 500 Monte Carlo experiments are performed for each input signal-to-noise ratio. As can be seen from the figure, the method has better target detection performance than the MTD method due to the joint compensation capability of the distance and the Doppler migration. When the detection probability is 0.9, the method has a signal-to-noise ratio gain of about 19 dB compared with the MTD method.

Claims

1. A radar detection and multi-dimensional parameter estimation method for highly maneuverable group targets, characterized in that, Includes the following steps: Construct a radar echo signal model for highly maneuverable group targets; Pulse compression technology was used to obtain pulse-compressed echo signals of highly maneuverable group targets; By performing coherent accumulation of signal energy through second-order phase compensation, slow-time dimension NUFFT and range dimension IFFT, group target detection and joint coarse estimation of range-velocity-acceleration are realized. An overcomplete dictionary is constructed based on the coarse estimation results of multidimensional parameters to sparsely represent the pulse compression signal of a highly maneuverable group of targets. A sparse Bayesian learning method is used to sparsely reconstruct the signals of each target in the cluster in the distance-velocity-acceleration estimation space, and obtain refined estimation results of multidimensional parameters of each target.

2. The radar detection and multi-dimensional parameter estimation method for highly maneuverable group targets according to claim 1, characterized in that, Constructed High-Mobility Group Target Echo Model Represented as ; in, and These represent the fast-time variable and the slow-time variable, respectively. , , , , , and These represent the pulse number, pulse repetition period, pulse width, speed of light, signal frequency modulation slope, signal wavelength, and number of targets, respectively. For the goal The echo amplitude, Indicate target Relative radar instantaneous slant range, , and The target The initial slant range, radial velocity, and acceleration relative to the radar.

3. The radar detection and multi-dimensional parameter estimation method for highly maneuverable group targets according to claim 2, characterized in that, High-mobility group target pulse compression echo signal obtained after pulse compression Represented as ; in, For the goal The amplitude of the pulse pressure signal, , and These are the distance-frequency variable, signal bandwidth, and carrier frequency, respectively.

4. The radar detection and multi-dimensional parameter estimation method for highly maneuverable group targets according to claim 3, characterized in that, Signal energy coherent accumulation results Represented as ; in, For azimuth frequency variables, It is a second-order phase compensation function. For acceleration search values, ; When the acceleration search value matches the target acceleration value The accumulated peak value is the largest. Based on constant false alarm rate processing and the position of the main peak, group target detection and joint coarse estimation of distance-velocity-acceleration are realized.

5. The radar detection and multi-dimensional parameter estimation method for highly maneuverable group targets according to claim 4, characterized in that, The sparse representation of the vectorized high-maneuverability group target pulse compression signal is as follows: ; in and Represent the observed signal vector and the sparse vector to be reconstructed, respectively. The noise vector has zero mean and zero variance. Gaussian distribution; An overcomplete dictionary constructed around the coarse estimates of the target parameters.

6. The radar detection and multi-dimensional parameter estimation method for highly maneuverable group targets according to claim 5, characterized in that, The sparse Bayesian learning iterative process of target signals in a highly maneuverable group is represented as follows: ; in for The mean of the posterior distribution. The standard deviation of noise. for The covariance matrix of the posterior distribution. , for The variance vector, and The first and second parts are respectively the first and second parts before and after the update. One hyperparameter, for The diagonal elements, for The One element; and These are the number of pulses and the number of sampling points in the distance dimension, respectively. After the iteration is complete, By performing inverse vectorization, sparse reconstruction results of high-maneuverability group target signals in the three-dimensional estimation space of range-velocity-acceleration are obtained, thereby achieving refined estimation of multi-dimensional parameters of each target.

7. A radar detection and multi-dimensional parameter estimation system for highly maneuverable group targets, characterized in that, The system for implementing the method according to any one of claims 1 to 6 comprises: The first module is used to construct a radar echo signal model of highly maneuverable group targets; The second module uses pulse compression technology to obtain pulse compression echo signals of highly maneuverable group targets; The third module performs coherent accumulation of signal energy through second-order phase compensation, slow-time dimension NUFFT and range dimension IFFT, realizing group target detection and joint coarse estimation of range-velocity-acceleration; The fourth module constructs an overcomplete dictionary based on the coarse estimation results of multidimensional parameters to sparsely represent the pulse compression signal of highly maneuverable groups of targets. The fifth module employs a sparse Bayesian learning method to sparsely reconstruct the signals of each target within the cluster in the distance-velocity-acceleration estimation space, thereby obtaining refined estimation results of the multidimensional parameters of each target.

8. 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 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-6.

Images