Large-size target low-frequency-band scattering characteristic extrapolation technology

Abstract

The invention discloses a large-size target low-frequency-band scattering characteristic extrapolation method, and relates to the technical field of radar target scattering characteristic testing and imaging. The method comprises the following steps of: establishing a geometric diffraction (GTD) scattering center model of a target by utilizing scattering test data of a radar cross section (RCS) of a large-size target at a high frequency band, and realizing scattering imaging by adopting a parameter spectrum estimation technology; and then a scattering center model of a target low-frequency band is obtained by using the frequency characteristic of the scattering center, and finally RCS scattering data of the target low-frequency band are reconstructed, so that low-frequency-band extrapolation of the scattering characteristic of the large-size target is realized. The method has high scattering characteristic extrapolation precision for a large-size target, the physical concept is clear, implementation is easy, the difficulty of low-frequency-band RCS testing can be solved, and the RCS testing bandwidth is expanded.

Description

Technical Field

[0001] The present invention relates to the field of electromagnetic field and microwave technology, and in particular to a radar target scattering characteristic testing technology and a radar target scattering characteristic imaging technology. Background Art

[0002] RCS is a key parameter that characterizes the reflection and scattering ability of a target for the emitted electromagnetic waves. The basic scattering characteristics of a target can be understood through RCS. Studying the RCS of a target is of great significance for stealth design, radar detection and target identification.

[0003] The target RCS can be obtained by means of testing. At present, the measurement method of the target RCS mainly adopts indoor compact field darkroom measurement and outdoor field measurement. In order to ensure the accuracy of the test data, the RCS measurement has high requirements for the test environment and test equipment, and the construction cost of the test site is very high, usually reaching tens of millions of yuan or even hundreds of millions of yuan. Since the wavelength of electromagnetic waves in the low-frequency band is relatively long, in order to meet the test requirements, it is necessary not only to increase the test site, but also to have higher requirements for the test equipment and test environment. Therefore, the RCS test field constructed in the early stage of China usually meets the L-band and above, and cannot achieve RCS testing in lower frequency bands, such as the P-band. In order to meet the RCS test requirements of the low-frequency band, the present invention proposes to use the large-size target echo RCS data obtained by high-frequency band testing, complete the fitting of the frequency characteristic curve by establishing a target broadband scattering center model, and then extrapolate it to the RCS value of the low-frequency band. The extrapolation of low-frequency band test data is conducive to reducing the measurement workload, reducing the requirements for existing test equipment, and increasing the test frequency band range of existing test fields. If a good prediction of the unknown low-frequency RCS value can be obtained, knowing the RCS size of the wide-band will be more conducive to analyzing the scattering characteristics of the target. Summary of the invention

[0004] The present invention intends to provide a method for extrapolating low-frequency RCS scattering of large-sized targets. The method is based on the principle of high-frequency scattering imaging of large-sized targets, and uses high-frequency scattering data obtained by testing the target in a compact microwave darkroom or an outdoor test field to establish a GTD scattering center model; then, based on the parameter spectrum estimation technology, large-sized target scattering center imaging is carried out, and a low-frequency scattering center model is established using the frequency characteristics of the scattering center; finally, the GTD scattering center model is used to reconstruct the low-frequency target RCS scattering data, and the low-frequency extrapolation of the scattering characteristics is realized to solve the problems raised in the above-mentioned background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A low-frequency band scattering characteristic extrapolation technology for large-size targets, the technical method specifically comprises the following steps:

[0007] S1. Obtain high-frequency RCS scattering data and establish a GTD scattering center model;

[0008] S2. Estimate the number of scattering centers, using the Gale circle algorithm to estimate the number of scattering centers of large-size targets;

[0009] S3. Estimate the position of the scattering center using the TLS-ESPRIT algorithm, where the TLS-ESPRIT algorithm adds a total least squares constraint based on the ESPRIT algorithm to suppress the influence of noise on the position estimation;

[0010] S4. Estimating scattering center type parameters, using the orthogonality of the signal and noise subspaces to estimate the scattering center type parameters;

[0011] S5. Estimate the scattering center intensity. The target backscatter electric field frequency response can be characterized by the GTD scattering center model. The scattering center intensity is extracted based on the least squares algorithm.

[0012] S6. Reconstruct the target low-frequency band RCS scattering data based on the estimated large-size target GTD scattering center model to achieve the extrapolation of low-frequency band RCS.

[0013] Furthermore, the step S1 specifically includes:

[0014] S11. Get the RCS scattering data of large-size targets in the high-frequency band in a compact darkroom or outdoor test. The data is the high-frequency band broadband RCS amplitude and phase data under given polarization and illumination angle conditions;

[0015] S12. Construct a GTD scattering center model for large-sized targets. Large-sized targets in the low frequency band, that is, the P band, are suitable for the GTD model. The scattering center parameters of the target can be extracted. Under broadband radar conditions, the GTD scattering center model of the target's backscattered electric field is:

[0016]

[0017] D i =exp(-4πjf m r i / c)

[0018] Where I is the number of scattering centers; A i is the scattering intensity; r i is the scattering center position vector; f m =f 0 +mΔf,f 0 is the initial frequency, Δf is the step frequency, m is the number of steps; c represents the propagation speed of electromagnetic waves; w(f m ) is additive complex Gaussian white noise; αi Characterizes the scattering type corresponding to the i-th scattering center, where the parameter α takes different values ​​according to the different scatterer structures and is an integer multiple of 1 / 2.

[0019] Furthermore, the step S2 specifically includes:

[0020] S21. The criterion for determining the number of scattering centers using the Gale circle algorithm is:

[0021]

[0022] Where D(L)∈[0,1] is an adjustment factor related to the number of snapshots; k∈[1,m-2], when k is increased from small to large, it is assumed that the first time GDE(k) appears negative is k 0 , then the number of scattering centers is k 0 -1.

[0023] Furthermore, the step S3 adopts an improved TLS-ESPRIT algorithm to estimate and extract the scattering center position parameters, specifically including:

[0024] S31. Select 2 P×1 dimensional vectors x from the target backscattered electric field data 1 、x 2 ,in:

[0025] x 1 (m) = [E(f m ),E(f m+1 ),...,E(f m+P-1 )] T

[0026] x 2 (m) = [E(f m+1 ),E(f m+2 ),...,E(f m+P )] T =x 1 (m+1)

[0027] In the formula, E(f m ),E(f m+1 ),...,E(f m+P-1 ) is the target backscattering field data corresponding to different frequencies. 1 (m), x 2 (m) constructing a matrix;

[0028] S32. Find x 1 (m) Autocorrelation matrix R zz , and for R zz Enter eigenvalue decomposition;

[0029] S33. Extraction of R zz The signal subspace U S , and divide it into 2 parts U S1 and U S2 , construct U S12 =[U S1 ,U S2 ];

[0030] S34. Calculate the following eigenvalue decomposition:

[0031] U S12 H U S12 =FΛΛ H , and decompose F into 4 sub-matrices of k×k, namely:

[0032]

[0033] S35. Calculation The position parameter r of the scattering center can be estimated i .

[0034] Furthermore, the step S4 estimates the scattering center type parameter using the orthogonality of the signal and noise subspaces, specifically including:

[0035] S41. Calculate α using the orthogonality of the signal subspace and the noise subspace in the observation space. i , using the GTD model to generate the signal pattern vector, namely:

[0036]

[0037] R zz Perform singular value decomposition, where V' is a matrix consisting of non-principal right singular vectors, which corresponds to the noise subspace of the signal;

[0038] S42. The inverse of the projection modulus of the signal pattern vector on the noise subspace can be expressed as:

[0039]

[0040] The more similar the signal pattern vector is to the signal, the smaller its projection modulus value on the noise subspace is. For S35, the position coordinates r of I scattering points have been obtained. i , then for every r i value, α selects different types; when α and position are r i When the true type at matches, the signal pattern vector is most consistent with the observed data, that is, the inverse of its projection value P(α, r) on the noise subspace is the largest, thereby determining the scattering types of different scattering centers.

[0041] Furthermore, the step S5 uses a least squares algorithm to estimate the scattering center intensity, specifically including:

[0042] S51. The scattered echo of the scattering center model can be expressed by the following formula:

[0043] X=Α′S * +w

[0044] In the formula, X=[E(f 0 ),E(f 1 ),...,E(f m-1 )] T is the target scattered echo data at different frequencies; S * =[A 1 ,A 2 ,...,A I ] T is the scattering intensity of different scattering centers; w is the Gaussian white noise at different frequencies;

[0045] S52. Get the matrix from the type parameter and position parameter:

[0046]

[0047] The calculation results are:

[0048] A′=[α′(α i ,r 1 ),α′(α 2 ,r 2 ),...,α′(α i ,r 1 )] T

[0049] S53.X is the known echo electric field data, so the estimated scattering intensity parameters are transformed into the unknown matrix S * According to the TLS-ESPRIT algorithm, we can get:

[0050] S * =(A′ H A′) -1 A′ H X

[0051] Furthermore, the step S6 implements low-frequency band RCS scattering extrapolation, specifically including:

[0052] S61. The above four scattering parameters are extracted by using high-frequency broadband RCS scattering data and scattering center parameter extraction algorithm, thereby constructing the GTD scattering center model of the target. The target backscattering electric field frequency response can be characterized by the GTD scattering center model:

[0053]

[0054] S62. Get the RCS frequency response calculation formula:

[0055]

[0056] The target RCS is extrapolated in the low frequency band according to the above formula.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. The method provided by the present invention effectively utilizes the RCS test data of the higher frequency band to realize the RCS extrapolation of the low frequency band, and can effectively expand the frequency band range of the existing RCS test system.

[0059] 2. The method provided by the present invention is only for processing test data, and there is no need to transform the existing RCS test system into a low-frequency band, thereby saving a lot of construction funds.

[0060] 3. The method provided by the present invention utilizes the GTD scattering center model of the target RCS scattering characteristics. The broadband frequency domain response characteristics of the model are the theoretical guarantee for realizing low-frequency extrapolation.

[0061] 4. The low-frequency band RCS extrapolation function of the large-size target of the present invention can also be applied and expanded to realize the RCS scattering extrapolation from the low-frequency band to the high-frequency band, and the fitting of RCS data within a wide frequency band. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 A flowchart of a technique for extrapolating the low-frequency scattering characteristics of a large-size target;

[0063] Figure 2 This is a schematic diagram of the principle of the Gale circle algorithm;

[0064] Figure 3 It is the flow chart of TLS-ESPRIT algorithm;

[0065] Figure 4 It is the flow chart of least squares algorithm;

[0066] Figure 5 Display diagram of the aircraft model used for RCS extrapolation testing;

[0067] Figure 6 This is a comparison diagram of the RCS extrapolation results at an azimuth of 0°;

[0068] Figure 7 This is a schematic diagram for comparing the RCS extrapolation results at an azimuth angle of 30°. DETAILED DESCRIPTION

[0069] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments:

[0070] like Figure 1 As shown, a low-frequency band scattering characteristic extrapolation technology for large-size targets includes the following specific steps:

[0071] S1. Obtain high-frequency RCS scattering data and establish a GTD scattering center model. For the scattering problem of large-size targets, the GTD model accurately describes the radar targets in the optical zone, and its expression is close to the actual target scattering mechanism. The parameters of the target scattering center are extracted based on the GTD model.

[0072] S2. Estimate the number of scattering centers. Use the Gale circle algorithm to estimate the number of scattering centers of large-sized targets. The Gale circle method uses the radius of the Gale circle of the scattered data signal covariance matrix to estimate the number of scattering centers (scattering sources). Usually, there is no obvious difference between the signal Gale circle and the noise Gale circle of the covariance matrix. It is necessary to perform a certain transformation on the covariance matrix according to the Gale circle theorem so that the signal Gale circle radius of the transformed covariance matrix is ​​significantly larger than the noise Gale circle radius, that is, the Gale disk of the transformed covariance matrix is ​​divided into two parts with different radii. The part with a larger radius is the signal Gale circle, and the part with a smaller radius is the noise Gale circle. In this way, a reliable estimation of the number of source signals is achieved based on the Gale circle radius.

[0073] S3. Estimate the position of the scattering center and use the least squares-rotation vector invariant technique (TLS-ESPRIT) to estimate the position of the scattering center. The ESPRIT algorithm is based on the subspace method and uses the second-order statistical characteristics of the echo signal to decompose the signal into different subspaces and estimate the position of the scattering center in the subspace. The TLS-ESPRIT algorithm adds a total least squares constraint to the ESPRIT algorithm to suppress the impact of noise on position estimation. Specifically, the TLS-ESPRIT algorithm decomposes the signal into two subspaces, then estimates the position of the scattering center in the first subspace, and then uses the total least squares method to correct it in the second subspace to obtain a more accurate position estimation result.

[0074] S4. Estimate the scattering center type parameters. Different scattering structures correspond to different scattering center type parameters. Even for the same scattering structure, different type parameters will correspond to different observation angles. Therefore, it is necessary to estimate and extract the scattering type parameters in real time to ensure the accuracy of the scattering center model in describing the target scattering characteristics. The orthogonality of the signal and noise subspaces is used to estimate the scattering center type parameters.

[0075] S5. Estimate the scattering center intensity. The target backscatter electric field frequency response can be characterized by the GTD scattering center model. The scattering center intensity is extracted based on the TLS-ESPRIT algorithm.

[0076] S6. Reconstruct the target low-frequency band RCS scattering data based on the estimated large-size target GTD scattering center model to achieve the extrapolation of low-frequency band RCS.

[0077] The step S1 specifically includes:

[0078] S11. Get the RCS scattering data of large-size targets in the high-frequency band in a compact darkroom or outdoor test. The data is the high-frequency band broadband RCS amplitude and phase data under given polarization and illumination angle conditions;

[0079] S12. Construct a GTD scattering center model for large-sized targets. Large-sized targets in the low frequency band, that is, the P band, are suitable for the GTD model. The scattering center parameters of the target can be extracted. Under broadband radar conditions, the GTD scattering center model of the target's backscattered electric field is:

[0080]

[0081] D i =exp(-4πjf m r i / c)

[0082] Where I is the number of scattering centers; A i is the scattering intensity; r i is the scattering center position vector; f m =f 0 +mΔf,f 0 is the initial frequency, Δf is the step frequency, m is the number of steps; c represents the propagation speed of electromagnetic waves; w(f m ) is additive complex Gaussian white noise; α i Characterizes the scattering type corresponding to the i-th scattering center, where the parameter α takes different values ​​according to the different scatterer structures and is an integer multiple of 1 / 2. For different typical scattering structures, the values ​​of α are shown in Table 1:

[0083] Table 1: α values ​​for typical scattering structures

[0084]

[0085]

[0086] The step S2 specifically includes:

[0087] S21. The criterion for determining the number of scattering centers using the Gale circle algorithm is:

[0088]

[0089] Where D(L)∈[0,1] is an adjustment factor related to the number of snapshots; k∈[1,m-2], when k is increased from small to large, it is assumed that the first time GDE(k) appears negative is k 0 , then the number of scattering centers is k 0 -1. The algorithm flow chart is as follows Figure 2 Shown

[0090] like Figure 3 As shown, the step S3 adopts the improved TLS-ESPRIT algorithm to estimate and extract the scattering center position parameters, which specifically includes:

[0091] S31. Select 2 P×1 dimensional vectors x from the target backscattered electric field data 1 、x 2 ,in:

[0092] x 1 (m) = [E(f m ),E(f m+1 ),...,E(f m+P-1 )] T

[0093] x 2 (m) = [E(f m+1 ),E(f m+2 ),...,E(f m+P )] T =x 1 (m+1)

[0094] In the formula, E(f m ),E(f m+1 ),...,E(f m+P-1 ) is the target backscattering field data corresponding to different frequencies. 1 (m), x 2 (m) constructing a matrix;

[0095] S32. Find x 1 (m) Autocorrelation matrix R zz , and for R zz Enter eigenvalue decomposition;

[0096] S33. Extraction of R zz The signal subspace U S , and divide it into 2 parts U S1 and U S2 , construct U S12 =[U S1 ,U S2 ];

[0097] S34. Calculate the following eigenvalue decomposition:

[0098] U S12 H U S12 =FΛΛ H , and decompose F into 4 sub-matrices of k×k, namely:

[0099]

[0100] S35. Calculation The position parameter r of the scattering center can be estimated i .

[0101] The step S4 estimates the scattering center type parameter by using the orthogonality of the signal and noise subspaces, and specifically includes:

[0102] S41. Calculate α using the orthogonality of the signal subspace and the noise subspace in the observation space. i , using the GTD model to generate the signal pattern vector, namely:

[0103]

[0104] R zz Perform singular value decomposition, where V' is a matrix consisting of non-principal right singular vectors, which corresponds to the noise subspace of the signal;

[0105] S42. The inverse of the projection modulus of the signal pattern vector on the noise subspace can be expressed as:

[0106]

[0107] The more similar the signal pattern vector is to the signal, the smaller the projection modulus value on the noise subspace is. The position coordinates r of I scattering points have been obtained above. i , then for every r i value, α selects different types; when α and position are r i When the true type at matches, the signal pattern vector is most consistent with the observed data, that is, the inverse of its projection value P(α, r) on the noise subspace is the largest, thereby determining the scattering types of different scattering centers.

[0108] like Figure 4 As shown, the step S5 uses the least squares algorithm to estimate the scattering center intensity, specifically including:

[0109] S51. The scattered echo of the scattering center model can be expressed by the following formula:

[0110] X=Α′S * +w

[0111] In the formula, X=[E(f 0 ),E(f1 ),...,E(f m-1 )] T is the target scattered echo data at different frequencies; S * =[A 1 ,A 2 ,...,A I ] T is the scattering intensity of different scattering centers; w is the Gaussian white noise at different frequencies;

[0112] S52. Get the matrix from the type parameter and position parameter:

[0113]

[0114] The calculation results are:

[0115] A′=[α′(α i ,r 1 ),α′(α 2 ,r 2 ),...,α′(α i ,r 1 )] T

[0116] S53.X is the known echo electric field data, so the estimated scattering intensity parameters are transformed into the unknown matrix S * According to the TLS-ESPRIT algorithm, we can get:

[0117] S * =(A′ H A′) -1 A′ H X

[0118] Preferably, the step S6 implements low-frequency band RCS scattering extrapolation, specifically comprising:

[0119] S61. The above four scattering parameters are extracted by using high-frequency broadband RCS scattering data and scattering center parameter extraction algorithm, thereby constructing the GTD scattering center model of the target. The target backscattering electric field frequency response can be characterized by the GTD scattering center model:

[0120]

[0121] S62. Get the RCS frequency response calculation formula:

[0122]

[0123] The target RCS is extrapolated in the low frequency band according to the above formula.

[0124] The specific implementation process is as follows:

[0125] like Figure 5 The target aircraft shown has a length of 36.68m and a wingspan of 38.00m. The RCS data in the frequency band 1 to 2GHz is used to extrapolate the RCS data in the frequency band 0.5 to 1GHz. Echo test angles: pitch angle 0°, azimuth angles 0° and 30°.

[0126] Step 1: Obtain the RCS data of the aircraft target in the high frequency band (1-2 GHz) at a pitch angle of 0° and an azimuth angle of 0°, including the amplitude and phase of the RCS;

[0127] Step 2: Process the echo data and first determine the number of scattering centers based on the Gale circle algorithm;

[0128] Step 3: Then use the TLS-ESPRIT algorithm to estimate the position of each scattering center;

[0129] Step 4: Estimate the scattering center type parameters based on the nearest neighbor criterion;

[0130] Step 5: Use the least squares method to estimate the intensity of each scattering center;

[0131] Step 6: Use the frequency domain characteristics of the GTD scattering center model to calculate the RCS in the low frequency band (0.5 GHz to 1 GHz).

[0132] Step 7: For the elevation angle of 0° and the azimuth angle of 30°, extrapolate using the same processing steps as above.

[0133] Figure 6 The extrapolated results for an azimuth angle of 0° are given for comparison, with the left figure for HH polarization and the right figure for VV polarization. The average RCS errors are 1.18 dB and 0.95 dB, respectively.

[0134] Figure 7 The extrapolated results at an azimuth angle of 30° are given, where the left figure is HH polarization and the right figure is VV polarization, with average RCS errors of 0.57dB and 0.58dB respectively.

[0135] The above is only an embodiment of the present invention, and the common knowledge such as the known specific technical solutions and / or characteristics in the solution is not described in detail here. It should be pointed out that for those skilled in the art, without departing from the technical solution of the present invention, several modifications and improvements can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A low-frequency band scattering characteristic extrapolation technology for large-size targets, characterized in that: The specific steps of the technical method include: S1. Obtain high-frequency RCS scattering data and establish a GTD scattering center model; S2. Estimate the number of scattering centers, using the Gale circle algorithm to estimate the number of scattering centers of large-size targets; S3. Estimate the position of the scattering center using the TLS-ESPRIT algorithm, where the TLS-ESPRIT algorithm adds a total least squares constraint based on the ESPRIT algorithm to suppress the influence of noise on the position estimation; S4. Estimating scattering center type parameters, using the orthogonality of the signal and noise subspaces to estimate the scattering center type parameters; S5. Estimate the scattering center intensity. The target backscatter electric field frequency response can be characterized by the GTD scattering center model. The scattering center intensity is extracted based on the least squares algorithm. S6. Reconstruct the target low-frequency band RCS scattering data based on the estimated large-size target GTD scattering center model to achieve the extrapolation of low-frequency band RCS.

2. The low-frequency band scattering characteristic extrapolation technology of a large-size target according to claim 1 is characterized in that: The step S1 specifically includes: S11. Get the RCS scattering data of large-size targets in the high-frequency band in a compact darkroom or outdoor test. The data is the high-frequency band broadband RCS amplitude and phase data under given polarization and illumination angle conditions; S12. Construct a GTD scattering center model for large-sized targets. Large-sized targets in the low frequency band, that is, the P band, are suitable for the GTD model. The scattering center parameters of the target can be extracted. Under broadband radar conditions, the GTD scattering center model of the target's backscattered electric field is: D i =exp(-4πjf m r i / c) Where I is the number of scattering centers; A i is the scattering intensity; r i is the scattering center position vector; f m =f0+mΔf, f0 is the initial frequency, Δf is the step frequency, m is the number of steps; c represents the propagation speed of electromagnetic waves; w(f m ) is additive complex Gaussian white noise; α i Characterizes the scattering type corresponding to the i-th scattering center, where the parameter α takes different values ​​according to the different scatterer structures and is an integer multiple of 1 / 2.

3. The low-frequency band scattering characteristic extrapolation technology of a large-size target according to claim 1 is characterized in that: The step S2 specifically includes: S21. The criterion for determining the number of scattering centers using the Gale circle algorithm is: Where D(L)∈[0,1] is an adjustment factor related to the number of snapshots; k∈[1,m-2], when k increases from small to large, assuming that the first negative number in GDE(k) corresponds to k0, the number of scattering centers is k0-1.

4. The low-frequency band scattering characteristic extrapolation technology of a large-size target according to claim 1 is characterized in that: The step S3 adopts the improved TLS-ESPRIT algorithm to estimate and extract the scattering center position parameters, which specifically includes: S31. Select two P×1 dimensional vectors x1 and x2 from the target backscattered electric field data, where: x1(m)=[E(f m ),E(f m+1 ),...,E(f m+P-1 )] T x2(m)=[E(f m+1 ),E(f m+2 ),...,E(f m+P )] T =x1(m+1) In the formula, E(f m ),E(f m+1 ),...,E(f m+P-1 ) is the target backscattering field data corresponding to different frequencies, and the matrix is ​​constructed according to x1(m) and x2(m); S32. Find the autocorrelation matrix R of x1(m) zz , and for R zz Enter eigenvalue decomposition; S33. Extraction of R zz The signal subspace U S , and divide it into 2 parts U S1 and U S2 , construct U S12 =[U S1 ,U S2 ]; S34. Calculate the following eigenvalue decomposition: U S12 H U S12 =FΛΛ H , and decompose F into 4 sub-matrices of k×k, namely: S35. Calculation The position parameter r of the scattering center can be estimated i .

5. The low-frequency band scattering characteristic extrapolation technology of a large-size target according to claim 1 is characterized in that: The step S4 estimates the scattering center type parameter by using the orthogonality of the signal and noise subspaces, and specifically includes: S41. Calculate α using the orthogonality of the signal subspace and the noise subspace in the observation space. i , using the GTD model to generate the signal pattern vector, namely: R zz Perform singular value decomposition, where V' is a matrix consisting of non-principal right singular vectors, which corresponds to the noise subspace of the signal; S42. The inverse of the projection modulus of the signal pattern vector on the noise subspace can be expressed as: The more similar the signal pattern vector is to the signal, the smaller its projection modulus value on the noise subspace is. For S35, the position coordinates r of I scattering points have been obtained. i , then for every r i value, α selects different types; when α and position are r i When the real type at matches, the signal pattern vector is most consistent with the observed data, that is, the inverse of its projection value P(α, r) on the noise subspace is the largest, thereby determining the scattering types of different scattering centers.

6. The low-frequency band scattering characteristic extrapolation technology of a large-size target according to claim 1 is characterized in that: The step S5 uses the least squares algorithm to estimate the scattering center intensity, which specifically includes: S51. The scattered echo of the scattering center model can be expressed by the following formula: X=A′S * +w In the formula, X=[E(f0),E(f1),...,E(f m-1 )] T is the target scattered echo data at different frequencies; S * =[A1,A2,...,A I ] T is the scattering intensity of different scattering centers; w is the Gaussian white noise at different frequencies; S52. Get the matrix from the type parameter and position parameter: The calculation results are: A′＝[α′(α i ,r1),α′(α2,r2),...,α′(α i ,r1)] T S53.X is the known echo electric field data, so the estimation of the scattering intensity parameters is transformed into finding the unknown matrix S * According to the TLS-ESPRIT algorithm, we can get: S * =(A′ H A′) -1 A′ H X。 7. The low-frequency band scattering characteristic extrapolation technology of a large-size target according to claim 1 is characterized in that: The step S6 implements low-frequency RCS scattering extrapolation, specifically including: S61. The above four scattering parameters are extracted by using high-frequency broadband RCS scattering data and scattering center parameter extraction algorithm, thereby constructing the GTD scattering center model of the target. The target backscattering electric field frequency response can be characterized by the GTD scattering center model: S62. Get the RCS frequency response calculation formula: The target RCS is extrapolated in the low frequency band according to the above formula.

Images