A method and system for cross-enhanced DOA localization using planar arrays

Abstract

This invention relates to the field of underwater acoustic positioning, and particularly to a cross-enhanced DOA positioning method and system for planar arrays. The method includes: acquiring and processing signals received by each element of an M*N planar array to obtain N column array signal vectors and M row array signal vectors; processing the signals received by each element using signal delay characteristics, and enhancing each column array signal vector and each row array signal vector along a selected signal enhancement direction; processing each enhanced column array signal vector and each row array signal vector using a one-dimensional spatial spectrum estimation method to obtain the angle of arrival (OA) of the target sound source, and using this OA as the direction for the next signal enhancement, thereby achieving directional tracking of the target. By using one-dimensional spatial spectrum estimation technology to process the two enhancement arrays separately, the computational load and complexity are significantly reduced while maintaining accuracy.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustic positioning, and in particular to a planar array cross-enhanced DOA positioning method and system. Background Technology

[0002] Ultra-short baseline (USBR) positioning systems, with their small size and easy installation, are widely used for underwater target location. They utilize multiple receiving elements to receive signals from underwater beacons, and then use the correlation between these signals to determine the time of arrival (TOA) and direction of arrival (DOA), thereby calculating the beacon's position. Based on the receiving array configuration, full-ocean-depth USBRs (typically exceeding 11,000 m) are divided into two types. One type uses several receiving elements (usually 4-8), such as the POSIDONIA II USBR from IXSEA. However, the signal received by a single element is relatively weak, and DOA accuracy cannot be guaranteed under low signal-to-noise ratio (SNR) conditions. The other type uses multiple elements to form several virtual elements, such as the HiPAP102 from Kongsberg. It divides 31 elements into four regions and calculates DOA based on the phase difference between the four virtual elements. This method uses virtual elements to enhance the SNR, but the final direction finding still uses four elements, limiting the accuracy of DOA for long-range positioning. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of the aforementioned ultra-short baseline (UBBR) localization methods by proposing a planar array-based cross-enhanced DOA localization and tracking method and system. It utilizes the array manifold of a uniform planar array to achieve cross-enhancing in two spatial dimensions, forming two sets of enhanced arrays composed of several array elements. The enhanced array's signal has a higher signal-to-noise ratio while maintaining the original array manifold. By processing the two enhanced arrays using one-dimensional spatial spectrum estimation techniques, the angle of arrival (OA) of the target sound source can be obtained. This OA is then used as the direction for the next signal enhancement, thereby achieving directional tracking of the target.

[0004] To achieve the above objectives, the present invention is implemented through the following technical solution.

[0005] This invention proposes a cross-enhanced DOA localization method for planar arrays, the method comprising:

[0006] Step 1. Collect and process the signals received by each element of the M*N planar array to obtain the N-column array signal vector and the M-row array signal vector respectively;

[0007] Step 2. Process the signals received by each array element using the signal delay characteristics, and enhance the signals of each column array signal vector and each row array signal vector in the selected signal enhancement direction;

[0008] Step 3. Use the one-dimensional spatial spectrum estimation method to process each enhanced column array signal vector and each row array signal vector to obtain the angle of arrival of the target sound source, and use the angle of arrival as the direction of the next signal enhancement, thereby realizing the directional tracking of the target.

[0009] As an improvement to the above technical solution, step 1 includes:

[0010] Step 1-1. Acquire the signals received by each element of the M*N planar array and perform preprocessing to obtain the complex signal g of each element channel. m，n (t), and the noise signal p introduced by each array element channel. m，n (t);

[0011] Step 1-2. Based on the complex signal g of each array element channel m，n (t) and noise signal p m，n (t), respectively, obtain the matrix relationship between the N-column array signal vector and the M-row array signal vector and the spatial signal vector emitted by the signal source.

[0012] As an improvement to the above technical solution, in step 1-1, an analog-to-digital converter is used to sample the signals received by each element of the planar array; the preprocessing includes digital bandpass filtering or demodulation.

[0013] As an improvement to the above technical solution, the matrix relationship between the N-column array signal vector and the M-row array signal vector obtained in steps 1-2 and the spatial signal vector emitted by the signal source is as follows:

[0014] The nth column array signal vector With spatial signal vector The matrix relationship between them is as follows:

[0015]

[0016] Where n = 1, 2, 3, ..., N; This represents the noise vector in the nth column. Space signal vector L represents the number of spatial signal sources, and the superscript T represents the matrix transpose; A n The M*L dimensional array manifold matrix is ​​represented as: The x-direction steering vector is used to determine the angle between the signal emitted by the l-th signal source and the x-axis direction, and is expressed as:

[0017]

[0018] Where l = 1, 2, ..., N;

[0019] The array signal vector in the m-th row With spatial signal vector The matrix relationship between them is as follows:

[0020]

[0021] Where m = 1, 2, 3, ..., M; Let m represent the noise vector in the m-th row. B m The N*L dimensional array manifold matrix is ​​represented as: The y-direction steering vector is used to determine the angle between the signal emitted by the l-th signal source and the y-axis direction, and is expressed as:

[0022]

[0023] Where, τ m，n，l The theoretical delay of the received signal in the m-th row and n-th column of the array element in the l-th direction of arrival is expressed as:

[0024] τ m，n，l =(m-1)*τ x，l +(n-1)*τ y，l

[0025] Where, τ x，l and τ y，l The unit delays in the x and y directions are respectively represented by the following formulas:

[0026]

[0027] τ y，l =d y cosα l sinβ l cos P(t)cos R(t) / c

[0028] Where c is the speed of sound, d x and d y These are the element spacings in the x and y directions, respectively; P(t) and R(t) represent pitch, roll, and roll respectively; the x and y directions are the row and column directions of the array, respectively.

[0029] As an improvement to the above technical solution, step 2 specifically includes:

[0030] Step 2-1. Utilize the delay characteristics of the signals of each element of the planar array to phase-shift or delay the signals and sum them to achieve coherent superposition of the signals. Map the planar array into a linear array in the x-direction and a linear array in the y-direction. Depending on the reference point, the two linear arrays are in L-shape or cross shape.

[0031] Step 2-2. Select the signal enhancement direction and define the unit vector D′ of the signal enhancement direction. l The corresponding pitch angle is α′ l The azimuth angle is β′ l Thus, the rotational unit delay in the x and y directions of the enhancement direction is obtained, and the signal enhancement matrices in the x and y directions are further defined respectively;

[0032] Steps 2-3: Using the signal enhancement matrix to phase shift or delay the array signals, coherent superposition of all signals in the x and y directions of the enhancement direction is achieved, completing the signal enhancement in the x and y directions, and obtaining the enhanced signal vector of each column array and each row array.

[0033] As an improvement to the above technical solution, step 2-2, selecting the signal enhancement direction, includes:

[0034] For applications with a single incoming wave direction, the previously located incoming wave direction is used as the signal enhancement direction to achieve directional feedback; the application with a single incoming wave direction includes localization sonar.

[0035] For applications with multiple incoming wave directions, the signal enhancement direction of a certain signal source in this positioning is set to the incoming wave direction of the previous positioning of that signal source, including depth sounding sonar, side-scan sonar and multibeam sonar.

[0036] As an improvement to the above technical solution, in steps 2-3, the enhanced signal vectors for each column array and each row array are obtained, wherein the enhanced signal vector of the nth column array in the x-direction is represented as:

[0037]

[0038] The array signal vector of the m-th row after signal enhancement in the y-direction is represented as:

[0039]

[0040] Among them, W x，l W y，l The signal enhancement matrices for the defined x and y directions are expressed as follows:

[0041]

[0042]

[0043] Where, τ′ x，l ,τ′ y，l Let x and y be the rotational unit delays for the l-th enhancement direction, respectively, and their expressions are as follows:

[0044]

[0045] τ′ y，l =d y cosα′ l sinβ′ l cos P(t)cos R(t) / c

[0046] Where, α′ l ,β′ l The unit vector D′ of the l-th signal enhancement direction is respectively l The corresponding pitch angle and azimuth angle.

[0047] As one of the improvements to the above technical solution, the one-dimensional spatial spectrum estimation method in step 3 includes the MUSIC algorithm and the ESPRIT algorithm;

[0048] When the spatial spectrum estimation method is the MUSIC algorithm, step 3 specifically includes:

[0049] Step 3-1. Calculate the covariance matrix of the enhanced array signal vector in the nth column of the x-direction and the enhanced array signal vector in the mth row of the y-direction respectively; perform eigenvalue decomposition on the covariance matrix and then use the orthogonality between the steering vector of the signal subspace and the noise subspace to obtain the power spectrum in the x-direction and y-direction respectively, and perform spectral peak search to obtain the angles corresponding to the maximum values, which are the x-direction angle and y-direction angle of the incoming wave respectively.

[0050] Step 3-2. Use the x-direction and y-direction angles of the future wave as the directions for the next signal enhancement, thereby achieving directional tracking of the target.

[0051] As one of the improvements to the above technical solution, the covariance matrix R of the enhanced x-direction nth column array signal vector is... x The formula for calculation is:

[0052]

[0053] Among them, R sx It is the signal covariance matrix, σ 2 Let I be the noise power, I be the identity matrix, the superscript T represents the matrix conjugate transpose, and E[ ] be the mathematical expectation;

[0054] For the covariance matrix R x After eigenvalue decomposition, we get:

[0055]

[0056] Among them, U S It is the signal subspace spanned by the eigenvectors corresponding to the large eigenvalues, and its corresponding eigenvalue diagonal matrix is ​​∑ s ;U N It is a noise subspace spanned by the eigenvectors corresponding to the small eigenvalues, and its corresponding eigenvalue diagonal matrix is ​​∑ N ;

[0057] Power spectrum P in the x-direction MU The expression is:

[0058]

[0059] in, For U S The guide vector.

[0060] This invention also proposes a planar array-based cross-enhanced DOA positioning system, the system comprising: an M*N planar array and a data processing module, wherein,

[0061] The M*N planar array includes M*N array elements for receiving signals;

[0062] The data processing module is used to acquire and process the signals received by each element of the M*N planar array to obtain the N-column array signal vector and the M-row array signal vector respectively; it is used to process the signals received by each element using the signal delay characteristics, and to enhance the signal of each column array signal vector and each row array signal vector in the selected signal enhancement direction; it is also used to process the enhanced column array signal vector and each row array signal vector respectively using the spatial spectrum estimation method to obtain the angle of arrival of the target sound source, and to use the angle of arrival as the direction of the next signal enhancement, thereby realizing the directional tracking of the target.

[0063] The advantages of this invention compared to the prior art are:

[0064] 1. Compared with the traditional two-dimensional spatial spectrum estimation technique for planar arrays, the cross-enhanced array signal has a higher signal-to-noise ratio and maintains the original array manifold, ensuring the accuracy of angle estimation. By using one-dimensional spatial spectrum estimation technique to process the two enhanced arrays separately, the computational load and complexity are greatly reduced while ensuring accuracy.

[0065] 2. The method of the present invention can also be used in sonar systems such as image sonar systems, multibeam sonar systems and other positioning sonar systems. The algorithm has high accuracy and is stable and reliable.

[0066] 3. The method of the present invention is applicable to all planar arrays and does not limit the number of array elements of the planar array, thus ensuring the accuracy of long-distance positioning of DOA. Attached Figure Description

[0067] Figure 1 This is a flowchart of the method of the present invention;

[0068] Figure 2 This is the effect of the preset angle on the angle estimation when the elevation angle of the incoming wave is 90 degrees.

[0069] Figure 3 This is the effect of the preset angle on the angle estimation when the elevation angle of the incoming wave is 65 degrees.

[0070] Figure 4 This is the effect of the preset angle on the angle estimation when the elevation angle of the incoming wave is 45 degrees.

[0071] Figure 5 This is a comparison chart of the angle estimation errors between the cross-enhanced DOA method and the two-dimensional MUSIC method. Detailed Implementation

[0072] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0073] Example 1

[0074] (I) This invention proposes a cross-enhanced DOA localization method for planar arrays, the basic operation method of which is as follows:

[0075] 1. Collect the three-dimensional attitude data of the carrier through attitude sensors, stabilize the attitude data in real time, and calculate the delay corresponding to each array element.

[0076] 2. The acquired array signals are preprocessed and represented in two orthogonal vector forms.

[0077] 3. The orthogonal vectors obtained in step 2 are cross-enhanced and mapped into an L-shaped or cross-shaped vertical linear array, while the spatial array guide vector remains unchanged in both directions.

[0078] 4. Perform one-dimensional spatial spectrum estimation on the two linear arrays obtained in step 3, and estimate the arrival angles in the two directions.

[0079] 5. Use the incoming wave angle obtained in step 4 as the preset angle input for the next cross-enhancement, and repeat the above steps to achieve target tracking.

[0080] The specific steps of the method are as follows:

[0081] 1. Attitude Acquisition and Stabilization

[0082] For an M*N planar array, where m = 1, 2, 3, ..., M, and n = 1, 2, 3, ..., N, the rotation matrix is ​​calculated as follows when the carrier undergoes attitude changes:

[0083]

[0084]

[0085]

[0086]

[0087] Γ P (t), Γ R (t) represents the bow roll. The rotation matrices in the three directions of pitch P(t), roll R(t), Γ(t) is the composite rotation matrix, and t is any sampling time.

[0088] First, the coordinates of each array element in the local level coordinate system are obtained by synthesizing the rotation matrix using the following formula.

[0089]

[0090] Wherein, the three-dimensional coordinate vector in the local leveling coordinate system at the m-th row and n-th column is U′ m，n =[x′ m，n y′ m，n z′ m，n ] T The three-dimensional coordinate vector in the carrier coordinate system at row m and column n is U. m，n =[x m，n y m，n z m，n ] T The subscript T denotes the transpose of the vector.

[0091] Then calculate vector U′ m，n and the unit vector V of the incoming wave direction l The included angle θ m，n，l (l = 1, 2, 3, ..., L), denoted as

[0092]

[0093] The |·| operation represents the modulus of a vector. The pitch angle is α. l The azimuth angle is β l The unit vector D of the incoming wave direction l It can be represented as

[0094] D l =[cosα l cosβl cosα l sinβ l sinα l ] T

[0095] Finally, the theoretical delay of each array element in the l-th incoming wave direction is calculated using the angle between the two vectors:

[0096]

[0097] Where c is the speed of sound. If the array is a uniform M*N planar array, the above theoretical delay can be rewritten as follows:

[0098]

[0099] Where d x and d y These represent the element spacing in the x and y directions, respectively.

[0100] Define rotation unit delay and τ y，l =d y cosα l sinβ l cos P(t)cos R(t) / c, then the above equation can be rewritten as

[0101] τ m，n，l =(m-1)*τ x，l +(n-1)*τ y，l (1)

[0102] 2. Signal Acquisition and Preprocessing

[0103] Sampling typically requires an analog-to-digital converter and preprocessing operations such as digital bandpass filtering or demodulation. The resulting complex signal for each array element channel is g. m，n (t), the noise introduced by the channel is p m，n (t) represents independent noise signals that are uncorrelated with each other. If there are L independent signal sources s... l (t), then the signal vector of the nth column array can be expressed as

[0104]

[0105] Wherein, signal vector noise vector Space signal vector The superscript T represents the matrix transpose. Therefore, an M*L dimensional array manifold matrix can be represented as...

[0106]

[0107] A n It can be further expressed as in The x-direction steering vector is used to determine the angle between the signal and the x-axis direction; it can be expressed as...

[0108]

[0109] Substituting equation (1) into the equation, we get

[0110]

[0111] Similarly, the array signal vector for the m-th row can be represented as:

[0112]

[0113] Wherein, signal vector noise vector Then the N*L dimensional array manifold matrix can be represented as

[0114]

[0115] B n It can be further expressed as in The y-direction steering vector is used to determine the angle between the signal and the y-axis direction, and it can be expressed as:

[0116]

[0117] Substituting equation (1) into the equation, we get

[0118]

[0119] The above M*N dimensional array signal can be represented as and Two orthogonal vector forms.

[0120] 3. Array Cross-Enhancement

[0121] By utilizing the delay characteristics of the column array signals and row array signals mentioned above, phase shifting or delaying the signals and summing them achieves coherent superposition. This operation can map the planar array into X-direction linear arrays and Y-direction linear arrays. Depending on the reference point, the two linear arrays can be L-shaped or cross-shaped. After cross-enhancement, the signal-to-noise ratio is enhanced, and the spatial array steering vector remains unchanged in both linear arrays.

[0122] For applications with a single incoming wave direction, such as localization sonar, the signal enhancement direction can be selected as the incoming wave direction of the previous localization, achieving directional feedback. For applications with multiple incoming wave directions, such as depth sounding sonar, side-scan sonar, and multibeam sonar, multiple signal enhancement calculations can be performed to enhance the signal in multiple directions separately. The unit vector D′ for the l-th signal enhancement direction is now defined. l The corresponding pitch angle is α′ l The azimuth angle is β′ l Similarly, we can obtain and τ′ y，l =d y cosα′ l sinβ′ l cos P(t)cos R(t) / c, where τ′ x，l and τ′ y，l These are the rotation unit delays in the x and y directions of the l-th enhancement direction, respectively.

[0123] Signal enhancement in the x-direction is achieved by coherently superimposing all array signals in that direction through phase shifting or delay. The signal enhancement matrix W is defined. x，l for

[0124]

[0125] According to equation (2), the array signal vector of the nth column after signal enhancement in the x-direction can be expressed as:

[0126]

[0127] make

[0128] A′ n =W x，l A n

[0129] A′ n The first act

[0130]

[0131] When the pitch angle is α′ l =α l The azimuth angle is β′ l =β′ l At that time, τ′ x，l =τ x，l , Can be rewritten as

[0132]

[0133] Comparing equation (2) and equation (3), it can be seen that the signal enhancement operation increases the signal amplitude but does not change the phase difference between the elements of the steering vector. Meanwhile, the independent and uncorrelated noise vectors... After signal enhancement, only incoherent superposition occurs. Therefore, the signal from the direction of arrival undergoes amplitude superposition, and the noise undergoes energy superposition. Thus, the signal enhancement gain is 10log. 10 M.

[0134] Similar to the above operations, signal enhancement can be performed in the y-direction by phase shifting or delaying the array signals to achieve coherent superposition of all array signals in that direction. Define the signal enhancement matrix R. y，l for

[0135]

[0136] According to equation (2), the array signal vector of the m-th column after signal enhancement in the v direction can be expressed as:

[0137]

[0138] When the pitch angle is α′ l =α l The azimuth angle is β′ l =β′ l At that time, τ′ y，l =τ y，l ,make

[0139] B′ m =W y，l B m

[0140] B′ n The first act

[0141]

[0142] Comparing equation (4) with equation (5), it can be seen that the signal enhancement operation enhances the signal amplitude without changing the relative phase between the elements of the steering vector. Meanwhile, independent and uncorrelated noise vectors... After signal enhancement, only incoherent superposition occurs. Therefore, the signal from the direction of arrival undergoes amplitude superposition, and the noise undergoes energy superposition. Thus, the signal enhancement gain is 10log. 10 N.

[0143] In summary, the cross-enhanced array signal improves the signal-to-noise ratio in a predetermined direction while maintaining the array manifold and steering vector unchanged, thus enabling [the processing of] [the signal-to-noise ratio]. and Spatial spectrum estimation is performed to obtain the angle between the incoming wave signal and the Y-axis direction and the angle between the incoming wave signal and the X-axis direction.

[0144] 4. Spatial Spectrum Estimation

[0145] Spatial spectrum estimation techniques utilize the orthogonality between the noise subspace and the signal subspace to estimate signal parameters and determine the direction of arrival. Commonly used algorithms are the MUSIC algorithm and the ESPRIT algorithm; here, the MUSIC method is used for spatial spectrum estimation.

[0146] The covariance matrix of the cross-enhanced y-direction array signal can be expressed as:

[0147]

[0148] Where R sx It is the signal covariance matrix, σ 2 Let I be the noise power, I be the identity matrix, the superscript T represent the matrix conjugate transpose, and E[ ] be the expected value. For R x Eigenvalue decomposition yields

[0149]

[0150] U S It is the signal subspace spanned by the eigenvectors corresponding to the large eigenvalues, and its corresponding eigenvalue diagonal matrix is ​​∑ S U N It is a noise subspace spanned by the eigenvectors corresponding to the small eigenvalues, and its corresponding eigenvalue diagonal matrix is ​​∑ N Then, by utilizing the orthogonality between the steering vector of the signal subspace and the noise subspace, the power spectrum can be obtained.

[0151]

[0152] For P MU Perform a peak search to obtain the angle γ corresponding to the maximum value. x , which is the angle in the x-direction of the incoming wave, where For U S The guide vector.

[0153] Change the input of the above calculation process to By repeating the above process, the y-direction angle γ of the incoming wave can be estimated. y .

[0154] (II) Estimation and Comparison of Computational Amount

[0155] The two-dimensional MUSIC method is a typical and commonly used method for traditional two-dimensional spatial spectrum estimation, characterized by high accuracy and good robustness. Table 1 compares the computational costs of the cross-enhanced DOA method, the two-dimensional MUSIC method, and the two-dimensional beamforming method.

[0156] Table 1 Comparison of computational complexity for the three calculation methods

[0157]

[0158]

[0159] It can be seen that when M and N are large, the cross-enhanced DOA method has fewer eigenvalue decompositions compared to the two-dimensional MUSIC method, and the computational cost of two one-dimensional peak searches is greatly reduced compared to one two-dimensional peak search. Therefore, the cross-enhanced DOA method has lower computational complexity.

[0160] (III) Simulation Results

[0161] Cross-enhancement processing requires a preset pitch angle α′ l and azimuth angle β′ l Typically, we need to preset the angle based on the previous processing result. Therefore, the preset angle will affect the angle error in another direction. The simulation parameters and results of this effect are as follows:

[0162] (1) Monte Carlo simulation times: 100 times / preset angle

[0163] (2) Number of planar array elements: 6*6

[0164] (3) Element spacing: half wavelength

[0165] (4) Preset angle input spacing: 2 degrees

[0166] from Figures 2-4 It can be concluded that when the signal-to-noise ratio exceeds 10dB, the preset angle has little impact on the angle estimation. Under low signal-to-noise ratio conditions, certain preset angles have a significant impact on the angle estimation, and these angles differ greatly from the actual angle of arrival. Generally, if the angle estimation result obtained in the previous step is used as feedback input, it will not have a significant impact on the angle estimation error in the next step.

[0167] The two-dimensional MUSIC method is a general method for estimating the direction of arrival (DOA) of a planar array, and it has a smaller angle estimation error compared to other DOA methods. The simulation results comparing the cross-enhanced DOA method proposed in this patent with the two-dimensional MUSIC method are as follows:

[0168] (1) Number of Monte Carlo simulations: 1000 per method

[0169] (2) Target direction: randomly selected in half space

[0170] (3) Number of planar array elements: 6*6

[0171] (4) Element spacing: half wavelength

[0172] from Figure 5As can be seen, when the signal-to-noise ratio exceeds 10 dB, both the cross-enhanced DOA method and the 2D MUSIC method have small angle estimation errors. However, under low signal-to-noise ratio conditions, the cross-enhanced DOA method degrades slightly.

[0173] In summary, the planar array cross-enhanced DOA localization and tracking method achieves similar accuracy to the traditional planar array two-dimensional spatial spectrum estimation method under high signal-to-noise ratio (SNR) conditions. In low SNR applications, accuracy may degrade when the direction of arrival of certain targets deviates significantly from the preset angle; however, high-precision direction finding can still be achieved by using the correct cross-enhanced preset angle for tracking.

[0174] Example 2

[0175] Embodiment 2 of the present invention designs a planar array-based cross-augmented DOA positioning system, the system comprising: a planar array and a data processing module, wherein,

[0176] The planar array includes multiple array elements for receiving signals;

[0177] The data processing module is used to acquire and process the signals received by each element of the M*N planar array to obtain the N-column array signal vector and the M-row array signal vector respectively; it is used to process the signals received by each element using the signal delay characteristics, and to enhance the signal of each column array signal vector and each row array signal vector in the selected signal enhancement direction; it is also used to process the enhanced column array signal vector and each row array signal vector respectively using the spatial spectrum estimation method to obtain the angle of arrival of the target sound source, and to use the angle of arrival as the direction of the next signal enhancement, thereby realizing the directional tracking of the target.

[0178] This invention proposes a cross-enhanced DOA localization and tracking method for planar arrays. It achieves cross-enhancing in two spatial dimensions to improve the signal-to-noise ratio. Furthermore, it utilizes one-dimensional spatial spectrum estimation techniques to estimate the DOA for both enhanced arrays, ensuring accuracy while significantly reducing computational load and complexity.

[0179] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A cross-enhanced DOA positioning method of planar array, the method comprising: Step 1. Collecting and processing signals received by each element of the M*N planar array to obtain cross N-column array signal vectors and M-row array signal vectors respectively; Step 2. Processing signals received by each element using signal delay characteristics, and performing signal enhancement on each column array signal vector and each row array signal vector in the selected signal enhancement direction; Step 3. Processing each column array signal vector and each row array signal vector after enhancement using a one-dimensional spatial spectrum estimation method to obtain the angle of arrival of the target sound source, and taking the angle of arrival as the direction of next signal enhancement to realize directional tracking of the target; wherein the step 2 specifically comprises: Step 2-1. Using the delay characteristics of signals received by each element of the planar array to perform phase shift or delay and summation to realize coherent superposition of signals, mapping the planar array into x-direction linear array and y-direction linear array, and according to the reference point, the two direction linear arrays are in L shape or cross shape; Step 2-2. Select the signal enhancement direction and define the unit vector D' of the signal enhancement direction l , and the corresponding pitch angle is α' l , and the azimuth angle is β' l , so as to obtain the rotation unit delay of the x direction and the y direction of the enhancement direction, and further define the signal enhancement matrix of the x direction and the y direction respectively; Step 2-3. Using signal enhancement matrix to realize coherent superposition of all signals in the x-direction and y-direction of the enhancement direction by phase shift or delay of array signals, completing signal enhancement in the x-direction and y-direction, and obtaining each column array signal vector and each row array signal vector after enhancement.

2. The cross-enhanced DOA positioning method of planar array according to claim 1, characterized in that, The step 1 comprises: Step 1-1. Collect and pre-process the signal received by each array element of the M*N planar array to obtain the complex signal g of each array element channel m,n (t), and the noise signal p introduced by each array element channel m,n Step 1-1. Collect and pre-process the signal received by each array element of the M*N planar array to obtain the complex signal g of each array element channel m,n (t), and the noise signal p introduced by each array element channel m,n Step 1-1. Collect and pre-process the signal received by each array element of the M*N planar array to obtain the complex signal g of each Step 1-2. Obtain the complex signal g of each array element channel m,n (t) and the noise signal p m,n (t), respectively, to obtain the matrix relationship between the N-column array signal vector and the M-row array signal vector of the cross and the spatial signal vector emitted by the signal source.

3. The cross-enhanced DOA positioning method of a planar array according to claim 2, characterized in that, In the step 1-1, an analog-to-digital converter is used to sample signals received by each element of the planar array; and the preprocessing comprises digital band-pass filtering or demodulation.

4. The cross-enhanced DOA positioning method of planar array according to claim 2, characterized in that, The matrix relationship between the N-column array signal vectors and M-row array signal vectors obtained in the step 1-2 and the spatial signal vector emitted by the signal source is as follows: the nth column array signal vector the matrix relationship between the spatial signal vector is where n = 1, 2, 3, …, N; represents the nth column noise vector, spatial signal vector L represents the number of spatial signal sources, and the upper index T represents matrix transposition; A n represents an M*L-dimensional array flow matrix, and is represented as: is an x-direction steering vector, and the x-direction steering vector is used to determine the included angle between the signal emitted by the lth signal source and the x-axis direction, and is represented as: wherein, l=1, 2, …, N; The mth row array signal vector The matrix relationship between the spatial signal vector is: where m = 1, 2, 3, …, M; represents the mth row noise vector, B m represents an N*L dimensional array flow pattern matrix, which is represented as: is a y-direction steering vector, the y-direction steering vector is used to determine an included angle between a signal emitted by the lth signal source and a y-axis direction, and is represented as: where τ m,n,l represents the theoretical delay of the received signal of the element in the mth row and the nth column in the lth DOA, and the expression is as follows: τ m,n,l = (m-1)*τ x,l +(n-1)*τ y,l Where, τ x,l and τ y,l The unit delays in the x and y directions are respectively represented by the following formulas: τ y,l = d y cos α l sin β l cos P(t) cos R(t) / c where c is the sound speed, d x and d y are the inter-element distances in the x and y directions, respectively; P(t), R(t) represent the yaw, pitch, roll, respectively; x and y directions are the row and column directions of the array, respectively.

5. The cross-enhanced DOA positioning method of planar array according to claim 1, characterized in that, In the step 2-2, the signal enhancement direction is selected, including: For single wave direction application, taking the wave direction positioned as the signal enhancement direction to realize directional feedback; the single wave direction application includes positioning sonar; For multiple wave direction application, the signal enhancement direction of a signal source positioned this time is set as the wave direction of the signal source positioned last time; the multiple wave direction application includes depth sounding sonar, side scan sonar and multi-beam sonar.

6. The cross-enhanced DOA positioning method of planar array according to claim 1, characterized in that, In the step 2-3, each column array signal vector after x-direction signal enhancement and each row array signal vector after y-direction signal enhancement are obtained, wherein the n-th column array signal vector after x-direction signal enhancement is represented as: the m-th row array signal vector after y-direction signal enhancement is represented as: where Wxand Wyare defined x-direction and y-direction signal enhancement matrices, respectively, expressed as: x,l y,l where Wxand Wyare defined x-direction and y-direction signal enhancement matrices, respectively, expressed as:​ where τ'xand τ'yare the unit delays in the x and y directions of the lth enhancement direction, respectively, and are expressed as: x,l where τ'xand τ'yare the unit delays in the x and y directions of the lth enhancement direction, respectively, and are expressed as: y,l where τ'xand τ'yare the unit delays in the x and y directions of the lth enhancement τ' y,l = d y cos α' l sin β' l cos P(t) cos R(t) / c wherein a′ l , b′ l are the corresponding pitch and azimuth angles, respectively, of the ith signal enhancement direction unit vector D′ l .

7. The cross-enhanced DOA positioning method of a planar array according to claim 6, characterized in that, The one-dimensional spatial spectrum estimation method in the step 3 comprises MUSIC algorithm and ESPRIT algorithm; When the spatial spectrum estimation method is MUSIC algorithm, the step 3 specifically comprises: Step 3-1. Calculating the covariance matrix of the n-th column array signal vector after x-direction signal enhancement and the m-th row array signal vector after y-direction signal enhancement respectively; after eigenvalue decomposition of the covariance matrix, the orthogonality of the signal subspace and the noise subspace is used to obtain the power spectrum of the x-direction and the y-direction, and the spectrum peak search is performed respectively to obtain the angle corresponding to the maximum value, which is the x-direction angle and the y-direction angle of the wave respectively. Step 3-2. The x-direction angle and the y-direction angle of the incoming wave are taken as the direction of next signal enhancement, so as to realize directional tracking of the target.

8. The cross-enhanced DOA positioning method of a planar array according to claim 7, characterized in that, the enhanced x-direction n-th column array signal vector x The calculation formula is: where R sx is the signal covariance matrix, σ 2 is the noise power, I is the identity matrix, the upper index T represents the matrix conjugate transpose, and E[] is the mathematical expectation. The covariance matrix R x is decomposed into eigenvalues where U S is the signal subspace spanned by the eigenvectors corresponding to the large eigenvalues, whose diagonal matrix is S ; U N is the noise subspace spanned by the eigenvectors corresponding to the small eigenvalues, whose diagonal matrix is N ; The power spectrum P in the x direction MU The expression is: wherein is the unit vector of U S is the unit vector of U 9. A cross-enhanced DOA positioning system of a planar array, characterized in that, The system comprises an M*N planar array and a data processing module, wherein, The M*N planar array comprises M*N elements for receiving signals; The data processing module is configured to collect and process the signals received by each element of the M*N planar array to obtain cross N-column array signal vectors and M-row array signal vectors respectively, to process the signals received by each element using signal delay characteristics, to perform signal enhancement on each column array signal vector and each row array signal vector in a selected signal enhancement direction, and to process each column array signal vector and each row array signal vector after enhancement using a spatial spectrum estimation method to obtain the angle of arrival of the target sound source, and to take the angle of arrival as the direction of next signal enhancement, so as to realize directional tracking of the target. The processing of the signals received by each element using signal delay characteristics and the signal enhancement on each column array signal vector and each row array signal vector in a selected signal enhancement direction specifically comprise: The delay characteristics of the signals of each element of the planar array are used to perform phase shift or delay and summation on the signals to realize coherent superposition of the signals, to map the planar array into an x-direction linear array and a y-direction linear array, and to make the two direction linear arrays into L shape or cross shape according to different reference points; A signal enhancement direction is selected and a unit vector D' of the signal enhancement direction is defined l with a corresponding pitch angle of a' l and an azimuth angle of β' l so as to obtain a rotation unit delay of the x direction and the y direction of the enhancement direction, and further define a signal enhancement matrix of the x direction and the y direction respectively; The phase shift or delay of the array signals using the signal enhancement matrix realizes coherent superposition of all signals in the x-direction and the y-direction of the enhancement direction, completes signal enhancement in the x-direction and the y-direction, and obtains each column array signal vector and each row array signal vector after enhancement.

Images