Method for detecting enhanced target acoustic scattering properties, storage medium, and sonar

Abstract

This invention discloses a detection method, storage medium, and sonar that enhances the acoustic scattering characteristics of a target. After parameter initialization, the energy value of the matched filter beamforming is estimated, the ratio of the matched filter beamforming energy value to the actual matched filter beamforming energy value is calculated, and the value is transformed to the frequency-wavenumber domain to obtain Q. (it) (f,k u ), and then according to Q (it) (f,k u Calculate the update rate Δs of the signal energy distribution function. (it) (t,sinθ), based on the update rate Δs of the signal energy distribution function (it) The signal energy distribution function (t, sinθ) is calculated after one update and iterated until convergence to obtain high-resolution imaging in the range and azimuth directions. This method can effectively improve the signal-to-mixing ratio of target echoes, reduce the false alarm probability, and enhance the detection capability of submerged and buried targets.

Description

Technical Field

[0001] This invention relates to the field of underwater target detection technology, specifically to a detection method, storage medium, and sonar that enhances the acoustic scattering characteristics of targets, thereby improving the detection capability of submerged and buried targets. Background Technology

[0002] The acoustic scattering characteristics of submerged and buried targets vary with their location and the incident angle of the sound waves, exhibiting spatial inhomogeneity. Acoustic scattering from the sediment layer and the target are coupled in both the spatial and temporal domains. Furthermore, clutter caused by strong scatterers in the sediment layer further complicates the detection of columnar targets. Therefore, improving the spatiotemporal resolution of the target and suppressing clutter are effective means to enhance the ability of active sonar to detect submerged and buried columnar targets. Traditional methods, such as low-frequency synthetic aperture sonar and adaptive spatiotemporal high-resolution processing, suffer from problems such as high requirements for the sonar platform (the platform needs to perform uniform linear motion) and insufficient robustness.

[0003] See [1] Yang T C. Deconvolved conventional beamforming for ahorizontal line array[J]. IEEE Journal of Oceanic Engineering, 2017, 43(1):160-172. [2] Yang T C. Performance analysis of superdirectivity of circular arrays and implications for sonar systems[J]. IEEE journal of oceanicengineering, 2018, 44(1): 156-166. TC Yang applied a deconvolution algorithm for image deblurring to conventional beamforming of uniform line arrays and circular arrays. Compared with the traditional MVDR beamforming method, the beam is narrower, the sidelobe level is lower, the output signal-to-noise ratio is higher, and the robustness is better.

[0004] However, with the improvement of active sonar resolution, the amount of clutter caused by strong scattering objects such as reefs, rocks, and artificial facilities in complex waters such as ports and lakes has also increased. Clutter, due to its strong correlation with the transmitted waveform, leads to the appearance of false targets. To address the clutter suppression problem, sonar waveform design and filtering, as well as spatiotemporal adaptive processing, can reduce the impact of clutter on echo detection to some extent. However, sonar waveform design and filtering are more effective at suppressing clutter from moving targets, but less effective for stationary targets. The difficulty in obtaining clutter samples and estimating their covariance matrix results in insufficient clutter suppression capability of spatiotemporal adaptive processing methods. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention improves the detection capability of submerged and buried targets by calculating the target's acoustic scattering coefficient and designing a high-resolution spatiotemporal detection method based on the target's acoustic scattering characteristics.

[0006] According to a first aspect, the present invention provides a detection method for enhancing the acoustic scattering characteristics of a target, comprising the following steps:

[0007] Step S1: Calculate the acoustic scattering coefficient of the target. Initialize the signal energy distribution function ;

[0008] According to the acoustic scattering coefficient Calculate the two-dimensional point spread function of matched filtering-beamforming ;

[0009] Step S2: Initialize the signal energy distribution function and 2D point extension function After two-dimensional FFT transformation to the frequency-wavenumber domain, we obtain and ,according to and Calculate the energy value of matched filter beamforming ; For time delay, For wave number.

[0010] Step S3: Calculate the matched filter beamforming energy value The ratio of the actual matched filter beamforming energy value to the actual energy value is transformed to the frequency-wavenumber domain to obtain... ;

[0011] Step S4: According to Calculate the update rate of the signal energy distribution function ;

[0012] Step S5: Update the signal energy distribution function according to its rate of change. Calculate the signal energy distribution function after one update. ;

[0013] Step S6: Perform an iterative convergence check. If convergence fails, return to step S2 and start the next iteration. Otherwise, use the RL algorithm to update the signal energy distribution function. Two-dimensional deconvolution processing is performed to obtain high-resolution images in the range and azimuth directions;

[0014] In the above process, This represents the number of iterations. It is the azimuth angle. This is the Fourier transform of the snapshot.

[0015] Furthermore, in step S1, the initialized signal energy distribution function for ,

[0016]

[0017] Two-dimensional point extension function for:

[0018]

[0019] in, The Fourier transform of the autocorrelation function of the signal; The conjugate of the acoustic scattering coefficient; The speed of sound is j; j is an imaginary number. ; For wave number, , Wavelength; The number of array elements; The spacing between array elements; For the first The time delay of each scatterer, and , For the first The distance between the scatterers; For the first The orientation of the scatterer.

[0020] Furthermore, in step S2, the matched filter-beamforming energy value for:

[0021]

[0022] Furthermore, in step S3,

[0023]

[0024] The energy output after the matched filter beam is formed;

[0025] .

[0026] Let be the signal energy distribution function, and ; The time delay of the m-th scatterer is... The orientation of the m-th scatterer.

[0027] Furthermore, in step S4,

[0028]

[0029] Furthermore, in step S5,

[0030] .

[0031] Furthermore, in step S6, the condition for iterative convergence is:

[0032]

[0033] in, This represents the Csiszar difference function.

[0034] Furthermore, in step S6, the RL algorithm is used to update the signal energy distribution function. Two-dimensional deconvolution processing is performed to obtain high-resolution images in the range and azimuth directions, which is achieved through the following iterative formula:

[0035]

[0036] When the target being detected is spherical, its acoustic scattering coefficient for:

[0037] ;

[0038] In the formula, ; For wave number, and , For frequency, For the speed of sound, The radius of the sphere; It is a Bessel function of the first kind; ;

[0039] When the target being detected is a cylindrical or irregularly shaped target, the acoustic scattering angle is set using COMSOL software. Change frequency The acoustic scattering coefficient was obtained by numerical simulation. .

[0040] When the target is a cylindrical or irregularly shaped target, its surface cannot be described in an orthogonal curvilinear coordinate system, thus preventing the rigorous solution of the acoustic scattering field using the separation of variables method. The acoustic scattering field of the target is modeled and analyzed using the finite element numerical simulation software COMSOL, thereby calculating the acoustic scattering coefficients of the target model. First, an accurate geometric model of the target is created in COMSOL. Then, using COMSOL's built-in pressure acoustics module, the incident sound pressure is set. The far-field scattered sound pressure was obtained by analyzing the numerical solution based on the Helmholtz-Kirchhoff integral method. as follows:

[0041]

[0042]

[0043] In the formula, The sound pressure level is the amplitude. For wave number, and , For frequency, For the speed of sound, For phase, For spatial coordinates, For the normalized wave direction, all parameters of the incident sound pressure are settable. The sound pressure level at the field location. For the normal gradient, It is the Green's function of the sound pressure field under the model and environmental conditions built in COMSOL, and it is the position vector of the field point and the position vector of the source point. The function.

[0044] Finally, the average incident sound pressure at the boundary of the geometric model is calculated using COMSOL's averaging operator. The scattering coefficient of complex targets can then be obtained. :

[0045]

[0046] According to a second aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, the computer program being executable by a processor to implement the steps of the method described above.

[0047] According to a third aspect, the present invention also provides an active sonar for detecting submerged and buried targets, including the computer-readable storage medium described above.

[0048] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0049] (1) Spatiotemporal high-resolution processing includes two aspects: temporal high-resolution processing and spatial high-resolution processing. For targets that are submerged or buried, their acoustic reflection cross-section is small, but their reflection intensity per unit area is large, while the sedimentary layer has a large reflection cross-section, but its reflection intensity per unit area is small. Through spatiotemporal high-resolution processing, the signal-to-mixing ratio of the target echo can be improved, and the false alarm probability can be reduced.

[0050] (2) The method of the present invention has better robustness. Based on the matched filtering-conventional beamforming method, it does not require solving the optimal weight coefficients based on the adaptive method (adaptive spatiotemporal high-resolution methods require solving the optimal weight coefficients, and data quality and solution method have a great impact on the calculation of the optimal weight coefficients), and does not have strict requirements on the detection platform (synthetic aperture processing methods require the detection platform to perform uniform linear motion or uniform circular motion), and the physical meaning of the calculation results is clear.

[0051] (3) The method of the present invention has a faster calculation speed. Based on the two-dimensional convolution theorem, it is implemented in the frequency-beam Fourier transform domain, which has a faster calculation speed and can meet the real-time processing requirements in engineering. Attached Figure Description

[0052] Figure 1 This is a flowchart of the method of the present invention;

[0053] Figure 2 The diagram shows the backscattering morphology function (scattering coefficient) of the pipeline in Example 1;

[0054] Figure 3 This refers to the two-dimensional point extension function in Example 1;

[0055] Figure 4 This is a diagram showing the results of detecting buried pipes in Example 1. Detailed Implementation

[0056] The present invention will be further described in detail below through specific embodiments.

[0057] Example 1

[0058] like Figure 1 As shown, this embodiment provides a spatiotemporal high-resolution detection method to enhance the acoustic scattering characteristics of a target. In this embodiment, based on active sonar, a linear frequency modulated signal with a frequency of 10kHz~20kHz and a pulse width of 10ms is used to detect buried oil and gas pipelines. Specifically, the method includes the following steps:

[0059] Step S1: Parameter initialization.

[0060] S11: Calculate the acoustic scattering coefficient of the target. For the first layer of seafloor sediments The acoustic scattering coefficient of a scatterer when the target is spherical. for:

[0061] When the target being detected is spherical, its acoustic scattering coefficient for:

[0062] ;

[0063] In the formula, ; For wave number, and , For frequency, For the speed of sound, The radius of the sphere; It is a Bessel function of the first kind; .

[0064] When the target being detected is a cylindrical or irregularly shaped target, the acoustic scattering angle is set using COMSOL software. Change frequency The acoustic scattering coefficient was obtained by numerical simulation. .

[0065] In this embodiment, the detection target is a buried oil and gas pipeline (cylindrical), therefore, its acoustic scattering coefficient is obtained. The process is as follows:

[0066] The acoustic scattering field of the target was modeled and analyzed using the finite element numerical simulation software COMSOL, thereby calculating the acoustic scattering coefficient of the target model. First, an accurate geometric model of the target was created in COMSOL. Then, the incident sound pressure was set using COMSOL's built-in pressure acoustics module. The far-field scattered sound pressure was obtained by analyzing the numerical solution based on the Helmholtz-Kirchhoff integral method. as follows:

[0067]

[0068]

[0069] In the formula, The sound pressure level is the amplitude. For wave number, and , For frequency, For the speed of sound, For phase, For spatial coordinates, For the normalized wave direction, all parameters of the incident sound pressure are settable. The sound pressure level at the field location. For the normal gradient, It is the Green's function of the sound pressure field under the model and environmental conditions built in COMSOL, and it is the position vector of the field point and the position vector of the source point. The function.

[0070] Finally, the average incident sound pressure at the boundary of the geometric model is calculated using COMSOL's averaging operator. The scattering coefficient of complex targets can then be obtained. :

[0071]

[0072] The acoustic scattering coefficient in this embodiment like Figure 2 As shown.

[0073] S12: Initialize the signal energy distribution function Initialized signal energy distribution function for ,

[0074]

[0075] S13: Based on the acoustic scattering coefficient Calculate the two-dimensional point spread function of matched filtering-beamforming ;

[0076]

[0077] Two-dimensional point extension function like Figure 3 As shown.

[0078] The Fourier transform of the autocorrelation function of the signal; The conjugate of the acoustic scattering coefficient; The target azimuth angle ranges from -45° to 45°; j is an imaginary number. ; For wave number, , For wavelength, Calculated at 15kHz; The number of array elements. ; For the spacing between array elements, ; For the first The time delay of each scatterer, and , For the first The element is relative to the first The distance between the scatterers; For the first The element is relative to the first The orientation of the scatterer.

[0079] Step S2: Initialize the signal energy distribution function and 2D point extension function After two-dimensional FFT transformation to the frequency-wavenumber domain, we obtain and ,according to and Calculate the energy value of matched filter beamforming ;

[0080]

[0081] Step S3: Calculate the matched filter beamforming energy value The ratio of the actual matched filter beamforming energy value to the actual energy value is transformed to the frequency-wavenumber domain to obtain... ;

[0082]

[0083] The energy output after the matched filter beam is formed;

[0084]

[0085] Let be the signal energy distribution function, and , The time delay of the m-th scatterer is... The orientation of the m-th scatterer.

[0086] Step S4: According to Calculate the update rate of the signal energy distribution function ;

[0087]

[0088] Step S5: Update the signal energy distribution function according to its rate of change. Calculate the signal energy distribution function after one update. ;

[0089]

[0090] Step S6: Perform an iterative convergence check. If convergence fails, return to step S2 and begin the next iteration. Otherwise, use the RL (Richard-Lucy) algorithm to update the signal energy distribution function. Two-dimensional deconvolution processing is performed to obtain high-resolution images in the range and azimuth directions. This is achieved through the following iterative formula:

[0091]

[0092] (it) represents the number of iterations; the condition for convergence is:

[0093]

[0094] in, This represents the Csiszar difference function.

[0095] Based on the above steps, the buried pipeline detection results are achieved as follows: Figure 4 As shown, it significantly reduces the amount and amplitude of reverberation and clutter.

[0096] The above examples illustrate the present invention only to aid in understanding it and are not intended to limit the scope of the invention. Those skilled in the art can make various simple deductions, modifications, or substitutions based on the principles of this invention.

Claims

1. A detection method for enhancing the acoustic scattering characteristics of a target, characterized in that, include: Step S1: Calculate the acoustic scattering coefficient of the target. Initialize the signal energy distribution function ; According to the acoustic scattering coefficient Calculate the two-dimensional point spread function of matched filtering-beamforming ; Step S2: Initialize the signal energy distribution function and 2D point extension function After a two-dimensional FFT transformation to the frequency-wavenumber domain, we obtain and ,according to and Calculate the energy value of matched filter beamforming ; Step S3: Calculate the matched filter beamforming energy value The ratio of the actual matched filter beamforming energy value to the actual energy value is transformed to the frequency-wavenumber domain to obtain... ; Step S4: According to Calculate the update rate of the signal energy distribution function ; Step S5: Update the signal energy distribution function according to its rate of change. Calculate the signal energy distribution function after one update. ; Step S6: Perform an iterative convergence check. If convergence fails, return to step S2 and start the next iteration. Otherwise, use the RL algorithm to update the signal energy distribution function. Two-dimensional deconvolution processing is performed to obtain high-resolution images in the range and azimuth directions; In the above process, For the number of iterations, It is the azimuth angle. For the Fourier transform of the snapshot; In step S1, the initialized signal energy distribution function for , ； Two-dimensional point extension function for: ； in, This is the Fourier transform of the autocorrelation function of the signal; The conjugate of the acoustic scattering coefficient; The speed of sound is j; j is an imaginary number. ; For wave number, , Wavelength; The number of array elements; The spacing between array elements; For the first The time delay of each scatterer, and , For the first The distance between the scatterers; For the first The orientation of the scatterer.

2. The method as described in claim 1, characterized in that, In step S2, the matched filter-beamforming energy value for:

3. The method as described in claim 2, characterized in that, In step S3 ； The energy output after the matched filter beamforming; ； Let be the signal energy distribution function, and , The time delay of the m-th scatterer is... The orientation of the m-th scatterer.

4. The method as described in claim 3, characterized in that, In step S4 ； In step S5 。 5. The method as described in claim 4, characterized in that, In step S6, the condition for iterative convergence is: ； in, This represents the Csiszar difference function.

6. The method as described in claim 5, characterized in that, In step S6, the RL algorithm is used to update the signal energy distribution function. Two-dimensional deconvolution processing is performed to obtain high-resolution images in the range and azimuth directions, which is achieved through the following iterative formula:

7. The method according to any one of claims 1-6, characterized in that, In step S1, for the first layer of seabed sediment... One scatterer, When the target being detected is spherical, its acoustic scattering coefficient for: ； In the formula, ; For wave number, and , For frequency, For the speed of sound, The radius of the sphere; It is a Bessel function of the first kind; ; When the target being detected is a cylindrical or irregularly shaped target, the acoustic scattering angle is set using COMSOL software. Change frequency The acoustic scattering coefficient was obtained by numerical simulation. .

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program can be executed by a processor to implement the steps of the method as described in any one of claims 1-7.

9. An active sonar for detecting submerged and buried targets, characterized in that, Includes the computer-readable storage medium as described in claim 8.

Images