Low-altitude detection radar inverse scattering method based on Born approximation

Abstract

The invention discloses a low-altitude detection radar inverse scattering method based on Born approximation, and the method comprises the following steps: 1, modeling: employing quantum-classical mixed scattering for modeling, introducing a Bose-Einstein condensed state coherent state operator # imgabs0 # to describe the quantum correlation effect between vegetation scatterers, adding dynamic dielectric coupling, and carrying out the calculation of the quantum correlation effect between the vegetation scatterers; carrying out real-time tracking on the vegetation water content, the surface roughness and the vegetation coupling effect; 2, filtering: performing three-dimensional filtering processing on the original radar echo data to obtain radar echo filtering data; step 3, self-adaptive full-waveform inversion is carried out; step 4, solving; according to the method, quantum-classical mixed scattering modeling is adopted, epsilon c (r, t) = epsilon o + kWv (t) * front h (t), the time-varying characteristic of the vegetation water content is considered, a dielectric constant disturbance function epsilon c (r, t) becomes a dynamic variable, the accuracy of vegetation inverse scattering detection is improved, the front h (t) considers the ground surface roughness and the vegetation coupling effect, and the detection accuracy of the vegetation inverse scattering detection is improved. And the precision of vegetation inverse scattering detection is improved.

Description

Technical Field

[0001] The present invention relates to the field of radar technology, and in particular to a low-altitude detection radar inverse scattering method based on Born approximation. Background Art

[0002] When simulating complex urban electromagnetic environments, there are huge problems with unknown quantities. The electromagnetic modeling and simulation platform based on the Born signal reconstruction method can solve this problem and quickly calculate the required parameters of the target.

[0003] Born signal reconstruction methods are typically based on the Born approximation, a linearized treatment of the wave equation in scattering problems. It approximates the effect of a scatterer on the wavefield as a perturbation, considering the scattered wavefield to be the linear superposition of the incident wavefield and the perturbation wavefield caused by the scatterer. Taking optical diffraction tomography as an example, this method measures information such as the intensity of scattered light and uses the Born approximation to establish the relationship between the scattered light and the refractive index distribution within the object, thereby reconstructing the object's internal structure. The basic idea is to group discrete sub-scatterers on the scatterer's surface and employ different mutual coupling calculation methods based on the positional relationships within the groups.

[0004] However, the traditional Born signal reconstruction method has the following problems in detecting vegetation areas: 1. Lack of coherent scattering. The traditional model assumes that vegetation scatterers are independent and uncorrelated. However, when low-frequency electromagnetic waves are incident, there are quantum-scale coherence effects and near-field interactions between scatterers, resulting in phase distortion; Multiple scattering cumulative errors. The first-order Born approximation only considers single scattering. Dense scatterers in vegetation areas trigger more than three scattering paths, and the cumulative phase error reaches Δφ>π / 2, forming a ghost (the measured ghost rate is >12%); 2. Dynamic dielectric properties are not modeled. The time-varying characteristics of vegetation water content (such as diurnal transpiration) make the dielectric constant perturbation function O(r) a dynamic variable. The traditional static model cannot capture the random fluctuations of ε(r,t); 3. Terrain-vegetation coupling effect. The surface roughness (correlation length lc) and the vegetation layer jointly modulate multipath reflections to produce coupled scattering fields. The existing algorithm is not decoupled.

[0005] In summary, a low-altitude detection radar inverse scattering method based on Born approximation is designed. Summary of the Invention

[0006] In order to overcome the above-mentioned shortcomings, the present invention provides a low-altitude detection radar inverse scattering method based on Born approximation.

[0007] The present invention achieves the above-mentioned purpose through the following technical solutions:

[0008] A low-altitude detection radar inverse scattering method based on Born approximation includes the following steps:

[0009] Step 1: Modeling, using quantum-classical mixed scattering modeling and introducing Bose-Einstein condensate coherent state operators Describe the quantum correlation effect between vegetation scatterers, add dynamic dielectric coupling, and track the vegetation moisture content and the coupling effect between surface roughness and vegetation in real time;

[0010] Step 2: Filtering: Perform three-dimensional filtering on the original radar echo data to obtain radar echo filtered data;

[0011] Step 3: Adaptive full waveform inversion, which segments and weights the data obtained in step 2 and then adapts it to the model in step 1;

[0012] Step 4: Solve the cross-correlation objective function in step 3 through Ising model mapping and quantum tunneling effect, and output the target scattering point cloud;

[0013] Step 5: Multipath decoupling imaging uses an imaging operator to separate the direct wave and multipath reflected waves through mathematical modeling, eliminating vegetation shadows;

[0014] Among them, step three includes the following steps:

[0015] S31, data segmentation and weighting, segmenting the original radar echo data and the radar echo filter data into N sub-regions, and calculating the cross-correlation objective function;

[0016] S32, adaptive regularization, dynamically adjusts the Tikhonov regularization parameters based on the error distribution graph to balance the data fitting and model complexity of S31 and avoid overfitting;

[0017] S33, multi-scale decomposition inversion, decomposes the data into low-frequency, medium-frequency, and high-frequency components, and independently inverts the scattering model of each scale. The low-frequency components constrain the large-scale structure, and the high-frequency components optimize the details and improve the convergent scattering angle;

[0018] S34, quantum scattering enhancement, injects quantum correction terms into the gradient to enhance the heat dissipation response of vegetation.

[0019] Preferably, the calculation formula for the quantum-classical mixed scattering modeling is:

[0020]

[0021] in, is the quantum correction term, describing the quantum correlation effect between vegetation scatterers, β=(k B T) -1 ;

[0022] εc (r, t) = ε o +kW v (t)·f rough (t), W v (t) is the time-varying vegetation water content, is the surface roughness attenuation function.

[0023] Preferably, in step 2, wavelet packet entropy filtering is used for time domain filtering of radar echoes, tensor low-rank decomposition is used for spatial domain filtering of radar echoes, and quantum derivative filter is used for frequency domain filtering of radar echoes.

[0024] Preferably, the quantum derivative filter is a frequency domain filter based on the Schrödinger equation;

[0025] The calculation formula is as follows:

[0026]

[0027] Used to simultaneously suppress ground clutter (low frequency) and radio frequency interference (high frequency), improving the signal-to-noise ratio by ≥15dB.

[0028] Preferably, the calculation formula of the cross-correlation objective function is as follows:

[0029]

[0030] in, is the original radar echo data, is the radar echo filtering data, ω k is the weight coefficient, which is used to focus on the strong signal area and suppress local noise interference.

[0031] Preferably, the regularization parameter is λ, and the calculation formula of λ is as follows:

[0032]

[0033] Among them, δ is the attenuation factor, and the empirical value is 0.01-0.1.

[0034] Preferably, the specific steps of step 4 are as follows:

[0035] S41, dynamic Hamiltonian evolution, through adiabatic evolution, the system transitions from the quantum superposition state to the classical ground state, obtaining the global minimum energy solution;

[0036] S42. Conversion of spin state to dielectric constant: converting the measurement results of quantum bits into dielectric constant distribution in physical space. The calculation formula is as follows: α is the linear mapping coefficient, which is determined by calibration experiments (typical value α≈0.1~0.5), ζ is the quantum fluctuation correction term coefficient, which is used to compensate for annealing noise. is the expected value of the spin in the z direction, obtained through multiple sampling statistics;

[0037] S43, iterative feedback mechanism, combines the fast initial solution of classical algorithms with the global optimization capability of quantum annealing, balancing efficiency and accuracy;

[0038] S44, quantum noise suppression, suppresses the inherent noise of quantum hardware through Boltzmann weighted averaging and improves inversion stability.

[0039] The beneficial effects of the present invention are as follows: in the low-altitude detection radar inverse scattering method based on Born approximation:

[0040] 1. Using quantum-classical hybrid scattering modeling, ε c (r, t) = ε o +kW v (t)·f rough (t), taking into account the time-varying characteristics of vegetation moisture content, the dielectric constant perturbation function ε c (r, t) becomes a dynamic variable, which improves the accuracy of vegetation backscatter detection. Similarly, f rough (t) Taking into account the coupling effect of surface roughness and vegetation, the accuracy of vegetation inverse scattering detection is further improved;

[0041] 2. In step three, quantum scattering enhancement is used to inject a quantum correction term into the gradient to enhance the heat dissipation response of vegetation. This solves the problem of phase distortion caused by quantum-scale coherence effects and near-field interactions between scatterers when low-frequency electromagnetic waves are incident. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0043] Figure 1 It is a diagram of the method steps of the present invention. DETAILED DESCRIPTION

[0044] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0045] like Figure 1 As shown, a low-altitude detection radar inverse scattering method based on Born approximation includes the following steps:

[0046] Step 1: Modeling, using quantum-classical mixed scattering modeling and introducing Bose-Einstein condensate coherent state operators Describe the quantum correlation effect between vegetation scatterers, add dynamic dielectric coupling, and track the vegetation moisture content and the coupling effect between surface roughness and vegetation in real time;

[0047] Step 2: Filtering: Perform three-dimensional filtering on the original radar echo data to obtain radar echo filtered data;

[0048] Step 3: Adaptive full waveform inversion, which segments and weights the data obtained in step 2 and then adapts it to the model in step 1;

[0049] Step 4: Solve the cross-correlation objective function in step 3 through Ising model mapping and quantum tunneling effect, and output the target scattering point cloud;

[0050] Step 5: Multipath decoupling imaging uses an imaging operator to separate the direct wave and multipath reflected waves through mathematical modeling, eliminating vegetation shadows;

[0051] Among them, step three includes the following steps:

[0052] S31, data segmentation and weighting, segmenting the original radar echo data and the radar echo filter data into N sub-regions, and calculating the cross-correlation objective function;

[0053] S32, adaptive regularization, dynamically adjusts the Tikhonov regularization parameters based on the error distribution graph to balance the data fitting and model complexity of S31 and avoid overfitting;

[0054] S33, multi-scale decomposition inversion, decomposes the data into low-frequency, medium-frequency, and high-frequency components, and independently inverts the scattering model of each scale. The low-frequency components constrain the large-scale structure, and the high-frequency components optimize the details and improve the convergent scattering angle;

[0055] S34, quantum scattering enhancement, injects quantum correction terms into the gradient to enhance the heat dissipation response of vegetation.

[0056] As a specific embodiment, the calculation formula for the quantum-classical mixed scattering modeling is:

[0057]

[0058] in, is the quantum correction term, describing the quantum correlation effect between vegetation scatterers, β=(k B T) -1 ;

[0059] ε c (r, t) = ε o +kW v (t)·f rough(t), W v (t) is the time-varying vegetation water content, is the surface roughness attenuation function.

[0060] As a specific embodiment, in step 2, the time domain filtering of the radar echo adopts wavelet packet entropy filtering, the spatial domain filtering of the radar echo adopts tensor low rank decomposition, and the frequency domain filtering of the radar echo adopts quantum derivative filter.

[0061] As a specific embodiment, the quantum derivative filter is a frequency domain filter based on the Schrödinger equation;

[0062] The calculation formula is as follows:

[0063]

[0064] Used to simultaneously suppress ground clutter (low frequency) and radio frequency interference (high frequency), improving the signal-to-noise ratio by ≥15dB.

[0065] As a specific embodiment, the calculation formula of the cross-correlation objective function is as follows:

[0066]

[0067] in, is the original radar echo data, is the radar echo filtering data, ω k is the weight coefficient, which is used to focus on the strong signal area and suppress local noise interference.

[0068] As a specific embodiment, the regularization parameter is λ, and the calculation formula of λ is as follows:

[0069]

[0070] Among them, δ is the attenuation factor, and the empirical value is 0.01-0.1.

[0071] As a specific embodiment, the specific steps of step 4 are as follows:

[0072] S41, dynamic Hamiltonian evolution, through adiabatic evolution, the system transitions from the quantum superposition state to the classical ground state, obtaining the global minimum energy solution;

[0073] S42. Conversion of spin state to dielectric constant: converting the measurement results of quantum bits into dielectric constant distribution in physical space. The calculation formula is as follows: α is the linear mapping coefficient, which is determined by calibration experiments (typical value α≈0.1~0.5), ζ is the quantum fluctuation correction term coefficient, which is used to compensate for annealing noise. is the expected value of the spin in the z direction, obtained through multiple sampling statistics;

[0074] S43, iterative feedback mechanism, combines the fast initial solution of classical algorithms with the global optimization capability of quantum annealing, balancing efficiency and accuracy;

[0075] S44, quantum noise suppression, suppresses the inherent noise of quantum hardware through Boltzmann weighted averaging and improves inversion stability.

[0076] The above description is for inspiration. Based on the above description, relevant personnel can make various changes and modifications without departing from the technical concept of this invention. The technical scope of this invention is not limited to the content of the specification, but must be determined according to the scope of the claims.

Claims

1. A low-altitude detection radar inverse scattering method based on Born approximation, characterized by: The following steps are involved: Step 1: Modeling, using quantum-classical mixed scattering modeling and introducing Bose-Einstein condensate coherent state operators Describe the quantum correlation effect between vegetation scatterers, add dynamic dielectric coupling, and track the vegetation moisture content and the coupling effect between surface roughness and vegetation in real time; Step 2: Filtering: Perform three-dimensional filtering on the original radar echo data to obtain radar echo filtered data; Step 3: Adaptive full waveform inversion, which segments and weights the data obtained in step 2 and then adapts it to the model in step 1; Step 4: Solve the cross-correlation objective function in step 3 through Ising model mapping and quantum tunneling effect, and output the target scattering point cloud; Step 5: Multipath decoupling imaging uses an imaging operator to separate the direct wave and multipath reflected waves through mathematical modeling, eliminating vegetation shadows; Among them, step three includes the following steps: S31, data segmentation and weighting, segmenting the original radar echo data and the radar echo filter data into N sub-regions, and calculating the cross-correlation objective function; S32, adaptive regularization, dynamically adjusts the Tikhonov regularization parameters based on the error distribution graph to balance the data fitting and model complexity of S31 and avoid overfitting; S33, multi-scale decomposition inversion, decomposes the data into low-frequency, medium-frequency, and high-frequency components, and independently inverts the scattering model at each scale; S34, quantum scattering enhancement, injects quantum correction terms into the gradient to enhance the heat dissipation response of vegetation.

2. The low-altitude detection radar inverse scattering method based on Born approximation according to claim 1, characterized in that: The calculation formula for the quantum-classical mixed scattering modeling is: in, is the quantum correction term, describing the quantum correlation effect between vegetation scatterers, ε c (r, t) = ε o +kW v (tf rough (t), W v (t) is the time-varying vegetation water content, is the surface roughness attenuation function.

3. The low-altitude detection radar inverse scattering method based on Born approximation according to claim 1, characterized in that: In the step 2, the time domain filtering of the radar echo adopts wavelet packet entropy filtering, the spatial domain filtering of the radar echo adopts tensor low rank decomposition, and the frequency domain filtering of the radar echo adopts quantum derivative filter.

4. The low-altitude detection radar inverse scattering method based on Born approximation according to claim 3 is characterized in that: The quantum derivative filter is a frequency domain filter based on the Schrödinger equation; The calculation formula is as follows: Used to simultaneously suppress ground clutter (low frequency) and radio frequency interference (high frequency), improving the signal-to-noise ratio by ≥15dB.

5. The low-altitude detection radar inverse scattering method based on Born approximation according to claim 1, characterized in that: The calculation formula of the cross-correlation objective function is as follows: in, is the original radar echo data, is the radar echo filtering data, ω k is the weight coefficient.

6. The low-altitude detection radar inverse scattering method based on Born approximation according to claim 5, characterized in that: The regularization parameter is λ, and the calculation formula of λ is as follows: Among them, δ is the attenuation factor, and the empirical value is 0.01-0.

1.

7. The low-altitude detection radar inverse scattering method based on Born approximation according to claim 1, characterized in that: The specific steps of step 4 are as follows: S41, dynamic Hamiltonian evolution, through adiabatic evolution, the system transitions from the quantum superposition state to the classical ground state, obtaining the global minimum energy solution; S42, conversion of spin state to dielectric constant, converting the measurement results of quantum bits into dielectric constant distribution in physical space; S43, iterative feedback mechanism, combines the fast initial solution of classical algorithms with the global optimization capability of quantum annealing, balancing efficiency and accuracy; S44, quantum noise suppression, suppresses the inherent noise of quantum hardware through Boltzmann weighted averaging and improves inversion stability.

Images