A deep sea propagation mode identification method based on sound wave angle of arrival broadband fluctuation characteristics

Abstract

The application discloses a deep-sea propagation mode identification method based on the sound wave arrival angle broadband fluctuation characteristics, and comprises the following steps: step 1, performing conventional beam forming processing on data to obtain array responses at different frequencies, and obtaining the relationship between the wave arrival angle and the frequency; step 2, based on the deduction, it is found that the convergence zone sound ray arrival angle is not sensitive to the frequency, the arrival angle fluctuates little with the frequency, the surface waveguide arrival angle is sensitive to the frequency, and the arrival angle fluctuates greatly with the frequency, and the fluctuation curve of the wave arrival angle with the frequency is extracted on the basis of the relationship between the wave arrival angle and the frequency; and step 3, performing smoothing processing on the fluctuation curve of the wave arrival angle with the frequency to reduce the non-periodic interference. The application can realize the identification of the passive receiving arrival mode.

Description

Technical fields:

[0001] This invention belongs to the field of signal processing, specifically relating to a method for identifying deep-sea propagation patterns based on the broadband fluctuation characteristics of the angle of arrival of sound waves. Background technology:

[0002] The propagation of acoustic signals in the deep sea can be mainly divided into four modes: direct wave, shadow zone, convergence zone, and surface waveguide propagation. Surface waveguide and convergence zone can both achieve long-distance propagation of underwater sound and are important long-distance propagation modes in the deep sea, thus attracting widespread attention.

[0003] Surface waveguides (SWBs) are formed by a weak positive gradient in the sound velocity profile caused by the isothermal layer at the ocean surface. Baker and Schulkin, based on experimental data, gave an empirical formula for the short-range sound propagation loss in SWBs. Porter studied the marine and acoustic characteristics of SWBs by changing characteristic parameters such as gradient, depth, and surface sound velocity. Duan et al. studied the characteristics of sound propagation in SWBs based on a normal mode model and analyzed the optimal depth for active sonar transmission and passive reception. SWBs are unstable channels with strong spatiotemporal variability. Compared with surface sound channels, deep-sea sound channels are less affected by seasonal changes, and the channel effect is more stable. Converging regions form in the deep sea under certain conditions. Hale first observed convergent regions in marine experiments and conducted a series of studies and analyses on convergent regions, giving an estimation formula for the average field strength of convergent regions. Urick analyzed the changes in convergent regions at different sound source depths and pointed out the necessary conditions for the formation of convergent regions. A.O. Williams explained the formation of convergence zones based on normal mode theory, pointing out that the convergence phenomenon is the result of the superposition of a large number of in-phase normal modes; Zhang Renhe discussed the convergence zone at the reversal point using normal mode theory and ray theory; Fan Peiqin et al. analyzed the formation mechanism of deep-sea convergence zone phenomena and gave a model for calculating the distance of convergence zones; Guo Li et al. applied the convergence zone effect to localization, giving a sound source localization method based on convergence zone sound intensity matching. The characteristics of convergence zone intensity, distance, etc., vary with the parameters of the marine underwater acoustic environment, and the signal correlation, arrival structure, and sound propagation within the convergence zone have received widespread attention. The convergence zone detection mode of sonar systems, based on the characteristic of low sound propagation loss in the convergence zone, can detect underwater targets within the convergence zone. In the deep sea, sonar needs to utilize different sound propagation modes according to the actual marine environment to achieve long-range underwater acoustic communication and detection.

[0004] However, there is currently no research on the identification of propagation modes in surface waveguides and convergence regions. Summary of the Invention:

[0005] The technical problem to be solved by the present invention is to provide a deep-sea propagation mode identification method based on the broadband fluctuation characteristics of the angle of arrival of sound waves, which can realize the identification of passive reception arrival modes.

[0006] The technical solution of this invention is to provide a method for identifying deep-sea propagation patterns based on the broadband fluctuation characteristics of sound wave angle of arrival, comprising the following steps:

[0007] Step 1: Perform conventional beamforming processing on the data to obtain the array response at different frequencies, and obtain the angle of arrival-frequency relationship;

[0008] Step 2: Based on the derivation, it was found that the angle of arrival of the sound ray in the convergence region is not sensitive to frequency and the angle of arrival fluctuates little with frequency, while the angle of arrival of the surface waveguide is more sensitive to frequency and fluctuates greatly with frequency. Based on the angle of arrival-frequency relationship, the curve of the fluctuation of the angle of arrival with frequency was extracted.

[0009] Step 3: Considering that the angle of arrival (AHA) may fluctuate with frequency due to noise interference or inaccurate estimation of conventional beamforming, the curve of AHA fluctuating with frequency is smoothed to reduce non-periodic interference.

[0010] Step 4: Quantify the degree of fluctuation of the target angle of arrival with frequency by calculating the mean square error of the angle of arrival as a function of frequency, and use it as the basis for identifying the convergence zone and surface waveguide.

[0011] As a preferred embodiment, in step 2 of the design process, the derivation of the surface waveguide angle of arrival undulation characteristics is as follows:

[0012] According to normal mode theory, the sound pressure in the far field can be represented as a superposition of a series of normal modes:

[0013]

[0014] In the formula, (r s ,z s (r,z) and (r,z) are the positions of the target and the receiver, respectively. Let k be the mode function of the m-th normal mode. m R is the horizontal wavenumber of the corresponding mode; M is the mode number; r s We can assume it to be 0; the signals received by each element of the vertical line array are represented as follows:

[0015] x = [p(r, z1; z] ] s ),p(r,z2;z s ),...,p(r,z N ;z s )] T (2)

[0016] In the formula, z i N represents the element depth, and N represents the number of elements. T The transpose operator is used; the output of array beamforming is represented as:

[0017] y = w(θ)H x (3)

[0018] In the formula, w(θ) is the weighting vector, θ is the guiding angle, and () H As the conjugate transpose operator, in the desired direction, the weighted vector is represented as:

[0019] w(θ) = [1, e ikdsinθ ,e ik2dsinθ ,...,e ik(N-1)dsinθ ] T (4)

[0020] In the formula, k is the signal wavenumber, and d is the element spacing; the output energy of conventional beamforming is expressed as...

[0021] P(θ=yy H =w(θ) H xx H w(θ) (5)

[0022] Substituting equations (1) and (2) into equation (3) yields

[0023]

[0024] remember

[0025]

[0026] Equation (6) is expressed as

[0027]

[0028] In the formula, A m Let Ψ be the amplitude of the m-th mode. m To sample the space of the m-th mode, substituting equation (9) into equation (5), we can obtain...

[0029]

[0030] Equation (10) can be converted into two parts:

[0031]

[0032] In the formula, the first part is the response of the array to all modes, and the second part is the coupling response of the modes; from equation (8), we can obtain

[0033]

[0034] When i = j It only contains 1 / r and 1 / k i Since there are two variables, this term has no effect on the angle of arrival; when i≠j, and It is a complex number containing simple harmonic terms that vary with distance and horizontal wavenumber difference. Therefore, when the distance remains constant, the angle of arrival will change with (k). i -k j Fluctuations. The conjugate terms are combined and expressed as a cosine function:

[0035] T i,j =C i,j cos((k i -k j )r+φ i,j (13)

[0036] In the formula, C i,j For amplitude, φ i,j For the initial phase, the expressions are as follows:

[0037]

[0038] Therefore, equation (11) can be expressed as

[0039]

[0040] The first term in the above equation expresses the average value of the angle of arrival as a function of distance and frequency; the second term represents the magnitude of the fluctuations with distance and horizontal wavenumber difference; the relationship between frequency and horizontal wavenumber difference is expressed as follows:

[0041]

[0042] In the formula, θ i and θ j are the angles of arrival for the i-th and j-th modes, respectively.

[0043] When the target distance remains constant, the period of the angle of arrival fluctuation can be obtained from the second term of formula (16) as follows:

[0044] k i -k j =2π / r (18)

[0045] Substituting equation (18) into equation (17) yields

[0046]

[0047] In the formula, f i,j To correspond to the fluctuation periods of the i-th and j-th modes, we can consider f i,j This corresponds to the fluctuation period of the angle of arrival.

[0048] As a preferred option, in the design process of step 2, the derivation of the convergence zone arrival angle fluctuation characteristics is as follows:

[0049] The condition for the formation of a convergence region is that the difference in horizontal wavenumbers between adjacent normal modes within a specified range remains a relatively stable constant.

[0050] |k m+1 -k m |≈C (20)

[0051] Then at a certain horizontal distance r, there is

[0052] |k m+1 rk m r|≈2π (21)

[0053] Further expressed as

[0054] |k i -k j |r≈2π(ij) (22)

[0055] At this point, within this distance range, a subset of the normal modes superimposes to form a periodic convergence. Substituting equation (22) into equation (16), we can obtain the energy output of the conventional beamforming signal in the convergence region as follows:

[0056]

[0057] Comparing equations (24) and (16), it can be seen that the second term of the energy output of the conventional beamforming signal in the convergence zone does not fluctuate with distance and frequency. Therefore, the angle of arrival in the convergence zone does not change periodically with frequency.

[0058] Therefore, in theory, convergence zones and surface waveguides can be identified by observing fluctuations in the frequency domain of the angle of arrival.

[0059] Compared with the prior art, the present invention has the following advantages after adopting the above solution:

[0060] This invention addresses the need for deep-sea propagation mode identification in long-range underwater acoustic communication and detection. Based on normal mode theory, it derives the broadband angle of arrival fluctuation characteristics under different deep-sea propagation modes and discovers that the angle of arrival under surface waveguides and convergence zones exhibits different fluctuation patterns with frequency. This pattern can provide guidance for passive reception arrival mode judgment and realizes the identification of two deep-sea propagation modes: convergence zone and surface waveguide. Attached image description:

[0061] Figure 1 This describes the processing flow of a deep-sea propagation mode identification method based on the broadband fluctuation characteristics of the sound wave angle of arrival.

[0062] Figure 2 This is a schematic diagram of the simulation environment.

[0063] Figure 3Angle of arrival-frequency relationship diagram for the target in the convergence zone.

[0064] Figure 4 The wave arrival angle-frequency relationship diagram for surface waveguide targets.

[0065] Figure 5 Angle of arrival (AHA) of the target in the convergence zone fluctuates with frequency.

[0066] Figure 6 The graph shows the wave arrival angle of a surface waveguide target as a function of frequency.

[0067] Figure 7 The mean squared variance of the angle of arrival at different distances. Detailed implementation method:

[0068] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0069] like Figure 1 As shown, the deep-sea propagation pattern identification method based on the broadband fluctuation characteristics of the sound wave angle of arrival of the present invention includes the following steps:

[0070] Step 1: Perform conventional beamforming processing on the data to obtain the array response at different frequencies, and obtain the angle of arrival-frequency relationship;

[0071] Step 2: Based on the derivation, it was found that the angle of arrival of the sound ray in the convergence region is not sensitive to frequency and the angle of arrival fluctuates little with frequency, while the angle of arrival of the surface waveguide is more sensitive to frequency and fluctuates greatly with frequency. Based on the angle of arrival-frequency relationship, the curve of the fluctuation of the angle of arrival with frequency was extracted.

[0072] Step 3: Considering that the angle of arrival (AHA) may fluctuate with frequency due to noise interference or inaccurate estimation of conventional beamforming, the curve of AHA fluctuating with frequency is smoothed to reduce non-periodic interference.

[0073] Step 4: Quantify the degree of fluctuation of the target angle of arrival with frequency by calculating the mean square error of the angle of arrival as a function of frequency, and use it as the basis for identifying the convergence zone and surface waveguide.

[0074] In step 2 of the design process, the derivation of the surface waveguide angle-of-arrival undulation characteristics is as follows:

[0075] According to normal mode theory, the sound pressure in the far field can be represented as a superposition of a series of normal modes:

[0076]

[0077] In the formula, (r s ,z s (r,z) and (r,z) are the positions of the target and the receiver, respectively. Let k be the mode function of the m-th normal mode.m R is the horizontal wavenumber of the corresponding mode; M is the mode number; r s We can assume it to be 0; the signals received by each element of the vertical line array are represented as follows:

[0078] x = [p(r, z1; z] ] s ),p(r,z2;z s ),…,p(r,z N ;z s )] T (25)

[0079] In the formula, z i N represents the element depth, and N represents the number of elements. T The transpose operator is used; the output of array beamforming is represented as:

[0080] y = w(θ) H x (26)

[0081] In the formula, w(θ) is the weighting vector, θ is the guiding angle, and () H As the conjugate transpose operator, in the desired direction, the weighted vector is represented as:

[0082] w(θ) = [1, e ikdsinθ ,e ik2dsinθ ,...,e ik(N-1)dsinθ ] T (27)

[0083] In the formula, k is the signal wavenumber, and d is the element spacing; the output energy of conventional beamforming is expressed as...

[0084] P(θ=yy H =w(θ) H xx H w(θ) (28)

[0085] Substituting equations (1) and (2) into equation (3) yields

[0086]

[0087] remember

[0088]

[0089] Equation (6) is expressed as

[0090]

[0091] In the formula, A m Let Ψ be the amplitude of the m-th mode. m To sample the space of the m-th mode, substituting equation (9) into equation (5), we can obtain...

[0092]

[0093] Equation (10) can be converted into two parts:

[0094]

[0095] In the formula, the first part is the response of the array to all modes, and the second part is the coupling response of the modes; from equation (8), we can obtain

[0096]

[0097] When i = j It only contains 1 / r and 1 / k i Since there are two variables, this term has no effect on the angle of arrival; when i≠j, and It is a complex number containing simple harmonic terms that vary with distance and horizontal wavenumber difference. Therefore, when the distance remains constant, the angle of arrival will change with (k). i -k j Fluctuations. The conjugate terms are combined and expressed as a cosine function:

[0098] T i,j =C i,j cos((k i -k j )r+φ i,j (36)

[0099] In the formula, C i,j For amplitude, φ i,j For the initial phase, the expressions are as follows:

[0100]

[0101] Therefore, equation (11) can be expressed as

[0102]

[0103] The first term in the above equation expresses the average value of the angle of arrival as a function of distance and frequency; the second term represents the magnitude of the fluctuations with distance and horizontal wavenumber difference; the relationship between frequency and horizontal wavenumber difference is expressed as follows:

[0104]

[0105] In the formula, θ i and θ j are the angles of arrival for the i-th and j-th modes, respectively.

[0106] When the target distance remains constant, the period of the angle of arrival fluctuation can be obtained from the second term of formula (16) as follows:

[0107] k i -k j =2π / r (41)

[0108] Substituting equation (18) into equation (17) yields

[0109]

[0110] In the formula, f i,j To correspond to the fluctuation periods of the i-th and j-th modes, we can consider f i,j This corresponds to the fluctuation period of the angle of arrival.

[0111] As a preferred option, in the design process of step 2, the derivation of the convergence zone arrival angle fluctuation characteristics is as follows:

[0112] The condition for the formation of a convergence region is that the difference in horizontal wavenumbers between adjacent normal modes within a specified range remains a relatively stable constant.

[0113] |k m+1 -k m |≈C (43)

[0114] Then at a certain horizontal distance r, there is

[0115] |k m+1 rk m r|≈2π (44)

[0116] Further expressed as

[0117] |k i -k j |r≈2π(ij) (45)

[0118] At this point, within this distance range, a subset of the normal modes superimposes to form a periodic convergence. Substituting equation (22) into equation (16), we can obtain the energy output of the conventional beamforming signal in the convergence region as follows:

[0119]

[0120] Comparing equations (24) and (16), it can be seen that the second term of the energy output of the conventional beamforming signal in the convergence zone does not fluctuate with distance and frequency. Therefore, the angle of arrival in the convergence zone does not change periodically with frequency.

[0121] Therefore, in theory, convergence zones and surface waveguides can be identified by observing fluctuations in the frequency domain of the angle of arrival.

[0122] Figure 2The simulation environment is as follows: The simulation conditions are a 5000m deep sea, a 32-element vertical array receiver with an element spacing of 2m, an array center depth of 50m, a sound source depth of 50m, and a frequency range of 250Hz–350Hz. Two similar simulation environments are shown in the figure, the only difference being the presence of a 100m thick surface waveguide in the sound velocity profile indicated by the dashed line. In this simulation environment, a normal mode model is used for sound field simulation modeling, with the convergence zone ranging from 63km to 67km.

[0123] Figure 3 An angle-of-arrival (AOA)-frequency relationship diagram for targets in the convergence zone; Figure 4 This is the angle-of-arrival (AOA)-frequency relationship diagram for a surface waveguide target. The acoustic field at different frequencies of the surface waveguide and the convergence zone at a distance of 67 km was simulated and modeled using Kraken, obtaining the sound pressure level data of the array's received signal. Conventional beamforming processing was then performed on the data to obtain the array response at different frequencies, resulting in the AOA-frequency relationship diagram, as shown below. Figure 3 and Figure 4 As shown, the angle of arrival of the sound rays in the convergence region is not sensitive to frequency, and the angle of arrival fluctuates little with frequency. However, the angle of arrival of the surface waveguide is more sensitive to frequency, and the angle of arrival fluctuates greatly with frequency.

[0124] Figure 5 Angle of arrival (AHA) of a target in the convergence zone as a function of frequency. Figure 6 This is a frequency fluctuation plot for the angle of arrival (AHA) of a surface waveguide target. Based on the AHA-frequency plot, the fluctuation of AHA with frequency is extracted. Considering that the AHA with frequency can fluctuate due to noise interference or inaccurate estimation in conventional beamforming, a smoothing process is performed to reduce non-periodic interference, resulting in the following: Figure 5 and Figure 6 In the figure, the angle of arrival (AHA) of the convergence zone approximates a horizontal straight line with frequency, remaining constant regardless of frequency, with an AHA of 3.5°. The AHA of the surface waveguide fluctuates around 0° with an amplitude of 3°, showing a periodic change with frequency. As derived in this paper, the period of the AHA of the surface waveguide is a quantity that varies with distance and frequency. Therefore, identifying the surface waveguide and convergence zone by calculating the period is difficult in practical applications. A better approach is to quantify the fluctuation of the target's AHA with frequency as a basis for identifying the convergence zone and the surface waveguide. Calculating the root mean square (RMS) of the AHA variation with frequency in the figure, the MMS of the convergence zone target is 0; the MMS of the surface waveguide target is 1.64, demonstrating clear separability.

[0125] Figure 7The mean square error of the angle of arrival fluctuation at different distances is used. To investigate the impact of signal-to-noise ratio (SNR) variation on the propagation mode identification method, simulations were conducted at different distances. To ensure that the selected distances covered both the convergence zone and the surface waveguide, 21 distances were selected between 63km and 67km in the convergence zone, with each distance point spaced 200m apart; 21 distances were also selected between 63km and 67km in the surface waveguide, with each distance point spaced 200m apart; and then ten more distances were selected between 10km and 100km. The input SNR was 0dB, and other environmental conditions were the same as those used in simulations. Figure 1 The mean square error of the angle of arrival fluctuation at points of the same or different distances is as follows: Figure 7 As shown in the figure, the mean square error of the angle of arrival undulation of the surface waveguide and the convergence region is clearly separable at different distances.

[0126] This invention discloses that the angle of arrival under surface waveguides and convergence zones exhibits different fluctuation patterns with frequency. This pattern can provide guidance for passive receiver arrival mode determination, thereby enabling deep-sea propagation mode identification.

[0127] 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 method for identifying deep-sea propagation patterns based on the broadband fluctuation characteristics of sound wave angle of arrival, characterized in that: Includes the following steps, Step 1: Process the data to obtain the array response at different frequencies and obtain the angle of arrival-frequency relationship; Step 2: Based on the derivation, extract the curve of the fluctuation of the angle of arrival with frequency based on the angle of arrival-frequency relationship; Step 3: Smooth the curve of the angle of arrival as a function of frequency to reduce non-periodic interference. Step 4: Quantify the degree of fluctuation of the target angle of arrival with frequency by calculating the mean square error of the angle of arrival as a function of frequency, and use it as the basis for identifying the convergence zone and surface waveguide.

2. The deep-sea propagation pattern identification method based on the broadband fluctuation characteristics of the sound wave angle of arrival as described in claim 1, characterized in that: In step 2 of the design process, the derivation of the surface waveguide angle-of-arrival undulation characteristics is as follows: According to normal mode theory, the sound pressure in the far field can be represented as a superposition of a series of normal modes: In the formula, (r s ,z s (r,z) and (r,z) are the positions of the target and the receiver, respectively. Let k be the mode function of the m-th normal mode. m R is the horizontal wavenumber of the corresponding mode; M is the mode number; r s We can assume it to be 0; the signals received by each element of the vertical line array are represented as follows: x=[p(r,z1;z s ),p(r,z2;z s ),...,p(r,z N ;z s )] T (2) In the formula, z i N represents the element depth, and N represents the number of elements. T The transpose operator is used; the output of array beamforming is represented as: y=w(θ) H x (3) In the formula, w(θ) is the weighting vector, θ is the guiding angle, and () H As the conjugate transpose operator, in the desired direction, the weighted vector is represented as: w(θ)[1,e ikdsinθ ,e ik2dsinθ ,...,e ik(N-1)dsinθ ] T (4) In the formula, k is the signal wavenumber, and d is the element spacing; the output energy of conventional beamforming is expressed as... P(θ)=yy H =w(θ) H xx H w(θ) (5) Substituting equations (1) and (2) into equation (3) yields remember Equation (6) is expressed as In the formula, A m Let Ψ be the amplitude of the m-th mode. m To sample the space of the m-th mode, substituting equation (9) into equation (5), we can obtain... Equation (10) can be converted into two parts: In the formula, the first part is the response of the array to all modes, and the second part is the coupling response of the modes; from equation (8), we can obtain When i = j It only contains 1 / r and 1 / k i Since there are two variables, this term has no effect on the angle of arrival; when i≠j, and It is a complex number containing simple harmonic terms that vary with distance and horizontal wavenumber difference. Therefore, when the distance remains constant, the angle of arrival will change with (k). i -k j Fluctuations, combining conjugate terms to represent a cosine function: T i,j =C i,j cos((k i -k j )r+φ i,j ) (13) In the formula, C i,j For amplitude, φ i,j For the initial phase, the expressions are as follows: Equation (11) can be expressed as The first term in the above equation expresses the average value of the angle of arrival as a function of distance and frequency; the second term represents the magnitude of the fluctuations with distance and horizontal wavenumber difference; the relationship between frequency and horizontal wavenumber difference is expressed as follows: In the formula, θ i and θ j These are the angles of arrival for the i-th and j-th modes, respectively; When the target distance remains constant, the period of the angle of arrival fluctuation can be obtained from the second term of formula (16) as follows: k i -k j =2π / r (18) Substituting equation (18) into equation (17) yields In the formula, f i,j To correspond to the fluctuation periods of the i-th and j-th modes, we can consider f i,j This corresponds to the fluctuation period of the angle of arrival.

3. The deep-sea propagation pattern identification method based on the broadband fluctuation characteristics of the sound wave angle of arrival as described in claim 1, characterized in that: In step 2 of the design process, the derivation of the convergence zone arrival angle fluctuation characteristics is as follows: The condition for the formation of a convergence region is that the difference in horizontal wavenumbers between adjacent normal modes within a specified range remains a relatively stable constant. |k m+1 -k m |≈C (20) Then at a certain horizontal distance r, there is |k m+1 r-k m r|≈2π (21) Further represented as |k i -k j |r≈2π(ij) (22) At this point, within this distance range, a subset of the normal modes superimposes to form a periodic convergence. Substituting equation (22) into equation (16), we can obtain the energy output of the conventional beamforming signal in the convergence region as follows:

4. The deep-sea propagation pattern identification method based on the broadband fluctuation characteristics of the angle of arrival of sound waves according to claim 1, characterized in that: In step 2, based on the derivation, the angle of arrival of the converging region's acoustic rays fluctuates less with frequency compared to the angle of arrival of the surface waveguide.

5. The deep-sea propagation pattern identification method based on the broadband fluctuation characteristics of the angle of arrival of sound waves according to claim 1, characterized in that: In step 1, the data is processed using conventional beamforming to obtain the array response at different frequencies, thus obtaining the angle of arrival-frequency relationship.

Images