A matrix filtering based horizontal array matching positioning method

Abstract

The application discloses a horizontal array matching positioning method based on matrix filtering. A high-order parabolic equation model is adopted to calculate sound field propagation matrices at different steps, and a system propagation matrix in a sound source-receiving range is obtained through multiplication operation of the different step matrices; according to the depth of the sound source and the positions of the horizontal array elements, matrix coefficients are extracted from the system propagation matrices at different distances to form a new sparse system propagation matrix; subsequently, the sparse system propagation matrix is processed; a matrix filter set is formed; finally, the horizontal array receiving signal is processed by using the matrix filter set to form a distance-depth ambiguity diagram, and a target positioning result is obtained. The application can effectively improve the energy ratio of the main lobe and the side lobe in the depth dimension, reduce the number of high noise points in the whole two-dimensional ambiguity diagram, improve the depth and distance resolution of target positioning, and has important value for performance improvement of target positioning in an actual environment.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of underwater target passive positioning, and particularly relates to a horizontal array matching positioning method based on matrix filtering. BACKGROUND

[0002] When a water target is sailing, it radiates sound signals that can propagate over a long distance underwater, which is the basis for realizing underwater target passive positioning. By modeling underwater sound propagation, analyzing the time-space-frequency interference structure of sound signals underwater, constructing a measured-simulated data matching relationship, and forming different underwater target passive positioning methods. For example, a matching field positioning method based on sound pressure matching (Bartlett linear estimator, minimum variance distortionless response beamformer, multiple constraint processor, etc.), a matching mode positioning method based on normal mode, and a matching beam positioning method based on different beam incident energies. Due to the influence of the sea surface interface, when underwater sound propagates, a standing wave field is formed in the vertical direction, and a traveling wave field is formed in the horizontal direction. Because the interference structure of the standing wave field is complex, the vertical array passive positioning method is more likely to obtain better positioning results. However, in actual situations, in order to obtain better target direction finding results, a horizontal array is usually used as the main passive receiving array on the equipment, so it is of great value to improve the performance of the horizontal array passive positioning. The current horizontal array matching positioning method has high sidelobe of the ambiguity diagram, low depth and distance resolution, and it is difficult to obtain reliable target depth and distance results in actual situations. SUMMARY

[0003] The application provides a horizontal array matching positioning method based on matrix filtering, which can effectively improve the energy ratio of the main lobe and the sidelobe in the depth dimension, reduce the number of high noise points in the entire two-dimensional ambiguity diagram, and improve the depth and distance resolution of target positioning.

[0004] The technical solution of the application is to provide a horizontal array matching positioning method based on matrix filtering, and the steps are as follows,

[0005] Step 1: A high-order parabolic equation model is used to calculate the sound field propagation matrix at different steps;

[0006] Step 2: The system propagation matrix in the sound source-receiving range is obtained through multiplication operation of the sound field propagation matrix at different steps;

[0007] Step 3: According to the depth of the sound source and the position of the horizontal array element, the matrix coefficients are extracted from the system propagation matrix at different distances to form a new sparse system propagation matrix;

[0008] Step 4: The sparse system propagation matrix is processed, the least square estimation method is used, and a diagonal load factor is added to form a matrix filter set;

[0009] Step 5, the horizontal array received signal is processed by using matrix filter bank to form range-depth ambiguity figure, and target positioning result is obtained.

[0010] As preferred, in step 1, the parabolic equation is expressed as a multi-step matrix equation by using two-dimensional high-order Padé approximation, which is as follows,

[0011] R l U l =S l U l-1

[0012] Wherein, R l ,S l are ocean environmental parameter matrixes, and U l is different distance sound pressure vector.

[0013] As preferred, in step 2, the process of recursively solving the different depth sound field value vector at the next horizontal distance from the different depth sound field value vector at the last horizontal distance can be expressed in the following form:

[0014]

[0015] Wherein, the matrix T(r+dr) is the propagation matrix at the distance horizontal array r+dr distance, and dr is the distance grid interval.

[0016] As preferred, in step 3, in order to obtain the system propagation matrix at different distances, it is necessary to multiply the propagation matrices with different number of intervals, and the system propagation matrix at r+dr distance can be expressed as:

[0017] PM(M*dr)=T(M*dr)T((M-1)*dr)…T(m*dr)…T(2*dr)T(1*dr)

[0018] Wherein, PM is the system propagation matrix at different distances.

[0019] As preferred, in step 4, when the received array sound pressure is known, the sound source position can be estimated by using matrix filter bank, assuming that the sound source is located at the distance m*dr, the received array is located at the depth N0*dz, and the distance is (M1~M2)*dr, then it satisfies:

[0020]

[0021] Wherein, P zs ((1:N)*dz) is the initial sound field vector, N is the sound source depth grid number, and dz is the depth grid interval.

[0022] As preferred, in step 4, the least square method is used and the matrix diagonal loading factor is added, and the range-depth ambiguity estimation result is obtained.

[0023] P zs = (PM T PM+δI) -1 PM T P

[0024] The symbol (·) in the formula T represents the conjugate transpose of a complex matrix, and the diagonal loading factor δ can be adjusted according to different signal-to-noise ratios, and the lower the signal-to-noise ratio, the larger the diagonal loading factor.

[0025] Compared with the prior art, the application has the following advantages:

[0026] The horizontal array matching positioning method based on matrix filtering provided by the application can effectively improve the energy ratio of the main lobe and the sidelobe in the depth dimension, reduce the number of high noise points in the entire two-dimensional ambiguity diagram, and improve the depth and distance resolution of target positioning. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is the implementation process diagram of the application.

[0028] Figure 2 is the ocean waveguide parameter diagram.

[0029] Figure 3 is the sound source positioning result obtained by using the conventional matching processor and the method of the application.

[0030] Figure 4 is the depth-intensity result at the sound source distance calculated by using the conventional matching processor and the method of the application.

[0031] Figure 5 is the distance-intensity result at the sound source depth calculated by using the conventional matching processor and the method of the application. DETAILED DESCRIPTION

[0032] The application will be further described in the specific embodiments in combination with the drawings:

[0033] Referring to the implementation process diagram shown in Figure 1 , the horizontal array matching positioning method based on matrix filtering involved in the application adopts a high-order parabolic equation model to calculate a sound field propagation matrix; then, a system propagation matrix is obtained through multiplication operation; and matrix coefficients are extracted from the system propagation matrix at different distances according to the sound source depth and the horizontal array element position to form a new sparse system propagation matrix; subsequently, a least square estimation method is used, and a diagonal loading factor is added to form a matrix filter set; finally, the horizontal array received signal is processed by using the matrix filter set to form a distance-depth ambiguity diagram.

[0034] wherein,

[0035] The underwater acoustic field calculation formula of the parabolic equation model of the two-dimensional high-order Padé approximation can be expressed in the form of a multi-step matrix equation:

[0036] R l U l =S l U l-1

[0037] where R l , S l are matrices of marine environmental parameters, and U l is a vector of sound pressure at different distances.

[0038] The process of recursively solving the different depth sound field value vector at the next horizontal distance from the different depth sound field value vector at the previous horizontal distance can be written in the following form from the most intuitive point of view:

[0039]

[0040] where the matrix T(r+dr) is the propagation matrix at the distance level r+dr, and dr is the distance grid interval.

[0041] In order to obtain the system propagation matrix at different distances, it is necessary to include the multiplication of propagation matrices with different number intervals. Assuming that PM is the system propagation matrix at different distances, the system propagation matrix at the distance r+dr can be expressed as:

[0042] PM(M*dr)=T(M*dr)T((M-1)*dr)…T(m*dr)…T(2*dr)T(1*dr)

[0043] Since the system propagation matrix is a matrix connecting the sound source and the receiving position, when the sound pressure on the receiving array is known, the sound source position can be estimated by a matrix filter set. Assuming that the sound source is located at the distance m*dr, the receiving array is located at the depth N0*dz, and the distance is (M1~M2)*dr, then it satisfies:

[0044]

[0045] where P zs ((1:N)*dz) is the initial sound field vector, N is the number of source depth grids, and dz is the depth grid interval.

[0046] Thus, the least squares method can be used to obtain the distance-depth ambiguity estimation result by adding a matrix diagonal loading factor:

[0047] P zs =(PM T PM+δI) -1 PM T P

[0048] The symbol (·) in the above formula T δ represents the conjugate transpose of a complex matrix, and δ is the diagonal loading factor, which can be adjusted according to different signal-to-noise ratios. The lower the signal-to-noise ratio, the larger the diagonal loading factor should be.

[0049] See Figure 2 The diagram shows the ocean waveguide parameters, including ocean acoustic parameters and source information. The horizontal array is a 100-element array with an element spacing of 5m and a water depth of 40m. The source depth is 100m, the frequency is 200Hz, and the distance is 9km.

[0050] Figure 3 The results of sound source localization obtained using a conventional matched field processor and the method presented are given. It can be seen that the conventional matched field processor has higher sidelobe intensity, while the sidelobe intensity is greatly reduced in the calculation results of the method, resulting in higher resolution and lower overall energy outside the main lobe.

[0051] Further details can be found in [link to relevant document]. Figure 4 and Figure 5 , Figure 4 The depth-intensity results at the sound source distance calculated using the two methods described above are presented. It can be seen that the target depth main lobe width is narrower and higher in the results calculated by this method, and the depth resolution is higher. Figure 5 The distance-intensity results at the sound source depth obtained by the two methods mentioned above are presented. It can be seen that the target distance main lobe width is narrower and higher in the calculation results of this method, and the distance resolution is higher.

[0052] As can be seen, the horizontal matrix matching localization method based on matrix filtering of the present invention can effectively improve the energy ratio of the main lobe to the side lobe in the depth dimension, reduce the number of high noise points in the entire two-dimensional ambiguity map, and improve the depth and distance resolution of target localization. It has great innovation and is of great value for improving the performance of target localization in real-world environments.

[0053] The above description only illustrates preferred embodiments of the present invention and should not be construed as limiting the scope of the claims. Any equivalent procedural modifications made using this specification are included within the patent protection scope of this invention.

Claims

1. A matrix filter based horizontal array matching positioning method, characterized in that: The method steps are as follows, Step 1, using a high-order parabolic equation model, the sound field propagation matrix at different steps is calculated; Step 2, the system propagation matrix in the sound source-receiving range is obtained through multiplication operation of the sound field propagation matrix at different steps; Step 3, according to the sound source depth and horizontal array element position, the matrix coefficients are extracted from the system propagation matrix at different distances to form a new sparse system propagation matrix; Step 4, the sparse system propagation matrix is processed, the least square estimation method is used, and a diagonal load factor is added to form a matrix filter set; Step 5, the matrix filter set is used to process the horizontal array receiving signal to form a distance-depth ambiguity diagram and obtain the target positioning result.

2. The matrix-filter-based horizontal array matching positioning method according to claim 1, characterized in that: In step 1, the two-dimensional high-order Padé approximation parabolic equation is expressed as a multi-step matrix equation, which is specifically as follows, R l U l = S l U l-1 where l = 1, 2,..., L, R l l is a matrix of marine environmental parameters, U l is a vector of sound pressure at different distances.​ 3. The matrix-filter-based horizontal array matching localization method of claim 1, wherein: In step 2, the process of recursively solving the different depth sound field value vectors at the next horizontal distance from the different depth sound field value vectors at the previous horizontal distance is expressed in the following form: Wherein, the matrix T(r+dr) is the propagation matrix at the distance horizontal array r+dr distance, and dr is the distance grid interval.

4. The matrix-filter-based horizontal array matching localization method of claim 1, wherein: In step 3, the system propagation matrix at r+dr distance is expressed as: PM(M*dr)=T(M*dr)T((M-1)*dr)…T(m*dr)…T(2*dr)T(1*dr) Wherein, PM is the system propagation matrix at different distances.

5. The matrix-filter-based horizontal array matching localization method of claim 1, wherein: In step 4, when the sound pressure on the receiving array is known, the sound source position can be estimated through the matrix filter set. Assuming that the sound source is located at the distance m*dr, the receiving array is located at the depth N0*dz, and the distance is (M1~M2)*dr, then it satisfies: where P zs ((1 :N)*dz) is the initial sound field vector, N is the number of depth grid of sound source, and dz is the depth grid interval.

6. The matrix-filter-based horizontal array matching localization method of claim 1, wherein: In step 4, the least square method is used and a diagonal load factor of the matrix is added to obtain the distance-depth ambiguity estimation result: P zs = (PM T PM+δI) -1 PM T P The symbol (·) in the formula T δ represents the conjugate transpose of a complex matrix, and δ is the diagonal loading factor.

Images