A phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular array.

By using pre-stored phase angle offset and beam interpolation phase compensation methods, the problem of obtaining arbitrary pointing angles in frequency domain fast beamforming is solved, and the effect of rapid beamforming of arbitrary angles by uniform circular array is realized.

CN116381656BActive Publication Date: 2026-05-26THE 715TH RES INST OF CHINA SHIPBUILDING IND CORP

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
THE 715TH RES INST OF CHINA SHIPBUILDING IND CORP
Filing Date
2022-12-07
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing technologies struggle to quickly obtain beams with arbitrary pointing angles when forming uniform circular arrays using frequency-domain fast beamforming methods, and conventional beamforming methods suffer from low computational efficiency and slow computation speed.

Method used

By using pre-stored phase angle offset values ​​and beam interpolation phase compensation methods, beam data with specified and offset angles is generated using conventional beamforming. The phase angle difference is normalized, and combined with frequency domain fast beamforming and parabolic interpolation, approximate frequency domain beam data is reconstructed.

Benefits of technology

While maintaining the computational efficiency of fast beamforming in the frequency domain, it can quickly obtain beams with arbitrary pointing angles, thus improving computational speed and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116381656B_ABST
    Figure CN116381656B_ABST
Patent Text Reader

Abstract

This invention discloses a phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular ring array. The method is applied to uniform circular ring arrays and includes the following steps: First, a simulated single-frequency signal at a specified angle is generated using array-related parameters. This simulated single-frequency signal is then used for conventional beamforming to generate beam data at the specified angle and beam data with a biased angle. Second, the phase angles of the two sets of beam data are calculated separately, and the difference in phase angles is normalized by frequency and angle. The beam pointing angle is then changed to obtain an omnidirectional unit phase angle offset value. Third, conventional frequency-domain fast beamforming is performed. After calculating the energy of the frequency-domain beam data, parabolic interpolation is performed to obtain the interpolated energy spectrum. This invention overcomes the shortcomings of low computational efficiency and slow calculation speed of conventional beamforming when omnidirectional beams are required, and also solves the problem that conventional frequency-domain fast beamforming is not easy to obtain beams with arbitrary pointing angles.
Need to check novelty before this filing date? Find Prior Art

Description

Technical fields:

[0001] This invention belongs to the field of underwater acoustic array signal processing technology, specifically relating to a phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular array. Background technology:

[0002] Gaining spatial array gain through beamforming processing of pre-formed beam directions to improve detection capabilities is a fundamental technique in sonar equipment. However, when the number of pre-formed beams is large, the computational efficiency is low. For ring arrays with uniformly distributed elements, a frequency-domain fast beamforming method utilizing element-space FFT has been developed. This method can form omnidirectional beams that are integer multiples of the number of elements at once, significantly improving computational efficiency. However, the beam pointing of this method is based on the element pointing direction, and the omnidirectional beams are evenly distributed, which is not conducive to forming beams with arbitrary pointing angles. Therefore, there is an urgent need to design a method that can quickly obtain beams with arbitrary pointing angles while maintaining the computational efficiency of frequency-domain fast beamforming. Summary of the Invention:

[0003] The technical problem to be solved by this invention is to provide a phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular array. This method overcomes the shortcomings of low computational efficiency and slow computational speed of conventional beamforming when omnidirectional beams are required, and also solves the problem that traditional frequency domain fast beamforming is not easy to obtain beams with arbitrary pointing angles. While taking into account the computational efficiency of frequency domain fast beamforming, it can quickly obtain beams with arbitrary pointing angles.

[0004] The technical solution of this invention is to provide a phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular ring array. This method is applied to a uniform circular ring array, and the steps are as follows.

[0005] First, a simulated single-frequency signal at a specified angle is generated using array-related parameters. Then, conventional beamforming is performed using this simulated single-frequency signal to generate beam data at the specified angle and beam data with a biased angle.

[0006] Secondly, the phase angles of the two sets of beam data are calculated respectively, and the difference in phase angles is normalized by frequency and angle. The beam pointing angle is changed to obtain the omnidirectional unit phase angle offset value.

[0007] Next, traditional frequency domain fast beamforming is performed. After calculating the energy of the frequency domain beam data, parabolic interpolation is performed to obtain the interpolated energy spectrum.

[0008] Next, search for the beam in the fast beamforming preform that is closest to the specified beam angle, and calculate the deviation angle value between the two beams;

[0009] Finally, the approximate frequency domain beam data is reconstructed by combining the pre-stored unit phase angle offset value and the interpolated energy spectrum.

[0010] This invention is mainly applied to uniform circular arrays, where the array parameters are: array radius R, number of array elements N, array elements arranged in an even number, and the coordinates of each element are (x...). n ,y n (n = 1, 2, ..., N), the sound source is in the far field, the sound signal arrives at each array element in the form of a plane wave, and the incident direction of the signal is θ. m The array element pointing angle is According to the theory of frequency domain fast beamforming, spatial FFT can be used to quickly obtain BN omnidirectional beams, where B is a positive integer, and the pointing angle of each beam is determined by the pointing angle of the array elements. Based on the baseline, they are distributed at equal intervals, that is

[0011]

[0012] If we want to obtain M omnidirectional beams, with each beam pointing as follows:

[0013]

[0014] In most cases This achievement realizes the goal of obtaining high-performance beamforming while maintaining computational efficiency in the frequency domain. The beam that is being pointed at.

[0015] This invention mainly comprises two parts: pre-stored phase offset angle and beam interpolation phase compensation.

[0016] in,

[0017] 1. Pre-stored phase offset angle:

[0018] First, obtain the pointing angle of each beam. The signal Rs received by each array element is constructed based on the beam pointing angle. n =AS n A is the signal strength, S n =cos(2πf(t+τ) n )) represents the time-domain signal received by each array element, where t is the sampling time, and τ n =(x n cosθ m +y n sinθ m ) / c represents the reception of each array element. The incident signal has a time delay relative to the origin, where c is the speed of sound and f is the signal frequency. Under the same beam, the simulated signal is fixed, and conventional beamforming methods are used to... Angle-based beamforming processing is performed to obtain frequency domain beam data BeamRs(f). Then, a fixed offset angle Δα is selected, and conventional beamforming methods are used to... Perform beamforming processing at an angle to obtain the frequency-domain beam data BeamPy(f). Respectively obtain the phase angles of the two sets of frequency-domain beam data, that is, β Rsm (f) = tan -1 (BeamRs(f)), β Pym (f) = tan -1 (BeamPy(f)), and subtract the phase angles corresponding to the same frequency to obtain

[0019] Δβ m (f) = β Rsm (f) - β Pym (f) (3)

[0020] Perform frequency and angle normalization on Δβ m (f) to obtain the unit phase angle offset value of the current beam:

[0021]

[0022] Repeat the above operations for M beams to obtain the unit phase angle offset values β m , m = 1, 2,..., M. The above calculations can be completed in advance, without occupying the hardware resources of real-time calculations, without affecting the calculation efficiency of actual beamforming, and finally only M data are generated, and the occupation of hardware storage resources can be ignored.

[0023] 2. Beam interpolation phase compensation:

[0024] First, perform fast frequency-domain beamforming on BN beams, where BN < M and B is as large as possible to maximize the computational advantage of fast frequency-domain beamforming and minimize the beam angle error, and obtain the frequency-domain beam data B Fm (f), m = 1, 2,..., BN. Calculate the energy spectrum of each frequency point by obtaining the energy of the frequency-domain beam data:

[0025]

[0026] In the above formula represents taking the conjugate of a complex number, and P F is a two-dimensional matrix of BN * F, where F represents the number of frequency points. Perform M-point parabolic interpolation on the energy spectra of BN beams at each frequency point to obtain

[0027]

[0028] In the above formula, paraInter() represents parabolic interpolation calculation, is a two-dimensional matrix of M * F.

[0029] Secondly, based on the beam pointing angle, and combining equations (1) and (2), find the beam m' that is closest to the pointing direction of beam m, i.e.

[0030]

[0031] In the above formula, min() represents the calculation of the minimum value index. This yields the deviation angle value. The unit phase angle offset value β obtained by combining equation (4) m The beam energy value obtained by equation (6) This allows us to construct approximate frequency domain beam data for beam m:

[0032]

[0033]

[0034] Where i is the imaginary part identifier, m∈1,2,...M,m'∈1,2,...,BN.

[0035] Finally, regarding B m (f) By superimposing the energy and frequency points, the spatial energy spectrum P of the M beams can be obtained:

[0036]

[0037] Compared with the prior art, the present invention has the following advantages after adopting the above solution:

[0038] This invention overcomes the shortcomings of low computational efficiency and slow computation speed of conventional beamforming when omnidirectional beamforming is required, and solves the problem that traditional frequency domain fast beamforming is not easy to obtain beams with arbitrary pointing angles. Furthermore, this invention proposes a method of pre-storing the phase angle and then interpolating and compensating for the beam, which can quickly obtain beams with arbitrary pointing angles while taking into account the computational efficiency of frequency domain fast beamforming. Attached image description:

[0039] Figure 1 This is a flowchart illustrating the principle of pre-stored phase angle in an embodiment of the present invention;

[0040] Figure 2 This is a flowchart illustrating the phase compensation principle in an embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of the array layout in an embodiment of the present invention;

[0042] Figure 4 This is a comparison diagram of phase compensation results and CBF in an embodiment of the present invention;

[0043] Figure 5 The diagram shows the phase compensation result and CBF error in the embodiment of the present invention. Detailed implementation manners:

[0044] The present invention will be further described in detail in conjunction with the accompanying drawings in the following detailed implementation manners:

[0045] The following combines Figure 1 and Figure 2 to describe the detailed implementation manners of the present invention in detail:

[0046] A phase compensation method for quickly forming beams at any angle by a uniform circular array, which is applied to a uniform circular array, specifically includes the following steps.

[0047] Step 1: Generate a simulated single-frequency signal Rs at a specified angle according to the array parameters n ;

[0048] Step 2: Use the generated simulated signal to perform conventional beamforming to generate beam data BeamRs(f) at a specified angle and beam data BeamPy(f) at a certain deviated angle;

[0049] Step 3: Calculate the phase angles of the two groups of beam data respectively, and normalize the difference of the phase angles in terms of frequency and angle to obtain a pre-stored unit phase angle offset value β m ;

[0050] Step 4: Change the beam pointing angle, and repeat Steps 1 to 3 to obtain the pre-stored unit phase angle offset values β of all M beams;

[0051] Step 5: Perform traditional fast beamforming in the frequency domain for BN (BN < M) beams to obtain frequency-domain beam data B F (f);

[0052] Step 6: Calculate the energy of the frequency-domain beam data B F (f) to obtain the energy spectrum P F (f) of each frequency point and perform M-point parabolic interpolation to obtain the interpolated energy spectrum

[0053] Step 7: Search and determine the beam number m' closest to the specified angle in B F (f), where m' ∈ 1, 2,..., BN;

[0054] Step 8: Calculate the deviation angle value, and construct an approximate frequency-domain beam data B m in combination with the pre-stored unit phase angle offset value β and the energy spectrum m (f);

[0055] Step 9: Change the beam, and repeat Steps 6 to 8 to obtain the frequency-domain beam data B(f) of all M beams. After calculating the energy, the frequency points are superimposed to obtain the spatial energy spectrum P of the M beams.

[0056] Specifically, the following is a combination Figure 3 The embodiments of the present invention will be further described in detail below:

[0057] In step one, taking an equally spaced circular arc array with N (N≥1) elements as an example, the array radius is R, and the array is arranged with an even number of elements. The coordinates of each element are (x... n ,y n (n = 1, 2, ..., N), the sound source is in the far field, the sound signal arrives at each array element in the form of a plane wave, and the incident direction of the signal is θ. m The array element pointing angle is The pointing angle of each beam is obtained according to equation (2). The signal Rs received by each array element is constructed based on the beam pointing angle. n =AS n A is the signal strength, S n =cos(2πf(t+τ) n )) represents the time-domain signal received by each array element, where t is the sampling time, and τ n =(x n cosθ m +y n sinθ m ) / c represents the reception of each array element. The direction of the incident signal is the time delay relative to the origin of the coordinate system, where c is the speed of sound and f is the signal frequency.

[0058] In step two, using Rs n signal pair A conventional beamforming method is used to obtain frequency domain beam data BeamRs(f). Then, a fixed offset angle Δα is selected, and the conventional beamforming method is applied to... Beamforming is performed on the angle to obtain frequency domain beam data BeamPy(f);

[0059] In step three, the phase angles of the two sets of frequency domain beam data are obtained, i.e., β. Rsm (f)=tan -1 (BeamRs(f)), β Pym (f)=tan -1 (BeamPy(f)), Δβ is obtained by calculating according to equation (3). m (f) The unit phase angle offset value β is obtained by normalization calculation according to equation (4). m And store the calculation results;

[0060] In step four, the beam pointing is changed, and steps one through three are repeated to obtain the unit phase angle offset value β for all M beams. m Let m = 1, 2, ..., M and store the calculation results;

[0061] In step five, as Figure 3 shown in the formation, perform fast beamforming in the frequency domain for BN beams, where BN < M and B is as large as possible, maximizing the computational advantage of fast beamforming in the frequency domain, minimizing the beam angle error, and obtaining the frequency-domain beam data B Fm (f), m = 1, 2,..., BN;

[0062] In step six, calculate the energy spectrum P at each frequency point according to equation (5) Fm (f), m = 1, 2,..., BN, f = 1, 2,..., F, and perform parabolic interpolation on the two-dimensional energy spectrum P F in the beam dimension to obtain the estimated two-dimensional energy spectrum

[0063] In step seven, according to the pointing angle According to equation (7), find the fast beamforming preformed beam closest to and determine its beam number m', m' ∈ 1, 2,..., BN;

[0064] In step eight, calculate the angle difference from and Combined with the pre-stored unit phase angle offset value β in step four m and the estimated energy spectrum obtained in step six m and the estimated energy spectrum obtained in step six the approximate frequency-domain beam data B of the m-th beam can be constructed according to equation (8) m (f), m = 1, 2,..., M;

[0065] In step nine, change the beam pointing and repeat steps six to eight to obtain the approximate frequency-domain beam data B(f) of all M beams. After performing energy frequency point superposition according to equation (10), the spatial energy spectrum P of the M beams can be obtained. ]>

[0066] As Figure 4 shows, a comparison diagram is given between the spatial energy spectrum of conventional beamforming of a 128-element equally spaced circular array with the target at 0° and 360 omnidirectional beams, and the spatial energy spectrum of first performing fast beamforming in the frequency domain for 256 beams and then obtaining 360 omnidirectional beams through phase compensation according to this achievement. The blue solid line is the result of conventional beamforming, and the red dashed line is the result of phase compensation according to this achievement.

[0067] As Figure 5 shows, the error between conventional beamforming and the spatial energy spectrum of this embodiment is given.

[0068] Table 1 below shows the average calculation time of different methods after 1000 simulations on the same hardware device with the same parameters.

[0069] Table 1: Comparison of Results with CBF Calculation Time

[0070]

[0071] according to Figure 4 , Figure 5 As can be seen from the results in Table 1, this achievement can quickly obtain a beam with the same direction as conventional beamforming while taking into account the computational efficiency of fast beamforming in the frequency domain.

[0072] 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 structural or procedural modifications made using this specification are included within the patent protection scope of the present invention.

Claims

1. A phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular ring array, the method being applied to a uniform circular ring array, characterized in that: The steps are as follows: First, a simulated single-frequency signal at a specified angle is generated using array-related parameters. Then, conventional beamforming is performed using this simulated single-frequency signal to generate beam data at the specified angle and beam data with a biased angle. Secondly, the phase angles of the two sets of beam data are calculated respectively, and the difference in phase angles is normalized by frequency and angle. The beam pointing angle is changed to obtain the omnidirectional unit phase angle offset value. Next, fast beamforming in the frequency domain is performed. After calculating the energy of the frequency domain beam data, parabolic interpolation is performed to obtain the interpolated energy spectrum. Next, search for the beam in the fast beamforming preform that is closest to the specified beam angle, and calculate the deviation angle value between the two beams; Finally, the approximate frequency domain beam data is reconstructed by combining the pre-stored unit phase angle offset value and the interpolated energy spectrum. Among them, the simulated single-frequency signal is used. Beam pointing angle Perform conventional beamforming to obtain frequency domain beam data Then select a fixed offset angle. Using conventional beamforming methods to Beamforming processing is performed on the angle to obtain frequency domain beam data. ; Pre-stored unit phase angle offset value Obtain it in the following ways: The phase angles of the two sets of frequency domain beam data were obtained respectively. Right now , , According to the formula Calculation obtained ; The unit phase angle offset value is obtained by normalization calculation using the following formula. And store the calculation results. In the formula For frequency values, It is a fixed offset angle.

2. The phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular array according to claim 1, characterized in that: The specific steps are as follows: Step 1: Generate a simulated single-frequency signal at a specified angle based on the array parameters. ; Step two: Use the generated simulation signal to perform conventional beamforming to generate beam data at a specified angle. Beam data with bias angle ; Step 3: Calculate the phase angle for each of the two sets of beam data, and normalize the difference in phase angle for both frequency and angle to obtain the pre-stored unit phase angle offset value. ; Step four, change the beam pointing angle, and repeat steps one through three to obtain all the beams. Pre-stored unit phase angle offset value for each beam ; Step 5, do Frequency domain rapid beamforming of individual beams Obtain frequency domain beam data ; Step 6: Process the frequency domain beam data Determine the energy and obtain the energy spectrum at each frequency point. and do Point parabolic interpolation to obtain the interpolated energy spectrum ; Step 7, Search Confirmation The beam number closest to the specified angle , ; Step 8: Calculate the deviation angle value and combine it with the pre-stored unit phase angle offset value. and energy spectrum Step nine of constructing an approximate frequency domain wave: beam data. ; Change the beam and repeat steps six through eight to obtain all the beams. Frequency domain beam data of each beam The result can be obtained by superimposing the frequency points after calculating the energy. Spatial energy spectrum of each beam .

3. The phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular array according to claim 2, characterized in that: In step one, it is assumed that the equally spaced circular arc array has Each array element, The array radius is Arrange the array according to an even number of elements, with the coordinates of each element as follows: The sound source is in the far field, and the sound signal arrives at each array element in the form of a plane wave. The incident direction of the signal is... The array element pointing angle is Obtain the pointing angle of each beam. The signal received by each array element is constructed based on the beam pointing angle. , For signal strength, For each array element to receive the time-domain signal, where Sampling time, For each array element to receive The incident signal has a time delay relative to the origin. is the speed of sound, and f is the signal frequency.

4. The phase compensation method for rapidly forming beams at arbitrary angles using a uniform circular array according to claim 2, characterized in that: In step eight, approximate frequency domain beam data is constructed as follows. : According to the pointing angle and obtained from step seven Recent fast beamforming preformed beam angle Calculate the angle difference Combined with the unit phase angle offset value pre-stored in step four And the estimated energy spectrum obtained in step six Approximate frequency domain beam data is constructed based on equation (3). , In the formula This is a marker for the imaginary part. , The phase angle of the preformed frequency domain beam obtained by fast beamforming, where f is the frequency value.