Underwater target positioning method based on underwater sound particle vibration velocity polarization processing

Abstract

The invention discloses an underwater target positioning method based on underwater sound particle vibration velocity polarization processing. The method comprises the steps that a vector hydrophone array is arranged, underwater acoustic signals are collected, and vector receiving signals of the vector hydrophone array are written into a tensor form; performing high-order singular value decomposition on a tensor receiving signal of the vector hydrophone array, and truncating an n-mode expansion left singular matrix to solve a noise subspace of each dimension information matrix; constructing an array manifold matrix of the vector hydrophone array and writing the array manifold matrix into a polarization array manifold tensor; and carrying out azimuth-polarization two-dimensional spectrum search by utilizing the array manifold tensor and the noise subspace, and obtaining azimuth and polarization parameters of the target to be detected according to a two-dimensional spectrum peak value. According to the underwater target positioning method provided by the invention, the polarization parameter and the direction of arrival of the incident signal can be accurately estimated, the underwater sound signal information can be complemented on the basis of the characteristics such as frequency, wave number and mode by means of the polarization parameter, and the underwater target detection capability is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vector hydrophone signal processing and underwater target detection, and specifically to a method for underwater target positioning based on the polarization processing of underwater acoustic particle velocity. Background Technique

[0002] Due to the influence of boundaries, medium inhomogeneity, coherent multi-sources, etc. in the actual ocean waveguide, the vibration direction of the particle deviates from the sound wave propagation direction, showing a phenomenon similar to polarization. In the traditional underwater acoustic model dominated by sound pressure, this objective microscopic phenomenon and fact cannot be faced squarely, making it impossible to truly unify the waveguide environment models of different sound fields.

[0003] Re-recognizing the physical properties of the acoustic particle velocity vector, especially the polarization property, is of great significance for deepening the basic science and engineering application research of underwater acoustics. Moreover, through polarization, a bridge of interconnection can be built between underwater acoustics and the research of electromagnetics, optics, and seismology. The disciplines intersect with each other, jointly promoting the development of wave physics and its applications, and providing more powerful support for solving underwater acoustic problems.

[0004] The acoustic particle velocity vector adds new polarization characteristics on the basis of the characteristics such as frequency, wave number, and mode that traditional underwater acoustics focuses on. It contains the key information of the interaction between underwater targets and the underwater acoustic waveguide. By measuring and processing the acoustic particle velocity vector, the characteristics of underwater targets and the underwater acoustic waveguide can be revealed respectively. The introduction of acoustic particle velocity polarization unifies the underwater acoustic models that interfere with each other, such as spherical waves, plane waves, and waveguide waves, that is, different underwater acoustic models are incorporated into a unified underwater acoustic polarization model, and the relevant information of underwater targets and the underwater acoustic waveguide is given through high-dimensional orthogonal decomposition. Compared with the acoustic field model with serious mismatch in underwater acoustic matched field processing, the underwater acoustic polarization model is always acoustically matched, and the focus has shifted from model mismatch to polarization parameter estimation. The tensor representation and tensor decomposition of the acoustic particle velocity vector polarization are helpful to obtain higher information fidelity and processing gain, providing a new perspective and opening up new ideas for solving many underwater acoustic problems. Summary of the Invention

[0005] Aiming at the problem that the waveguide environment models of different sound fields in the traditional model cannot be truly unified in form, the present invention proposes a method for underwater target positioning based on the polarization processing of underwater acoustic particle velocity. The present invention establishes an underwater acoustic polarization processing framework based on the invariance of the ellipse geometry and its tensor processing method, converts the equal-mode basis processing such as the matched sound field (mode) based on the accurate modeling of the sound field into a multi-dimensional parameter estimation problem based on the invariance of the polarization ellipse geometry, thereby reducing the dependence on the propagation model and the influence of ocean environment mismatch, and significantly improving the detection ability of the vector hydrophone array for underwater targets in complex ocean environments.

[0006] The object of the present invention is achieved as follows, and the steps are as follows:

[0007] Step 1: Arrange a vector hydrophone array in the form of a uniform linear array underwater to collect underwater acoustic signals of sound pressure scalars and particle velocity vectors.

[0008] Step 2: Obtain the polarization vectors of each vector hydrophone and the time delay vector of the vector hydrophone array through underwater acoustic signal acquisition.

[0009] Step 3: Calculate the steering matrix of the vector hydrophone array based on the polarization vectors and the time delay vector, and reconstruct it into a third-order tensor form, that is, obtain the steering tensor.

[0010] Step 4: Obtain the third-order tensor polarization output information of the vector hydrophone array based on the steering tensor of the vector hydrophone array.

[0011] Step 5: Perform high-order singular value decomposition on the third-order tensor polarization output information of the vector hydrophone array.

[0012] Step 6: Truncate and solve the left singular matrix of the 1-mode, 2-mode, and 3-mode expanded by high-order singular value decomposition to obtain the noise subspace matrix of each dimension.

[0013] Step 7: Use the steering tensor and the noise subspace to perform azimuth-polarization two-dimensional spectrum search, and determine the azimuth angle and polarization parameters of the detection target according to the direction of arrival and polarization parameters corresponding to the azimuth-polarization two-dimensional spectrum peak obtained by the search.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention proposes an underwater target positioning method based on underwater acoustic particle velocity polarization processing, introduces polarization parameters into the signal model, solves the mismatch of the short-range distance model in the actual ocean waveguide of the plane wave model, and improves the underwater target detection accuracy.

[0015] The polarization signal model of the present invention can compromise between traditional plane wave processing and classical matched field (mode) processing to better adapt to the actual received sound field, can not only make full use of sound field information, especially particle velocity information, but also reduce the model mismatch caused by spatio-temporal variations in complex environments. Description of the Drawings

[0016] The drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure;

[0017] Figure 1 It is a schematic diagram of the underwater acoustic particle velocity polarization ellipse of the vector hydrophone uniform linear array and the receiving points of array elements provided by an embodiment of the present invention;

[0018] Figure 2It is a flowchart of an underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing provided by an embodiment of the present invention;

[0019] Figure 3 It is a one-dimensional spatial spectrum comparison diagram of the horizontal azimuth angle estimation results of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (traditional long vector model and tensor model) and the traditional long vector MUSIC DOA estimation algorithm based on plane wave model provided by an embodiment of the present invention;

[0020] Figure 4 It is a one-dimensional spatial spectrum comparison diagram of the elevation angle estimation results of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (traditional long vector model and tensor model) and the traditional long vector MUSIC DOA estimation algorithm based on plane wave model provided by an embodiment of the present invention;

[0021] Figure 5 It is a one-dimensional spatial spectrum comparison diagram of the elliptic inclination angle estimation results of the MUSIC DOA - polarization parameter joint estimation algorithm of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (traditional long vector model and tensor model) provided by an embodiment of the present invention;

[0022] Figure 6 It is a one-dimensional spatial spectrum comparison diagram of the ellipticity angle estimation results of the MUSIC DOA - polarization parameter joint estimation algorithm of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (traditional long vector model and tensor model) provided by an embodiment of the present invention;

[0023] Figure 7 It is a two-dimensional spatial spectrum diagram of the DOA estimation results of the MUSIC DOA - polarization parameter joint estimation algorithm of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (traditional long vector model) provided by an embodiment of the present invention;

[0024] Figure 8 It is a two-dimensional spatial spectrum diagram of the DOA estimation results of the MUSIC DOA - polarization parameter joint estimation algorithm of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (tensor model) provided by an embodiment of the present invention;

[0025] Figure 9 It is a two-dimensional spatial spectrum diagram of the polarization parameter estimation results of the MUSIC DOA - polarization parameter joint estimation algorithm of the underwater target localization method based on vector hydrophone array for underwater acoustic particle velocity polarization processing (traditional long vector model) provided by an embodiment of the present invention;

[0026] Figure 10It is the two-dimensional spatial spectrogram of the polarization parameter estimation result of the tensor model MUSIC DOA - polarization parameter joint estimation algorithm for the underwater target localization method based on the vector hydrophone array and underwater acoustic particle velocity polarization processing provided by the embodiment of the present invention;

[0027] Figure 11 It is the comparison diagram of the root mean square error of the horizontal azimuth angle estimation results of the underwater target localization method (traditional long vector model and tensor model) based on the vector hydrophone array and underwater acoustic particle velocity polarization processing and the traditional long vector MUSIC DOA estimation algorithm based on the plane wave model;

[0028] Figure 12 It is the comparison diagram of the root mean square error of the elevation angle estimation results of the underwater target localization method (traditional long vector model and tensor model) based on the vector hydrophone array and underwater acoustic particle velocity polarization processing and the traditional long vector MUSIC DOA estimation algorithm based on the plane wave model;

[0029] Figure 13 It is the comparison diagram of the root mean square error of the elliptic inclination angle estimation results of the tensor model MUSIC DOA - polarization parameter joint estimation algorithm for the underwater target localization method based on the vector hydrophone array and underwater acoustic particle velocity polarization processing with the traditional long vector model;

[0030] Figure 14 It is the comparison diagram of the root mean square error of the ellipticity angle estimation results of the tensor model MUSIC DOA - polarization parameter joint estimation algorithm for the underwater target localization method based on the vector hydrophone array and underwater acoustic particle velocity polarization processing with the traditional long vector model. Specific Embodiments

[0031] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] Combined with Figure 1 and Figure 2 , the underwater target localization method based on the vector hydrophone array and underwater acoustic particle velocity polarization processing proposed by the present invention first establishes a particle velocity polarization signal model, forms a polarization vector composed of the hydrophone sound pressure and particle velocity components containing DOA parameters and polarization ellipse parameters, and the time delay vector of the vector hydrophone uniform linear array, establishes a steering vector through the Kronecker product, and reconstructs it into a steering tensor. According to the above steering tensor, the received signal is transformed into a third-order tensor form, and the polarization signal subspace and noise subspace in each dimension are obtained through high-order singular value decomposition. Combining with the traditional MUSIC algorithm, a signal processing method based on underwater acoustic polarization is proposed to estimate the DOA of the incident signal and the polarization parameters of the particle velocity polarization ellipse at the positions of the receiving array elements.

[0033] Combined with Figure 2 , in this embodiment, the underwater target positioning method based on the polarization processing of the underwater acoustic particle velocity of the vector hydrophone array includes the following steps:

[0034] Step 1: Arrangement of the vector hydrophone array and construction of the polarization driving vector.

[0035] In the underwater environment, an array composed of M four-component vector hydrophones is arranged in the form of a uniform linear array. Each vector hydrophone can synchronously measure the sound pressure signal (scalar) and the X / Y / Z three components of the particle velocity (vector). The underwater acoustic signal is collected by using the vector hydrophone array, and the polarization driving vector composed of the sound pressure of a single vector hydrophone and the three components of the particle velocity is constructed as follows:

[0036]

[0037] where u is the polarization vector of a single vector hydrophone, u x is the polarization component in the X-axis direction, u y is the polarization component in the Y-axis direction, u z is the polarization component in the Z-axis direction, θ ∈ [-π, π] is the horizontal azimuth angle of the vector hydrophone, is the pitch angle of the vector hydrophone, β ∈ [-π / 2, π / 2] is the elliptical inclination angle of the vector hydrophone, and α ∈ [-π / 4, π / 4] is the ellipticity angle of the vector hydrophone.

[0038] Step 2: Construct the time delay vector of the vector hydrophone array.

[0039] Since the vector hydrophone array is a uniform linear array, its array time delay vector a is specifically:

[0040]

[0041] where λ is the wavelength of the incident signal, d is the spacing between adjacent vector hydrophones in the array, and M is the number of vector hydrophones.

[0042] Step 3: Construct the driving tensor of the array.

[0043] The driving matrix of the vector hydrophone array is formed by using the Kronecker product of the driving vector and the time delay vector of the vector hydrophone array, and it is written as the driving tensor of the array. The specific calculation formula is:

[0044]

[0045] where A is the driving matrix, u1, u2, u M are the polarization vectors of the 1st, 2nd, and Mth vector hydrophones, a is the time delay vector of the vector hydrophone array, represents the tensor product;

[0046] If the vector hydrophone array is a uniform linear array and the array elements are four-component vector hydrophones, the specific tensor of the array manifold is as follows:

[0047]

[0048] Among them,

[0049] In the formula, k is the number of incident signals received by the vector hydrophone array, is the slice of the array steering tensor of the k-th incident signal in the dimension of the number of signals, u xk , u yk and u zk are the polarization components of the particle velocity polarization vector of the k-th incident signal in the three directions of the velocity x-axis, y-axis, and z-axis, θ k is the horizontal azimuth angle of the k-th incident signal, is the elevation angle of the k-th incident signal, β k is the elliptic tilt angle of the particle velocity polarization ellipse of the k-th incident signal at the array element receiving point, α k is the ellipticity angle of the particle velocity polarization ellipse of the k-th incident signal at the array element receiving point, and M is the number of vector hydrophones in the array.

[0050] Step 4: Obtain the third-order tensor polarization output information of the vector hydrophone array according to the array steering tensor.

[0051] In L fast snapshots, the third-order tensor form output by the vector hydrophone array can be expressed as:

[0052]

[0053] Among them, is the third-order tensor output information, is the array steering tensor, ×3 is the 3-mode tensor product operation, is the sound source signal waveform, is the Gaussian noise tensor, and K is the total number of incident signals.

[0054] Step 5: Perform high-order singular value decomposition (HOSVD) on the third-order tensor output of the vector hydrophone array;

[0055] Performing high-order singular value decomposition on the array received signals in tensor form gives:

[0056]

[0057] Among them, is the core tensor of the output signal tensor , × n represents the tensor product operation of the n-th dimension, Un is the left singular matrix of the n-mode expansion of the output signal tensor , where n = 1, 2, 3 are the left singular matrices after 1 / 2 / 3-mode expansion respectively

[0058] Step 6: Truncate the left singular matrix of the n-mode expansion obtained by HOSVD, and solve the noise subspace matrix for each dimension

[0059] By truncating the left singular matrices of 1-mode, 2-mode and 3-mode expansions, we can obtain That is, U nN is the noise subspace matrix for each dimension composed of the matrix column vectors corresponding to the smaller singular values remaining after the larger singular values corresponding to the number of signal sources are arranged in descending order

[0060] Step 7: Use the array steering tensor and the noise subspace matrix to perform azimuth-polarization two-dimensional spectrum search. The polarization parameters corresponding to the peak of the two-dimensional spectrum are the ellipticity angle and elliptic tilt angle of the polarization ellipse of the particle velocity at that point, and the direction of arrival (DOA) value corresponding to the peak is the azimuth angle and elevation angle of the incident signal. The azimuth of the detection target can be determined based on the polarization parameters and the direction of arrival

[0061] Since the column vectors of are orthogonal to the 1-mode noise subspace, that is the row vectors of are orthogonal to the 2-mode noise subspace. Based on the direction of arrival and polarization parameter estimation of the long-vector MUSIC algorithm, the joint estimation formula of the direction of arrival-polarization parameter MUSIC for the underwater target localization method of underwater acoustic particle velocity polarization processing is obtained as follows

[0062]

[0063] where is the polarization array manifold tensor, and × n represents the tensor product operation of the nth dimension, and U nN is the matrix related to the noise for each dimension composed of the matrix column vectors corresponding to the smaller singular values remaining after the larger singular values corresponding to the number of signal sources are arranged in descending order n is the conjugate transpose of U nN ;

[0064] The polarization parameters corresponding to the peak of the azimuth-polarization two-dimensional spectrum are the ellipticity angle and elliptic tilt angle of the polarization ellipse of the particle velocity at that point, and the direction of arrival (DOA) value corresponding to the peak is the azimuth angle and elevation angle of the incident signal

[0065] Simulation example

[0066] ​Combined with Figures 3 to 14 , the number of array elements of the vector hydrophone uniform linear array is M = 8, the element spacing is d = 0.5 m, the sound source signal is a single-frequency signal with frequency f s = 10 Hz, the sampling frequency f = 500 Hz, the number of snapshots is 1000, and the background noise is zero-mean Gaussian white noise; the incident horizontal azimuth angle of the signal is θ = 40°, and the elevation angle The elliptical inclination angle β of the particle velocity polarization ellipse is 30°, the ellipticity angle α is 10°, and the signal-to-noise ratio SNR = 0 dB.

[0067] Figures 3 to 6 It is a one-dimensional spatial spectrum comparison diagram of the estimated results of the incident signal DOA parameters and the particle velocity polarization ellipse parameters of the receiving point obtained by the underwater target localization method based on the vector hydrophone array for the polarization processing of underwater acoustic particle velocity. It can be seen from the figure that when the plane wave model estimates a real signal containing polarization, due to the influence of model mismatch, there is an estimation error in estimating the horizontal azimuth angle. And because the elevation angle and the elliptical inclination angle of polarization are coupled together, the estimation error of the plane wave model in estimating the elevation angle is relatively larger than that of the horizontal azimuth angle.

[0068] Figures 7 to 10 They are respectively the two-dimensional spatial spectra of the estimated results of the incident signal DOA parameters (traditional long vector model and tensor model) and the particle velocity polarization ellipse parameters of the receiving point (traditional long vector model and tensor model) obtained by the underwater target localization method based on the vector hydrophone array for the polarization processing of underwater acoustic particle velocity. Under the condition of the same signal-to-noise ratio, both the long vector MUSIC algorithm and the MUSIC algorithm based on tensor decomposition can accurately estimate the polarization parameter information and azimuth information of the signal. The sidelobe performance of the MUSIC algorithm based on tensor decomposition is better than that of the long vector MUSIC algorithm, and the spectral peak is sharper. When showing the same color scale, it can be intuitively seen that the MUSIC algorithm based on tensor decomposition has higher resolution, which means that it can provide higher estimation accuracy of the sound source horizontal azimuth angle / elevation angle DOA parameters and the particle velocity polarization ellipse inclination angle / ellipticity angle parameters.

[0069] Figures 11 to 14 It is a root mean square error comparison diagram of the estimated results of the incident signal DOA parameters and the particle velocity polarization ellipse parameters of the receiving point obtained by the underwater target localization method based on the vector hydrophone array for the polarization processing of underwater acoustic particle velocity. To avoid the influence of random errors, 300 Monte Carlo simulations are carried out, the signal-to-noise ratio range is set as -20 dB: 2 dB: 20 dB, and the spatial spectrum estimation step size of the DOA parameters and the polarization parameters is 0.5°.

[0070] By Figure 11 and Figure 12It can be seen that the root mean square error of the MUSIC algorithm based on the plane wave model is much larger than that of the MUSIC algorithm based on the polarization model of underwater acoustic particle velocity. This is because the non-polarization model processes the real polarized signal, and the polarization loss caused by polarization mismatch leads to model mismatch and estimation error.

[0071] From Figure 11 and Figure 12 it can be seen that the root mean square error of the MUSIC algorithm based on tensor decomposition is less than that of the long vector MUSIC algorithm. This shows that the signal subspace obtained by the MUSIC algorithm based on tensor decomposition through singular value decomposition of the information matrix in each dimension is more accurate than that of the traditional long vector MUSIC algorithm, and thus has better estimation accuracy.

[0072] In summary, the present invention relates to an underwater target localization method based on the polarization processing of underwater acoustic particle velocity using a vector hydrophone array. Many specific details have been described above for a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein. Those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific implementations disclosed above. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention by those skilled in the art within the technical scope disclosed by the present invention shall be covered by the protection scope of the present invention.

Claims

1. An underwater target positioning method based on the polarization processing of underwater acoustic particle velocity, characterized in that, It includes the following steps: Step 1: Arrange a vector hydrophone array in the form of a uniform linear array underwater to collect underwater acoustic signals of sound pressure scalars and particle velocity vectors; Step 2: Obtain the polarization vectors of each vector hydrophone and the time-delay vector of the vector hydrophone array through underwater acoustic signal acquisition; Step 3: Calculate the steering matrix of the vector hydrophone array according to the polarization vectors and the time-delay vector, and reconstruct it into a third-order tensor form, that is, obtain the steering tensor; Step 4: Obtain the third-order tensor polarization output information of the vector hydrophone array according to the steering tensor of the vector hydrophone array; Step 5: Perform high-order singular value decomposition on the third-order tensor polarization output information of the vector hydrophone array; Step 6: Truncate and solve the noise subspace matrices of each dimension for the left singular matrices of the 1-norm, 2-norm, and 3-norm obtained by high-order singular value decomposition; Step 7: Use the steering tensor and the noise subspace to perform azimuth-polarization two-dimensional spectrum search, and determine the azimuth angle and polarization parameters of the detection target according to the direction of arrival and polarization parameters corresponding to the azimuth-polarization two-dimensional spectrum peak obtained by the search.

2. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, wherein The vector hydrophone is a four-component vector hydrophone. The polarization vector of the vector hydrophone in Step 1 is specifically: where u is the polarization vector of a single vector hydrophone, 1 corresponds to the polarization component of the sound pressure scalar channel, and u x is the polarization component in the X-axis direction of the vibration velocity, and u y is the polarization component in the Y-axis direction of the vibration velocity, and u z is the polarization component in the Z-axis direction of the vibration velocity, θ ∈ [-π, π] is the horizontal azimuth angle of the vector hydrophone, is the pitch angle of the vector hydrophone, β ∈ [-π / 2, π / 2] is the elliptical tilt angle of the vector hydrophone, and α ∈ [-π / 4, π / 4] is the ellipticity angle of the vector hydrophone.

3. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, characterized in that, In Step 2, the time-delay vector of the vector hydrophone array is specifically: where a is the time-delay vector of the vector hydrophone array, λ is the wavelength of the incident signal, d is the spacing between individual vector hydrophones, and M is the number of vector hydrophones.

4. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, wherein In Step 3, the calculation formula for the steering matrix of the vector hydrophone array according to the polarization vectors and the time-delay vector is specifically: where A is the driving matrix, u1, u2, u M are the polarization vectors of the 1st, 2nd, and Mth vector hydrophones, a is the time delay vector of the vector hydrophone array, denotes the tensor product; Reconstruct the steering matrix into a steering tensor. Specifically, decompose and reconstruct the steering matrix A in different incident directions to obtain the steering tensor. The expression of the polarization array manifold tensor is: where where k is the number of incident signals received by the vector hydrophone array, is the slice of the k-th incident signal in the dimension of the number of driving tensor signals, u xk , u yk and u zk are the polarization components of the particle velocity polarization vector of the k-th incident signal in the three directions of the velocity x-axis, y-axis and z-axis, θ k is the horizontal azimuth angle of the k-th incident signal, is the elevation angle of the k-th incident signal, β k is the ellipticity angle of the particle velocity polarization ellipse of the k-th incident signal at the array element receiving point, α k is the ellipticity angle of the particle velocity polarization ellipse of the k-th incident signal at the array element receiving point, and M is the number of vector hydrophones.

5. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, wherein In Step 4, the formula for obtaining the third-order tensor polarization output information of the vector hydrophone array according to the steering tensor of the vector hydrophone array is specifically: Among them, is the output information of the third-order tensor polarization. L is the number of snapshots during the acquisition process of the vector hydrophone array. is the driving tensor, and ×3 is the 3-mode tensor product operation. is the waveform of the sound source signal. is the Gaussian noise tensor, and K is the total number of incident signals.

6. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, characterized in that In Step 5, the expression for performing high-order singular value decomposition on the third-order tensor polarization output information of the vector hydrophone array is specifically: Among them, is the third-order tensor polarization output information, is the output information tensor of the core tensor, × n represents the tensor product operation of the nth dimension, U n is the output information tensor of the left singular matrix of the n-mode expansion, n = 1, 2, 3, 7. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, wherein In Step 6, the truncation and solution of the noise subspace matrices of each dimension for the left singular matrices of the 1-norm, 2-norm, and 3-norm obtained by high-order singular value decomposition are specifically: The left singular matrices expanded for 1-mode, 2-mode, and 3-mode respectively are truncated to remove the part of the signal subspace. Among them, the last M - K columns of U1 are retained to obtain the noise subspace matrix the last 4 - K columns of U2 are retained to obtain the noise subspace matrix the last L - K columns of U3 are retained to obtain the noise subspace matrix K is the total number of incident signals.

8. The underwater target positioning method based on the polarization processing of underwater acoustic particle velocity according to claim 1, wherein, In Step 7, the determination of the azimuth angle and polarization parameters of the detection target according to the direction of arrival and polarization parameters corresponding to the azimuth-polarization two-dimensional spectrum peak obtained by the search specifically uses the direction-of-arrival-polarization parameter MUSIC joint estimation formula, and the formula is as follows: Among them, P MUSIC-T is the two-dimensional spectral function value, is the polarization array manifold tensor, is the direction of arrival, β, α are polarization parameters, and × n represents the tensor product operation of the n-th dimension, and U nN is the left singular matrix U for the n-mode expansion n corresponding to the noise subspace matrix of each dimension, n = 1, 2, 3, is the conjugate transpose of U nN .

Images